Document of the World Bank Report No: ICR00003518 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-40180 IDA-47560) ON A CREDIT IN THE AMOUNT OF SDR 151.50 MILLION (US$ 210.42 MILLION EQUIVALENT) TO THE REPUBLIC OF INDIA FOR A TAMIL NADU HEALTH SYSTEMS PROJECT June 27, 2016 Health, Nutrition and Population Global Practice (GHNDR) South Asia Region CURRENCY EQUIVALENTS (Exchange Rate Effective May 31, 2016) Currency Unit = Rupees (Rs) Rs 67.29 = US$ 1.00 US$ 1.00 = SDR 0.71 FISCAL YEAR: April 1 – March 31 ABBREVIATIONS AND ACRONYMS AF Additional Financing AIDS Acquired Immune Deficiency Syndrome AMCs Annual Maintenance Contracts ANMs Auxiliary Nurse Midwifes BCC Behavior Change Communication BMEs Biomedical Engineers CAG Comptroller and Auditor General CD Country Director CEmONC Comprehensive Emergency Obstetric and Neonatal Care CMS College Management System C-section Caesarean Section CTFs Common Treatment Facilities CVD Cardio Vascular Disease DCA Development Credit Agreement DIR Detailed Implementation Review DM&RHS Directorate of Medical & Rural Health Services DMS Directorate of Medical Services DO Development Objective DoHFW Department of Health and Family Welfare DPH Directorate of Public Health EmONCs Emergency Obstetric and Neonatal Care EMRI Emergency Management and Research Initiative FM Financial Management GAAP Governance and Accountability Action Plan GoTN Government of Tamil Nadu HCWM Health Care Waste Management HMIS Health Management Information System HMS Hospital Management System IAS Indian Administrative Service ICR Implementation Completion and Results Report ICU Intensive Care Unit ICDS Integrated Child Development Scheme ICWM Infection Control and Waste Management IDA International Development Association IEC Information, Education and Communication IMR Infant Mortality Rate IO Intermediate Outcome IP Implementation Progress ISMR Institutional Services Monitoring Report i ISRs Implementation Status and Results Reports IT Information Technology IUFRs Interim Unaudited Financial Reports M&E Monitoring and Evaluation MCH Maternal and Child Health MIS Management Information System MMR Maternal Mortality Ratio MTR Mid-term Review NABH National Accreditation Board for Hospitals NCD Non-communicable Disease NGOs Non-governmental Organizations NHM National Health Mission NIE National Institute of Epidemiology OBGYN Obstetrician and Gynecologist OPD Out-Patient Department PAD Project Appraisal Document PDO Project Development Indicators PHCs Primary Health Centers PINs Patient Identification Numbers PMU Project Management Unit PP Project Paper PPPs Public Private Partnerships PWD Public Works Department QAG Quality Assurance Group QCE Quality Circle of Excellence RCH Reproductive and Child Health RF Results Framework SC/ST Scheduled Caste/Scheduled Tribe SCA Sickle Cell Anemia SHRDC State Health Data Resource Center SPU Strategic Planning Unit SPC Strategic Planning Cell TDP Tribal Development Plan TNCDW Tamil Nadu Corporation for Development of Women TNHSP Tamil Nadu Health Systems Project TNMSC Tamil Nadu Medical Services Corporation TOR Terms of Reference TPA Third Party Administrator UAS University Automation System Global Practice Director: Olusoji Adeyi Country Director: Onno Ruhl Practice Manager: Rekha Menon Project Team Leader: Bushra Binte Alam ICR Team Leader: Sangeeta C. Pinto ICR Authors: Joy de Beyer/Surendra Agarwal/Owen Smith ii INDIA - TAMIL NADU HEALTH SYSTEMS PROJECT Table of CONTENTS A. Basic Information....................................................................................................... v B. Key Dates ................................................................................................................... v C. Ratings Summary ....................................................................................................... v D. Sector and Theme Codes .......................................................................................... vi E. Bank Staff .................................................................................................................. vi F. Results Framework Analysis .................................................................................... vii G. Ratings of Project Performance in ISRs ............................................................... xxiii H. Restructuring (if any) ............................................................................................ xxiv I. Disbursement Profile ............................................................................................. xxiv 1. Project Context, Development Objectives and Design ............................................... 1 1.1. Context at Appraisal ......................................................................................... 1 1.2. Original Project Development Objective (PDO) and Key Indicators (as approved) .................................................................................................................... 3 1.3. Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification............................................................................ 3 1.4. Main Beneficiaries ............................................................................................ 5 1.5. Original Components (as approved) ................................................................ 5 1.7. Other significant changes ................................................................................. 9 2. Key Factors Affecting Implementation and Outcomes ......................................... 10 2.1. Project Preparation, Design and Quality at Entry........................................... 10 2.2. Implementation ............................................................................................... 13 2.3. Monitoring and Evaluation (M&E) Design, Implementation and Utilization 17 2.4. Safeguard and Fiduciary Compliance............................................................. 21 2.5. Post-completion Operation/Next Phase .......................................................... 23 3. Assessment of Outcomes ....................................................................................... 24 3.1. Relevance of Objectives, Design and Implementation ................................... 24 iii 3.2. Achievement of Project Development Objectives (rating: Phase 1 – Substantial; Phase 2 – High) ..................................................................................... 27 3.3. Efficiency (rating: Phase 1 – Substantial; Phase 2 - High)............................. 36 3.4. Justification of Overall Outcome Rating ........................................................ 37 3.5. Overarching Themes, Other Outcomes and Impacts ...................................... 38 3.6. Summary of Findings of Beneficiary Survey and/or Stakeholders Workshops 40 4. Assessment of Risk to Development Outcome ...................................................... 40 5. Assessment of Bank and Borrower Performance .................................................. 41 5.1. Bank Performance .......................................................................................... 41 5.2. Borrower Performance ................................................................................... 42 6. Lessons Learned..................................................................................................... 43 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners........ 46 Annex 1. Project Costs and Financing .......................................................................... 47 Annex 2. Outputs by Component ................................................................................. 48 Annex 3. Economic and Financial Analysis ................................................................. 63 Annex 4. Bank Lending and Implementation Support/Supervision Processes ............ 68 Annex 5. Beneficiary Survey Results ........................................................................... 70 Annex 6. Stakeholder Workshop Report and Results................................................... 71 Annex 7. Summary of Borrower's ICR ......................................................................... 82 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ....................... 85 Annex 9. Details on NSS 2004 and NSS 2014 Data on Access and utilization of health services by poorest 40% and scheduled tribe (ST) populations in Tamil Nadu ........... 86 Annex 10. Tamil Nadu Key Indicators – National Family Health Surveys 2015 and 2005 .............................................................................................................................. 93 Annex 11. List of Supporting Documents .................................................................... 96 iv A. Basic Information India: Tamil Nadu Country: India Project Name: Health Systems Project IDA-40180, IDA- Project ID: P075058 L/C/TF Number(s): 47560 ICR Date: ICR Type: Intensive Learning ICR GOVERNMENT OF Lending Instrument: SIL Borrower: INDIA Original Total XDR 73.90M Disbursed Amount: XDR 137.75M Commitment: Revised Amount: XDR 151.5M Environmental Category: B Implementing Agencies: Tamil Nadu Health Systems Project Project Management Unit (TNHSP PMU), Department of Health and Family Welfare (DoHFW), Tamil Nadu Medical Services Corporation (TNMSC), Public Works Department (PWD) Cofinanciers and Other External Partners: n/a B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 03/31/2003 Effectiveness: 01/27/2005 01/27/2005 05/18/2007 02/19/2010 04/29/2010 Appraisal: 06/28/2004 Restructuring(s): 06/28/2010 05/08/2013 08/07/2014 Approval: 12/16/2004 Mid-term Review: 11/26/2007 11/21/2007 Closing: 09/30/2010 09/15/2015 C. Ratings Summary C.1. Performance Rating by ICR Outcomes: Highly Satisfactory Risk to Development Outcome: Negligible Bank Performance: Highly Satisfactory Borrower Performance: Highly Satisfactory Note: The Outcome rating is S for the first phase, and HS for the second phase. Under the methodology for weighting ratings, the overall outcome score is 5.57, which rounds up to 6, HS, even though there were some small shortcomings in achievement of outcomes. v C.2. Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Highly Satisfactory Implementing Quality of Supervision: Highly Satisfactory Satisfactory Agency/Agencies: Overall Bank Overall Borrower Highly Satisfactory Highly Satisfactory Performance: Performance: C.3. Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Project Quality at Entry No None at any time (Yes/No): (QEA): Problem Project at any Quality of QALP-1 rating 2 (Likely Yes time (Yes/No): Supervision (QSA): to achieve DO) DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Health 80 80 Other social services 1 1 Sub-national government administration 19 19 Theme Code (as % of total Bank financing) Child health 17 17 Health system performance 33 33 Indigenous peoples 16 16 Injuries and non-communicable diseases 17 17 Population and reproductive health 17 17 E. Bank Staff Positions At ICR At Approval Praful Patel (Original Vice President: Annette Dixon Credit)/Isabel M. Guerrero (Additional Financing) Country Director: Onno Ruhl Michael Carter/N Roberto Zagha Practice Manager/ Anabel Abreu Rekha Menon Manager: AF: Julie McLaughlin Project Team Leader: Bushra Binte Alam Preeti Kudesia vi ICR Team Leader: Sangeeta C. Pinto ICR Primary Author: Joy de Beyer ICR major contributors Surendra Agarwal, Owen Smith F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document, PAD, p. 5): To significantly improve the effectiveness of the health system, both public and private, in Tamil Nadu through: (i) increased access to and utilization of health services, particularly by poor, disadvantaged and tribal groups; (ii) development and pilot testing of effective interventions to address key health challenges specifically non- communicable diseases; (iii) improved health outcomes, access and quality of service delivery through strengthened oversight of the public sector health systems and greater engagement of non-governmental sector; and (iv) increased effectiveness of public sector hospital services, primarily at district and sub-district levels. The PDO statement in the Development Credit Agreement (DCA, p. 17) is worded slightly differently: “public and private” is omitted from the main clause of the objective statement; item (ii) omits the reference to piloting interventions and (iv) omits “district and sub-district levels”. The changes make the statement a little less precise. Revised PDO July 2010 at Additional Financing (AF) (Project Paper PP, p. 6): To significantly improve the effectiveness of the health system in Tamil Nadu as measured by: (i) increased access to and utilization of maternal and neo-natal care services, particularly by poor, disadvantaged and tribal groups; (ii) effective non-communicable disease interventions scaled up throughout the state; (iii) improved health outcomes, access and quality of service delivery through strengthened oversight of the public sector health systems and greater engagement of non-governmental sector; and (iv) increased effectiveness of public sector hospital services, primarily at district and sub-district levels. Results Framework, baselines and Actual Values In May 2007 (Management Letter and AM 5/18/2007), the Country Director (CD) approved the revisions to the results framework and monitoring matrix to incorporate recommendations of a Bank-wide review of results monitoring frameworks and outcome indicators (completed in 2006). Minor deletions to the original PDO indicators are indicated in [square brackets], baseline data were added, two Intermediate Outcome (IO) Indicators were deleted (road traffic accident case fatality rate, and doctor absenteeism) and a new IO Indicator was added “Evaluation of pilots being implemented to enhance management of project facilities in terms of quality of care”. The results framework documents three sets of changes to the PAD Results Framework: Additional Financing (April 29 2010): Three PDO indicators were added to monitor quality of inpatient care, supply and equipment management, and patient satisfaction; one quality PDO indicator changed, and one PDO indicator change reflected the progression vii from NCD pilots to scaling up. Numerous changes in IO indicators were made to measure specific outputs and activities, indicators were dropped that duplicated the PDO indicators or were no longer relevant, and “Core Indicators” added as required by new Bank-wide guidelines. Restructuring: August 2014 restructuring paper lists 8 dropped, 2 new, 7 unchanged PDO indicators, 37 dropped IOIs/parts of IOIs, 11 new IOIs and one revised IOI. Please note: The order in which of some indicators are presented try make the table easier to follow (e.g. AF indicator for NCD activities follows PAD original indicator 2 which also refers to NCDs) so some indicator numbers differ from their numbers in the PAD. Indicators that were included in the original or revised project both as PDO and IOIs are reported once. (a) PDO Indicators Original Target Formally Actual Value Achieved Values (from Revised Indicator Baseline Value at Completion or approval Target Target Years documents) Values Indicator 1: Original. Total in-patient utilization (considering both the public and private sector) by the Dropped in 2014 poorest 40% of the population increased (as measured by an asset mix) Surgery with overnight 2.45% 14% stay: Other hospital stays: 1.41% 10% increase by 9% (Source: Ferguson 2008 (IPSOS Patient Patient Satisfaction Satisfaction Survey of Hospitalized cases per Survey, March 2007). 20% increase by 2015). 1,000 persons in last 2010 365 days by monthly per capita consumption 41.9 cases per 1,000 expenditure pattern for 32.1 cases per 1,000 (Source National Sample (National Sample Survey lower 40% MPCE Survey Organization – Organization – NSSO group NSSO 60th round (2004) 71st round (2014). Proportion of (poorest 40%) population 13.2% ailing (NSSO) reporting any ailment 8.5% ailing (NSSO) in last 15 days, percent 98% of those who accessed 76% any form of care Date 2004 and March 2007 9/30/2010 2014 SURPASSED. Increases in inpatient care were far above the 20% goal: 571% increase in surgery with overnight stay, 638% increase in other hospital stays, and 31% increase in hospitalization rate per 1,000 people in lowest 40% income group. Percent of those Comments who reported any ailment and accessed care increased by 29%, well above the original 20% target. NSSO 2014 excluded pre-existing disability from “any ailment”, 2004 had included it so data are not strictly comparable, but unlikely to make much difference. viii Indicator 2: 2007, AF (Originally an IOI. Not Increased utilization of out-patient and in-patient services by Tribal Groups included in 2014. Proportion of tribal 8.5 per 1,000 (all TN) 13.2 per 1,000 (all TN) population reporting 0.8% (ST) No target set for No target set 10.3% (ST) ailment in last 15 days increase for increase of these, percentage 76% (all TN) 97.5% (all TN) who accessed any form 96.2% (ST) 93.3% (ST) of care Hospitalized cases 13.1 per 1000 15.5/1000 th Date 2004 (NSSO 60 round) 9/30/2013 8/30/2014 2014 (NSSO 71st round) ACHIEVED. PAD suggested tentative target of 30% increase over baseline, to be agreed after baseline value determined (not done). The 18% increase in ST hospitalization is substantial, and 13 fold increase in ST reporting ailment shows Comments increased recognition of need for care, and makes the small fall in % of those reporting an ailment who accessed care less worrying – it still indicates a very large increase in utilization of services. The 2004 baseline data likely indicate very low recognition of symptoms/need for care rather than very low illness incidence. Completion of two rigorously evaluated pilots of clinic-based NCD prevention and Indicator 3: Original, control, [careful monitoring of the effectiveness of other NCD prevention activities,] modified in 2007 and assessment of the impact of these pilots on the development of a state-wide policy Cervical cancer and In 2007, hypertension pilots dropped completed, monitored and Pilots completed “careful Pilots, carefully analyzed. and rigorously monitoring of Evaluations, NA (Not Applicable) Rigorous assessment by evaluated. effectiveness Policy NIE led to policy decision Policy developed of other NCD to scale-up cervical cancer prevention and CVD interventions activities” state-wide. Date achieved 2004 2008 2007 05/10/2010 ACHIEVED. Cervical cancer pilot in Theni and Thanjavar districts (Feb. 2007-Jan. 2010): 488,084 targeted women (30-60 years) screened (94.2% of women in the target group in the two districts); those testing positive and confirmed were referred for treatment. Hypertension pilot in Sivagangai and Virudhnagar districts (Oct 2007- March 2010): 1.231 million adults in the target group were screened; 77,757 new cases Comments were diagnosed (suffering from hypertension) and provided treatment and followed up. Analysis by TNHSP found substantial improvements in diastolic and systolic blood pressures among clinic patients who were regularly followed up. Adults screened for hypertension also received counseling on life-style modification. In 2007, [careful monitoring of the effectiveness of other NCD prevention activities] was dropped from the indicator, to align with the main focus of the project. Indicator 4 – new at Scale-up of cancer cervix screening and cardio-vascular disease prevention and control AF (replaced OI 2) based on a comprehensive assessments of the pilots ix Cervical cancer screening ISR #13 notes pilot operational in Theni Clinical screening and that GoTN and Thanjavar districts, No explicit target follow-up treatment of added urban and interventions for set for number of hypertension, diabetes, areas to the Programs scaled up cardiovascular disease districts in which cancers of cervix and scale up plan. (CVD) prevention and to implement the breast scaled up to all 32 End date control pilot operational programs. districts in TN. changed in in Sivagangai and 2014 Virudhnagar districts. Date achieved April 2010 9/30/2013 9/15/2015 9/15/2015 ACHIEVED. In addition, (a) Preventive school based activities scaled up in 16,369 government and aided schools under the Sarva Shiksha Abhiyaan Program. Comments (b) Workplace interventions implemented in 400 worksites; and (c) Community based interventions through TNCDW reached 250,476 (97%) TN women’s self-help groups. Improved quality of care (QOC) in public hospitals as measured by a series of Indicator 5: Original, indicators [and implementation of a regulation/ accreditation system to improve quality modified in 2007 of care in private sector hospitals] (i) Bed occupancy rate 80% Maintain 81% (ii) Number of major surgeries 211,988 No target 125,537 (iii) Number of diagnostic services 20,031,677 10% improvement 25,842,226 (iv) Number of night time caesarians at 6,817 10% improvement 11,406 (12 months) CEmONCs Date achieved 2004 - 2005 9/30/2010 2009-2010 Baseline data for major surgeries (2004-2005) was an outlier -- annual data thereafter were 60-80% of the baseline level. In any case, this and bed occupancy measure hospital efficiency, not quality. SURPASSED targets for quality indicators (ii) Number of diagnostic tests increased by 29%, nearly 3 times the target; (iii) night time C-sections (indicating 24x7 Comments functionality of CEmONCs) increased 167%, 16.7 times the 10% target. Source: Institutional Services Monitoring Reports (ISMRs) prepared by hospitals. [Regulation/accreditation system for private sector] dropped in 2007, GoTN had intended to develop its own system, but decided instead to use the existing (fairly new) system of the National Accreditation Board of Hospitals, and to focus on quality in public hospitals, which are answerable to GoTN. (See text for discussion of weaknesses in this and several other indicators.) AF version of Improved quality of care as measured by (i) bed occupancy rate, (ii) number of indicator 5 (above). diagnostic services performed, and (iii) number of night time caesarians at CEmONCs. Dropped in 2014 Data for 6 months: (i) bed occupancy rate Annual target: (i) 81% (i) 81% (ii) number of (i) maintain rate diagnostic services (ii) maintain at (ii) 11.967 million (ii) 19.140 million performed roughly 12 million (iii) number of night tests annually time caesarians at (iii) maintain at (iii) 4,656 (iii) 10,551 CEmONCs 8,500 annually x April – Sept. 2009 Oct 2013 – March. 2014 Date achieved 9/30/2013 (6 months) (6 months) Comments SURPASSED (ii) by about 150% and (iii) by about 125%, (i) achieved. Also an IOI. CEmONCs should handle more than 50% of complicated deliveries for women Indicator 6: Original. belonging to SC/ST concurrently meeting the standards of quality of care. Modified in Dropped in 2014 2007 to: At least 23% of complicated maternal admissions at certified project CEmONCs (state-wide) are for SC/ST patients. July 2010 - Sept. 2010: 36.5% for 48 Phase 1 Percent of all CEmONCs, and 46.7% complicated deliveries 23% (estimated) Increase by 20% At least 23% for 31 Phase II at CEmONCs that are of baseline CEmONCs. SC/ST patients Oct. 2013 – March 2014: 34.5% for 55 CEmONCs, and 51.9% for EmONCs. Date achieved April 2004 2010 2014 2010 and 2013/2014 SURPASSED. A 20% increase on a baseline of 23% would be 27.6% Actual value at Comments end of project was 125-188% of this target increase, and 150-226% of the revised target threshold of “at least 23%”. Indicator 7: Added in Proportion of C-section deliveries amongst SC/ST mothers at secondary level 2014 CEmONCs % C-section deliveries at secondary level 28% 43% 40% CEmONCs that are SC/ST mothers/babes Date achieved 2007-2008 2014-2015 9/15/2015 (11.5 months) PARTIAL ACHIEVEMENT (93% of the target level). SC/ST mothers accounted for 28% of C-section deliveries at secondary level CEmONCs in 2007/08, 43.3% in 2012- Comments 2013 – reaching the target, but this fell slightly to 40% in 2014/15. Given the worrying increase in C-sections over the period (less in public facilities than private), failing to meet this target is not considered problematic. Indicator 8: AF “promoted” the Effective functioning of CEmONCs (state-wide) as measured by % of complicated Original IOI to OI. admissions and no increase in maternal and neonatal case fatality rates. Dropped in 2014 20% and 50% Risk adjusted maternal 13.33 reductions by 6.17%. mortality rate 2008 and 2010 Maintain 2009 rates. 10% and 20% Risk adjusted neonatal 4.08 reductions by 3.98% case fatality rate 2008 and 2010 Date achieved Nov. 2009 2008, 2010 9/30/2013 Oct 2013-Mar 2014 xi No Appraisal baseline data available so not possible to assess performance at 2010. SURPASSED for period 2010 to 2014. Risk adjusted maternal mortality fatality rate fell to 46% of baseline, risk adjusted neonatal case fatality rate fell to 97.5% of Comments baseline. The rates are adjusted for percentage of all admission that are complicated, which rose from 35% in 2009 to 71.9% of maternal and 47.8% of neonatal admissions in 55 CEmONCs, and to 59% of maternal and 32.7% of neonatal admissions in 50 EmONCs in the reporting period Oct 2013-Nov 2014. Indicator 9: Original Increased patient satisfaction with care (perceived quality of care as measured by (also an IOI) patient satisfaction surveys) (Scores are on a Likert Score 1-5) (i) overall satisfaction (i) 3.96, -79% of patients were score satisfied with the facility (ii) overall in-patient (ii) 3.99, (ii) 3.92 Score (iii) 3.87. (iii) overall out-patient (iii) 3.95 -86% of patients noted score continuous water (iv) Satisfaction with (iv) Maintain or Maintain or available at facilities patient amenities increase increase -98% said out-patients in-patients 3.74 (in-patients), dept and waiting area out-patients 3.72 (out-patients) were clean and hygienic, (v) Satisfaction with (v) -97% in-patients found cleanliness facilities such as labor and in-patients 3.51 (in-patients) ward rooms clean and out-patients 3.7 (out-patients) hygienic 2006 9/15/2015 Date achieved Fergusson Patient 9/30/2013 9/15/2015 (Survey Oct/Nov 2014, Satisfaction Survey 2006 IPSOS) ACHIEVED. The differences in overall satisfaction scores reported are not statistically significant, also, satisfaction is subjective and relative to expectations, which rose over time as facilities were greatly improved by the project. Changes in survey methodologies complicate make trend assessment. Objective measures such as wait time show greatly improved patient experience: in 2014, 84% of patients were satisfied with the wait period of 4 minutes to access outpatient care; 75% of in-patients found the 10 minutes registration time acceptable and 88% of in-patients found 16 Comments minutes for securing a bed acceptable. In so far as they are comparable, these are better than results from the 2011 survey: 80% patients waited no more than 20 minutes to access any services at OPD; 60% of inpatients perceived waiting times at emergency registration and access to doctors in emergency as short, 89% outpatients and 91% in- patients were satisfied with cleanliness of hospital, 81% out-patients and 88% in- patients said running water in taps was available; 96% of in-patients were satisfied with the admission process. Indicator 10: New at Strengthened state-level capacity of pharmaceuticals and medical supplies AF, dropped in 2014 procurement, repair and maintenance of medical equipment xii Well-functioning system A comprehensive is in place. TNMSC now Capacity for: state-wide system handles procurement and procurement of established for all Annual Maintenance pharmaceuticals and Baseline was 2010 status medical Contracts for medical medical supplies, quo. equipment, equipment. 48 Bio- repair and maintenance procurement, medical engineers in post of medical equipment maintenance and to maintain medical repair equipment of public health facilities. Date achieved April 2010 9/30/2013 9/15/2015 ACHIEVED. Established a system to track and improve utilization, repair and maintenance of equipment in health care facilities: electronic inventory of about 100,000 equipment items in hospitals, and team of 48 engineers. TNMSC capacity for managing pharmaceuticals and medical supplies procurement (using World Bank procedures) enhanced. Indicator 11: New in Number of public hospitals accredited by the National Accreditation Board for 2014 Hospitals (NABH) Public hospitals 0 12 12 fully accredited accredited Date achieved 2009-2010 2014-2015 9/15/2015 ACHIEVED – a major achievement; these were the first large public hospitals in India to undergo the rigorous accreditation process. Based on the positive experience and benefits for improving quality of service and health outcomes, GoTN is preparing Comments another 46 hospitals for accreditation of which 1 had entry level accreditation, and 3 had completed assessment and were awaiting results from the Quality Council of India as of 5/23/16, (b) Intermediate Outcome Indicator(s) Original Target Formally Actual Value Achieved Values (from Revised Indicator Baseline Value at Completion or Target approval Target Years documents) Values Reduced case fatality-rate in SC/ST maternal admissions in CEmONC hospitals. Indicator 1: Original, (i) No increase in case-fatality ratio for maternal admissions adjusted for the risk of Modified in 2007 increased proportion of complicated maternal admissions in project CEmONCs.. Not included in AF (ii) No increase in case-fatality ratio for total neonatal admissions adjusted for the risk Dropped in 2014 of increased proportion of complicated maternal admissions in project CEmONCs. (i) maternal (i) 19.55 50% reduction (i) 13.72 (ii) neonatal (ii) 5.24 20% reduction No increase (ii) 4.04 MMR for complicated 114 from baseline 80 maternal admissions Date 2006-2007 9/30/2010 2010 2009-2010 ACHIEVED. MMR in 2010 was 30% lower than baseline, IMR was 23% lower, so partially achieved against PAD target. Results greatly surpassed all revised targets. In 2007 baseline was added, target modified to no increase, denominator changed to Comments complicated maternal admissions (not all admissions) as per Management Letter and AM, 5/18/2007. The “no increase” target was inappropriate given expected (and actual) improvements in quality of care. xiii Indicator 2: Original Increased satisfaction and perceived quality of care in counseling centers (in hospitals) Not in AF or 2014 % of patients satisfied 90% or highly satisfied with -overall services 92% -counsellors’ behavior N/A – centers began 20% improvement % who would access operating in 2007 over mid-term 99% counselling services on their next visit and End line study, ORGCSR, recommend them to The Nielsen Company family and friends Date achieved 2004 9/30/2010 9/30/2015 ACHIEVED – based on high approval levels, since no mid-term level is available as base-line. In addition to counselling centers in hospitals, counselling centers were also Comments set up in tribal areas; overall satisfaction with counselling services in tribal health facilities increased from 85.6% in 2010 to 100% at end line in 2015. Indicator3: New at Health personnel receiving training (number): Doctors (OBGYN and pediatricians), AF. Dropped in 2014 medical officers in 1st referral units and nurses. Doctors trained 1,068 doctors 1,419 doctors NA Nurses trained 1,334 nurses 3,342 nurses Date achieved April 2010 9/30/2013 3/31/2014 SURPASSED. Training completed as per plan, target numbers exceeded by 33% for Comments doctors, 250% nurses. Training focused on skills for operationalizing CEmONCs Indicator 4: New at Number of health facilities constructed, renovated and/or equipped. Maternity wings AF. Dropped in 2014 constructed and equipped at selected medical college hospitals Value (Quantitative or 0 8 8 Qualitative) Date achieved April 2010 9/30/2013 9/30/2014 Comments ACHIEVED Indicator 5: New at Staffing of CEmONCs according to agreed norms (2 OBGYN, 2 pediatricians, 1 AF. Dropped in 2014 anesthetist) in 80 CEmONCs 75 CEmONCs with 4 OBGYN, 2 pediatricians and 2 anesthetists. Staffing at all 80 % of CEmONCs with Phase 1: 77% CEmONCs is at staffing that is at least 50 EmONCs with 2 Phase 2: 38% least 85% of 85% of norm OBGYN, 2 pediatricians agreed norms and 1 anesthetist. Date November 2009 9/30/2013 9/30/2013 SURPASSED All 80 CEmONCs staffed at 100% of norm or better in 2013. However, Comments the final AM notes that staffing has fallen below this level since due to rapidly increased demand for these highly qualified doctors for expanded programs. Indicator 6: New at Increased provision of health services to the tribal population AF. Dropped in 2014 xiv i) mobile outreach services (ii) NGO hospitals providing bed grants, (i) 12 (i) 20 (i) 20 (iii) NGO hospitals providing testing, (ii) 2 (ii) 4 (ii) 4 counseling and treatment services for (iii) 2 (iii) 3 (iii) 3 sickle cell anemia (iv) patient counselors (iv) 32 (iv) 32 (iv) 42 at primary and secondary health facilities in tribal area Date achieved April 2010 9/30/2013 3/31/2014 Comments ACHIEVED for 3 of 4 items and exceeded by 31% for number of counselors (Item iv) Indicator 7: new at Increased emergency transport services (to facilitate use of hospitals by poor and AF. Dropped in 2014 disadvantaged) Vehicles providing 730 (August 2015). emergency transport 385 585 885,452, of which Also monitor: 26,915 from tribal areas # people transported 113,570 No targets set 26% pregnant women. % pregnant women 20% road accident % of road accidents victims Date 2008-09 9/30/2013 April 2014-March 2015 Comments SURPASSED at 125% of target Indicator 8: new at Number of district hospitals with support services provided (laundry, cleaning, security AF. Dropped in 2014 & food distribution) Number of district 0 at least 20 48 hospitals with services Date April 2010 9/30/2013 3/31/2014 SURPASSED Actual value is 240% of target threshold. Resulted in improved services Comments – improved cleanliness etc, attested by patient satisfaction surveys and evaluations. Indicator 9: new in Total number of complicated SC/ST maternal admissions at certified CEmONCs (with 2014. at least 2 OBGYNs, 2 pediatricians and 1 anesthetist) Number 35,156 66,000 74,373 Date 2007-2008 2014-2015 Oct 2014 – Sept 15, 2015 Comments SURPASSED by 113% of target. Indicator 10: new at Number of tribal patients provided outpatient care through Mobile Outreach Vans AF, Dropped in 2014 Number 137,543 200,000 244,003 Date 2007-2008 2014-2015 Oct 2014 – Sept 15, 2015 Comments SURPASSED by 122% of target Indicator 11: new at Percentage of calls made by pregnant women, attended to by Emergency 108 services AF Dropped in 2014 Percent response 90.3% 96.8% 99.3% Date 2009-2010 2014-2015 Oct 2014 – Sept 15, 2015 Comments SURPASSED by 103% of target xv Indicator 12: new at Percentage of calls made for Road Traffic Accident victims, attended by Emergency AF. Dropped in 2014 108 services Percent response 66% 79.4% 87% Date 2009-2010 2014-2015 Oct 2014 – Sept 15, 2015 Comments SURPASSED by 110% of target Indicator 13: new at Number of performance based contracts delivering health care services in the project in AF. Dropped in 2014 Tamil Nadu Number 0 9 37 Date 2009-2010 2014-2015 9/15/2015 SURPASSED by 411% of target. (5 contracts for Regional Diagnostic Labs, 2 Comments Housekeeping services contracts and all 30 contracts for handling hospital waste) Indicator 14: Decrease in smoking rates, particularly among the poor and young, in pilot Original, changed in districts. Changed in 2007 to: Increase in awareness amongst 13-15 year olds of 2007. Dropped at AF. the risk of tobacco use in two pilot districts No data for original Original: % currently indicator smoke at least one Awareness in pilot 79-80% (Sivagangai) cigarette/day districts before project Increase in 5% decrease and interventions: 63-67% awareness Revised: Aware that 95-99 (Virudhunagar). (Sivagangai) and 71-94% “smoking is injurious (Virudhunagar) (page 29, to health” baseline survey) Date 2008 9/30/2010 9/30/2010 2010 Comments ACHIEVED. Weak indicator, awareness of harm does not correlate with use. Indicator 15: Decreased road traffic accident case fatality rate. Original. Activity Changed in 2007 to: Dropped at MTR/AF Increased use of helmets as measured by direct observation surveys. Dropped Date Nov/Dec 2007 (MTR) NOT ACHIEVED. Activity excluded from the project at MTR. The project did monthly helmet use surveys in 14 locations in 13 districts (April 2007-April 2009) Comments which showed mixed results across districts. An initial increase in use was followed by falls after GoTN reduced enforcement in response to strong public resistance. Indicator 16: new at Health promotion for prevention of CVD among school children carried out AF. Dropped in 2014 Number of schools where promotion 50 5,000 16,369 16,369 activities are done Date April 2010 9/30/2013 9/15/2015 9/15/2015 ACHIEVED. Original target far surpassed, target revised at restructuring of 2014. Comments Activities done in collaboration with education department. Indicator 17: (i) Clinic based NCD control pilots implemented according to plan. (ii) Proportion of Original. Dropped in hypertensives receiving effective treatment. (iii) Increased number of women being 2014 screened for cervical cancer. xvi (i) Clinic based pilots (i) 2 pilots completed, implemented per plan evaluated, informed (ii) Diagnosed cases of design of full program. 0 2 pilots completed hypertension receiving (ii) 1.231 million data not available 10% increase (tbc) treatment per screened for hypertension data not available 10% increase (tbc) protocols. (iii) 488,034 women (iii) Women screened (94% of targeted age for cervical cancer group) screened Date 2004 9/30/2010 January 2010 ACHIEVED. Original placeholder targets of 10% increase over baseline in the PAD RF were not revisited, baseline data were not collected as intended in a household survey in the pilot areas. Screening and treatment rates were very low before the pilot, so 10% increases are likely to have been exceeded by the large pilots. In 2007, (ii) and (iii) were changed to: “Increasing percentage of hypertensive patients (measured quarterly) correctly receiving treatment at facilities enrolled in the pilot”, and “Increased coverage of women between 30-60 years of age screened for cervical Comments cancer in pilot districts.” These targets were achieved, and the pilots completed, carefully evaluated and informed the design of the programs rolled out under the AF Results: (ii) Of 1.231 million people screened for hypertension, 77,757 new cases were diagnosed, about 68% followed up and 23% of those diagnosed were treated. (iii) 488,034 women (94% of the 518,000 women in the target age group) screened by Jan. 2010, only about 50% received further screening diagnostics, and the treatment rate was only 23%. Detailed evaluation (by NIE) informed the roll-out, with improved design to reduce loss-to-follow up and achieve higher treatment rates. Indicator 18: new at AF: Cancer cervix screening and cardio-vascular disease (CVD) prevention and AF, wording changed control as measured by number screened and treated (equipment, training provided) in 2014 to align with 2014: (i) Percentage of eligible women in age group 30-60 years screened for cancer of indicator definition. cervix and (ii) Percentage of eligible persons (both men and women) in age group >30 Dropped in 2014 years screened for hypertension (i) Percentage of (i) 488,084 (85% of (i) at least 50% in (i) 40% of (i) 71% women 30-60 women age 30-60 women aged 30-60) districts where eligible screened for cervical years screened for screened for cervical program scaled-up women. cancer (10.3 million cancer of cervix cancer in 2 pilot districts. during July 2012- (ii) Percentage of (ii) at least 50% of September 2015). eligible persons (men (ii) 1,231,259 (3.4% of people aged 30+ (ii) no change and women) age >30 people aged 30+) years where (ii) 77% of persons 30+ years screened for screened for hypertension program is screened for hypertension hypertension in 2 pilot districts. implemented (29.03 million persons). Date April 2010 9/30/13 9/15/2015 9/15/2015 SURPASSED. Cancer screening was 142% above the 50% target threshold and 178% above revised target (revision was not warranted). Hypertension screening was 154% above target threshold. The 2011 census was used for the denominator. The NCD programs were scaled up to all 32 districts, in 1,710 facilities (PHCs, secondary and tertiary). People screened as positive were referred for confirmation of diagnosis, and Comments then for treatment and/or life style counseling. TNHSP added screening and treatment for diabetes and breast cancer. During July 2012-Sept 2015, 23 million people were screened for diabetes, the 0.958 million detected positive were given treatment and lifestyle counseling; 12.5 million women were screened for breast cancer, 153,330 women were referred for further diagnosis and treatment. xvii Indicator 19: new at AF (Core Indicator). Number of Health facilities constructed, renovated and/or equipped Dropped in 2014 2,228 facilities 2,176 facilities (1889 PHCs, (1,859 PHCs, 274 2,330 (1,889 PHCs, 274 274 GH, and Number of facilities 0 GHs and 43 GH, 100 medical 65 Medical Medical College dispensaries and 67 college Hospitals) medical college hospitals) hospitals) Date 2009-2010 9/30/2013 9/15/2015 9/15/2015 SURPASSED at 107% of original target and 105% of target as revised at restructuring Comments in 2014 Indicator 20:New at AF (Core Indicator) Number of health personnel receiving training Dropped in 2014 Number of persons 0 105,000 398,285 trained Date April 2010 9/15/2015 6/30/2015 Comments SURPASSED at 379% Indicator 21: Operational HMIS being used for management decision making at project facilities Original 100% Computerized reporting Paper based HMIS computerized HMIS in all facilities in all Phase 1 facilities. operational. reporting at Roll-out to others begun project facilities Date 2004 9/30/2010 9/30/2010 Comments ACHIEVED Indicator 22: New at Operational HMS being used for decision making in 270 hospitals, and HMIS AF operational across the state - HMS operational in 264 secondary and 45 tertiary - HMS operational hospitals in 270 hospitals - HMIS operational Facilities reporting - HMS in 38 hospitals - Operational across TN in 2,300 health through HMIS HMIS across the facilities: 1,889 PHCs, state 274 GH, 70 municipal dispensaries, 67 medical college hospitals. Date 2010 9/30/2013 9/15/2015 SURPASSED at 114% of target number of hospitals, and also rolled out CMS (college management system) in 20 government medical colleges and Dr. MGR medical Comments university, and UAS (University Automation System) in Dr. MGR medical university. HMS being used for decision making, especially to improve service quality. Indicator 23: New in Number of health facilities where HMIS is used to submit monthly reports. 2014 xviii 2,300 2,228 1,889 PHCs, 274 GH, 70 Number of facilities 0 (GH -274, PHC – municipal dispensaries, 1889, MC – 65) 67 medical college hospitals. Date 2009-2010 9/15/2015 9/15/2015 Comments SURPASSED at 103% of target Indicator 24: Original, modified in (i) Short and medium term measures for health care waste management (HCWM) 2007. Part (ii) added implemented at project facilities. (ii) Retraining of staff of health care facilities. (iii) at AF, dropped in PPP with NGO/private partners for transportation and final waste disposal. 2014 (i) Number of (i) Plan HCWM plan (i) HCWM plan facilities where implemented implemented in 449 implemented in all in 270 health public facilities. HCWM plan is project and non- 0 facilities (AF) implemented project public 49,500+ personnel from facilities (ii) 44,000 over 449 facilities trained/ staff in 449 retrained in 9 regional (ii) Number of 0 facilities training centers in all facility staff trained/re- aspects of managing retrained. trained (AF) health care waste. (iii) PPPs operating 0 (iii) Evaluation of (iii) HCWM PPPs piloted and and assessed. plan assessed, then PPPs set up PPP pilots implemented with 30 Common in 449 health Treatment Facilities facilities – (CTFs) to collect, Target for disinfect and dispose of 9/15/15 (set in waste. End-line 2014) evaluation done. Date 2004 9/30/2010 9/30/2013 8/5/2015 SURPASSED. The HCWM plan, which was broadened to an Infection Control and Waste Management (ICWM) plan was implemented more broadly than originally planned, and is one of the standout successes of the project. Original indicator (i) was modified by replacing “short and medium term measures” by “HCWM plan” in 2007. Comments At Additional Financing, “Retraining of staff of health care facilities” was added, and the target number of facilities specified. Implementation included an intensive behavior change campaign. Effective coordination was established with the State Pollution Control Board and municipal bodies, who do regular quality assurance checks at all treatment facilities. Indicator 25: Rigorous evaluation of 12 PPPs completed in terms of measured gains in Original access, quality and cost-effectiveness. Independent Evaluations completed of Evaluation 0 evaluation of PPPs all PPPs which informed completed scale-up decisions Date 2004 9/30/2010 9/30/2010 xix ACHIEVED. Evaluation studies of the PPPs included: 2 pilots for waste treatment, 2 PPPs for bed grant schemes (tribal areas), 2 PPPs for sickle cell anemia (tribal areas), several PPPs for mobile van outreach programs (tribal areas), PPP for emergency transport (ambulance), several PPPs for the provision of patient counsellors at Comments CEmONCs and non-CEmONCs facilities. Following evaluation studies, PPPs were scaled up for implementation during Additional Financing. Nearly all PPPs (including for housekeeping services and laboratories) have been absorbed by GoTN into its own regular sector health sector budget. Indicator 26: New at Maintain hospital based Quality Circles of Excellence, as measured by submission of AF. Dropped in 2014 monthly reports 267 secondary care 270 hospitals hospitals have Quality Number of hospitals 80 CEmONCs reporting reporting on Circles of Excellence (3 reporting on 20 on agreed quality of care agreed quality of converted into medical indicators monthly indicators care indicators college hospitals), and are reporting online monthly Date April 2010 9/30/2013 3/31/2014 Comments ACHIEVED Indicator 27: New at Number of health personnel of secondary hospitals trained to improve quality of care, AF (Core Indicator). including hospital management, rational use of drugs and skills based training Dropped in 2014 Additional 80 CMOs trained in hospital management and 739 1,000 staff trained staff trained in hospital in hospital administration; 1,692 in Number of people 355 staff trained in management and Quality Indicators, 1,915 trained hospital management 900 staff trained in in rational use of drugs; rational use of 37,468 doctors and nurses drugs trained in CEmONC skills, medical equipment use and NCDs. Date April 2010 9/30/2013 2014 SURPASSED at 213% of target for rational drug use training, and at 117% for Comments hospital management training (total number includes 355 trained before 2010) Indicator 28: New at AF (Core Indicator) Hospitals accredited including provision of civil works, equipment and training Dropped in 2014 Capacity for strengthening 12 hospitals upgraded and hospitals for the successfully accredited. process of In the process, Hospitals accredited 12 ongoing accreditations accreditation Department of Health’s established in learned how to manage Department of the accreditation process. Health Date April 2010 9/30/2013 9/15/2015 Comments ACHIEVED xx Indicator 29: New at Fully operational Project Management Unit (PMU) integrated into the Department of AF. Dropped in 2014 Health PMU increasingly PMU integrated integrated in Department into the PMU operational of Health. PMU’s work Department of absorbed by Department Health of Health by project-end. Date April 2010 9/30/2013 9/15/2015 Comments ACHIEVED Indicator 30: New at Mechanism established for planning and implementing IEC activities in the health AF. Dropped in 2014 sector, and for monitoring of PPP activities Support NRHM All IEC materials for establishment developed during the of a coordination project and results of unit in Directorate monitoring of PPP Mechanisms None of Health for all activities are stored on the established health TNHSP website available communication for use by National activities in the Health Mission and state Directorate of Health. Date April 2010 9/30/2013 9/30/2014 PARTIALLY ACHIEVED. The target statement omitted PPP coordination, which is explicitly part of the Indicator, so is included in the ICR assessment for this indicator. Mechanisms for monitoring PPPs have been established and institutionalized; the Comments move to performance-based contracts strengthens monitoring. Project PPP activities have been handed over to NRHM and Government of Tamil Nadu for management and financing. The PMU included an IEC coordinating unit, but a unit was not established in Health Directorate for state-wide coordination. Indicator 31: New at Establish SHDRC (State Health Data Resource Center) in Tamil Nadu AF. Dropped in 2014 SHDRC SHDRC established and None established operational. Date April 2010 9/30/2013 8/5/2015 ACHIEVED, The SHDRC collates, mines and runs higher order analytics on data Comments from over 20 Directorates of the Health Dept., to provide easy to use dashboards for administrators and managers, and help drive continued improvement. Indicator 32:Original Upgradation and repairs of project facilities completed according to plan. Dropped in 2014 Decision to take up Phase III districts based on All civil works completed Facility upgrading performance of 0 as planned, in several completed per plan Phase II districts phases. in implementation of software activities Date 2004 9/30/2010 9/30/2010 xxi ACHIEVED. All planned civil works to provide need-based additional infrastructure and enhance its quality completed and handed over – 35 Phase I works and 190 hospitals in Phase II. Planned “soft” activities including staff training undertaken and a Comments system to better utilize, repair and maintain equipment is in place. At Additional Financing, a decision was taken to keep civil works to bare minimum – upgrade 8 CEmONCs and complete upgrades at 12 facilities needed for NABH accreditation. (i) Increased number of laboratory tests, x-rays and other diagnostics at project Indicator 33:Original facilities. (ii) Reduction in equipment downtime. Added at AF: (iii) Equipment Changed at AF. provided to selected district hospitals in order to ensure the provision of a full range of Dropped in 2014. services, as per agreed norms (i) 20% increase (i) covered in (i) Data reported in PDO (i) Number diagnostics from baseline OI indicator 5 above Not available (ii) Dropped at (ii) no data (ii) equipment (ii)Declining trend AF (iii) Included in facility downtime in downtime (iii) no target upgrading indicator 32 Date 2004 9/30/2010 4/29/2010 9/15/2015 ACHIEVED. Diagnostics data captured in PDO indicator 5 above. System for equipment maintenance and repair set up and working, as reported above. Regular Comments reporting by hospitals and availability of engineers has greatly improved notification of problems, repair and maintenance of equipment. Indicator 34: Original. Item (ii) (i) Availability of staff according to norms at project facilities, [(ii) reduction in doctor dropped in 2007. (i) absenteeism as recorded in supervisor’s logbook] dropped in 2014 Support (i) Manpower Of the 267 project Number of project additional according to facilities, facilities with facilities having staff in contractual norms in all staff per sanctioned posts position against staff at 270 (project) hospitals were: (i) Doctors - 241; sanctioned posts: Doctors project (ii) Nurses – 259; (iii) - 116; Nurses – 150; hospitals in (ii) Absenteeism Technicians – 258; and Technicians – 40; and for accordance reduced by 50% (iv) all three categories: all three categories: 22. with agreed 211. norms Date March 2005 9/30/2010 4/29/2010 9/30/2010 PARTIALLY ACHIEVED 79% of the 267 project hospitals had staff as per norms by 2010. For each cadre, achievement was much closer to the very ambitious target: 90% for doctors, 97% for nurses and for technicians, the reason the ICR team Comments considers this partially achieved (rather than not met). Manpower norms were revised and rationalized depending on hospital size, in the early years of the project. Part (ii) on doctor absenteeism dropped in 2007. Indicator 35: Added Evaluation of pilots being implemented to enhance management of project in 2007. Dropped at AF facilities facing difficulties. Completed and system NA Evaluation implemented across all public hospitals. Date 2004 9/30/2010 9/30/2010 xxii ACHIEVED. The system to grade hospitals every month into A, B, C and D categories on the basis of 20 performance indicators, with poor grades triggering action plans, was evaluated and found useful for helping improve management and operations Comments in poorly performing project hospitals. PMU identified 65 poorly performing hospitals agreed on action plans to improve performance, with continued monthly monitoring. System was implemented across all hospitals. Indicator 36: new at TNMSC strengthened per agreed norms AF. Dropped in 2014 TNMSC strengthened per Biomedical engineers agreed plan, hired by the project. None including TNMSC strengthening mainstreaming of was undertaken with state biomedical government funds. engineers Date April 2010 9/30/2013 3/31/2014 ACHIEVED. See PDO Indicator 10. Biomedical engineers hired under the project Comments were absorbed as regular GoTN staff. This complemented TNMSC strengthening undertaken with state government funds. G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 05/07/2005 Satisfactory Satisfactory 7.50 2 11/04/2005 Satisfactory Moderately Satisfactory 7.52 3 05/10/2006 Moderately Satisfactory Moderately Unsatisfactory 7.52 4 11/07/2006 Moderately Satisfactory Moderately Unsatisfactory 7.71 5 05/03/2007 Moderately Satisfactory Moderately Satisfactory 11.89 6 10/18/2007 Satisfactory Moderately Satisfactory 18.98 7 04/09/2008 Satisfactory Satisfactory 22.80 8 10/09/2008 Satisfactory Satisfactory 29.54 9 03/23/2009 Satisfactory Moderately Satisfactory 42.93 10 09/18/2009 Satisfactory Moderately Satisfactory 63.64 11 04/21/2010 Satisfactory Moderately Satisfactory 88.24 12 05/14/2011 Satisfactory Satisfactory 113.87 13 06/06/2011 Satisfactory Satisfactory 113.87 14 02/09/2012 Satisfactory Satisfactory 135.03 15 09/12/2012 Satisfactory Satisfactory 145.13 16 04/27/2013 Satisfactory Satisfactory 166.22 17 08/17/2013 Satisfactory Satisfactory 170.88 18 01/15/2014 Satisfactory Satisfactory 178.05 19 07/30/2014 Satisfactory Satisfactory 186.77 20 10/06/2014 Satisfactory Satisfactory 186.77 21 12/15/2014 Satisfactory Satisfactory 195.33 22 06/16/2015 Satisfactory Satisfactory 199.45 23 09/02/2015 Satisfactory Satisfactory 207.67 xxiii H. Restructuring (if any) Board ISR Ratings at Amount Restructuring Approved Restructuring Disbursed at Reason for Restructuring & Key Date(s) PDO Restructuring in Changes Made DO IP Change USD millions Minor changes to Results Framework approved by CD in response to 05/18/2007 S S 11.89 recommendations of a Bank-wide review. Not processed as a restructuring. Reallocation of proceeds among 02/19/2010 S MS 87.08 categories. Approval of Additional Financing, changes in PDO, indicators and targets, 04/29/2010 Yes S MS 88.59 extension of Closing Date by 3 years to 9/30/2013. Reallocation of proceeds among categories, to finance taxes and modify 06/28/2010 S MS 92.02 definition of incremental operating costs. CD approved extension of Closing Date 05/08/2013 S S 166.22 by one year to 9/30/2014. CD approved extension of Closing Date by 11.5 months to 9/15/2015. Results 08/07/2014 S S 186.77 Framework was rationalized to be more relevant and reduce number of indicators. I. Disbursement Profile xxiv 1. Project Context, Development Objectives and Design 1.1. Context at Appraisal Tamil Nadu was the 7th most populous Indian state (65 million in 2001), one of the five most urbanized, and had the second lowest population growth rate (1.43%). Annual GDP growth was averaging more than 6%. Good “social determinants” of health, and high coverage and utilization of health services (e.g., 93% of children fully immunized, 89% of births in health facilities) had contributed to steady improvement in infant- and under 5 mortality and other health outcomes. The state’s human development and health indicators were among India’s best. Health sector trends and challenges. Despite this significant progress in health and access to services, Tamil Nadu’s infant mortality rate (IMR) of 52 per 1000 live births in 1999 was still much higher than in Sri Lanka and Kerala (IMRs of 12 and 16 respectively), and maternal mortality (MMR) had stagnated at 110 (2003).1 Audits of maternal deaths indicated that facilities -- especially in disadvantages areas -- could not all provide comprehensive emergency obstetric and neonatal care (CEmONC), and more needed to be done to improve and ensure quality of care in all facilities in a systematic and well-organized way. Tamil Nadu’s burden of disease from non-communicable diseases (NCDs) was large and growing. Heart disease, diabetes, and cancers were already the leading cause of death, and traffic deaths among the worst in India. NCDs and smoking (a key NCD risk factor) were especially high among the poor, who were therefore most burdened by the economic effects of illness, health care costs, lost productivity and premature death. Although use of health services was far more equitable than in most states, hospitalization among the poor was only 37% of the rate among the wealthy, and scheduled castes and scheduled tribes (SC/ST) had very limited access to health services. Health outcomes were relatively poor in districts and blocks within districts with pockets of SC/ST populations. Total health spending in Tamil Nadu was low, and predominantly out-of-pocket. Public health expenditure was less than US$3 per capita per year, and had fallen from 7.5% of the state budget in the mid-1980s to 5.8% in 2001. Nearly 75% of the health budget went on salaries, leaving very little for consumables, equipment, infrastructure and maintenance. Most central government funding for health was for primary care, leaving secondary care chronically under-funded. State Health Policy and Capacity Gaps. The Government of Tamil Nadu (GOTN) Health Policy of 2003 laid out ambitious goals, including reducing IMR to 15 per 1000 and MMR to 50 per 100,000 live births by 2020, and doing more to address key non-communicable diseases and injuries while sustaining vigorous efforts to control communicable diseases including HIVAIDS. The strategy focus was on improving the health status of the general population, with special emphasis on low-income communities and families. However, the state lacked experience in strategic and financial planning for the health sector, and in quality improvement activities such 1 Tamil Nadu Health Indicators at a Glance, 2014-2015, Directorate of Family Welfare, Government of Tamil Nadu (GoTN)) 1 as setting technical standards, ensuring quick adoption of technological advances, oversight of the private sector, accreditation of health facilities, efficient management of public sector hospitals, and monitoring health policy impact on vulnerable populations. Rationale for Bank involvement. GoTN requested funding and technical advice to help implement its new Health Policy and improve the quality of care across the state’s health system. In addition to substantial funding, IDA provided experience in health system strengthening, and the ability to help GoTN build capacity to develop, evaluate and implement quality assurance mechanisms, test innovative interventions to reduce NCDs, rigorously evaluate innovations to decide which to scale-up, and to collaborate better with the private sector to help achieve state health policy goals. Although Tamil Nadu had not had a state-level health project before, as a progressive state with relatively high capacity in the health sector, it was considered a good place to put into practice the shift in state level health system projects recommended by a 2002 major review of all State Health Systems projects in India.2 Additional Financing was approved in 2010 to continue successful project activities, expand the scope of some, and roll-out the successful pilot of NCD activities state-wide. The additional rationale for the AF was that the innovations in the project would be a valuable model for other states in India and other countries. The project was fully aligned with the core goals and strategic principles laid out in the World Bank India Country Strategy for FY05-08 (CS), approved in 2004: to help improve the quality of life especially for India’s poorest citizens and help India move closer to achieving the MDGs; selectively expand lending in health (and other specific areas); and focus on outcomes to help India achieve its development goals. Tamil Nadu was one of the 12 states where over 90 percent of India’s poor lived, and the project explicitly aimed to expand health service access and utilization by poor, disadvantaged and tribal groups. The CS noted that the project followed the guidelines for Bank engagement in the health sector and would “break new ground in forging public-private partnerships …strengthening oversight of private providers, increasing public expenditure on health and reorienting health facilities to ensure service for the poor…”3 The project also reflected “some of the most important elements of the fast reform scenario - most of which are embodied in the Tenth Plan”: “Refocus health, education and social safety net programs on outcomes. Improve the private market for health care through training, public information and accreditation.” (CS p9). Rigorous assessment of the cost-effectiveness of private provision of publicly financed services would generate information for the broader debate on how to increase the effectiveness and efficiency of publicly financed services across India. This would contribute to the third strategic principle of the CS: to expand the role of the Bank Group as a politically realistic knowledge provider and generator. The project would also help implement the Government’s core strategy for poverty reduction, embodied in India’s Tenth Five-Year Plan for 2002/3 to 2006/7. The first two items in the Plan’s core strategy were: (1) Improve governance and service delivery, placing greater reliance 2 India State Health Systems, Quality Enhancement Review, July 2002. The suggestions included putting more emphasis on sectoral reforms, beginning to integrate disease programs into mainstream service delivery, giving more attention to financing issues, and more engagement with the private sector. 3 CS p26 and Annex 5, which lays out the guidelines for Bank engagement in the health sector. 2 on the private sector and on public sector reforms to deliver accountability, reduce opportunities for corruption and improve the speed and effectiveness of government at all levels. (2) Second, reduce poverty, including by better access to health care. The project was designed to contribute directly to higher-level MDG objectives to which India was fully committed: reducing maternal, child and infant mortality and premature and preventable mortality among adults. 1.2. Original Project Development Objective (PDO) and Key Indicators (as approved) Project Appraisal Document (p 5): The Project Development Objective was “to significantly improve the effectiveness of the health system, both public and private, in Tamil Nadu through: (i) increased access to and utilization of health services, particularly by poor, disadvantaged and tribal groups; (ii) development and pilot testing of effective interventions to address key health challenges specifically non-communicable diseases; (iii) improved health outcomes, access and quality of service delivery through strengthened oversight of the public sector health systems and greater engagement of non-governmental sector; and (iv) increased effectiveness of public sector hospital services, primarily at district and sub-district levels.” The PDO statement in the Development Credit Agreement (DCA, p. 17) is worded slightly differently: “public and private” is omitted from the main clause of the objective statement; item (ii) omits the reference to piloting interventions and (iv) omits “district and sub-district levels”. The differences make the PDO statement less precise. The four key outcome indicators (PAD, p. 31-32) were: (a) Total in-patient utilization (considering both the public and private sector) by the poorest 40% of the population increased (as measured by an asset mix), (b) Completion of two rigorously evaluated pilots of clinic-based NCD prevention and control, careful monitoring of the effectiveness of other NCD prevention activities, and assessment of the impact of these pilots on the development of state-wide policy. (c) Improved quality of care (QOC) in public hospitals as measured by a series of indicators (including management of indicator conditions, patient outcomes, and quality control mechanisms) and implementation of regulation accreditation system to improve quality of care in private sector hospitals. (d) An increase in the number of complicated deliveries by women in the Scheduled Caste/ Scheduled Tribe (SC/ST) population that are handled by CEmONCs that meet standards for quality and neo-natal care. The PAD Results Framework in Annex 3 listed a slightly different version: “CEmONCs should handle more than 50% of the complicated deliveries for women belonging to the SC/ST concurrently meeting the standards of quality of care” (p.29); a few pages later the table detailing the measurement strategy for the indicators gave a target of 20% improvement on the baseline which was still to be determined (p.32). 1.3. Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification Additional Financing (AF) was approved on April 29, 2010 to enable successful NCD pilots to be scaled up across the state; and to support the continuation of successful activities. The PDO and key indicators were changed to reflect the focus of the activities that would continue to be funded and the change from piloting NCD interventions to implementing them across the state. 3 Reference to the private sector part of the health system was dropped from the PDO, an appropriate change since the project focus was mainly on the public sector, and because GoTN has little direct influence or oversight of private sector health providers. The revised PDO in the AF Project Paper was: “To significantly improve the effectiveness of the health system in Tamil Nadu as measured by: (i) increased access to and utilization of maternal and neo-natal care services, particularly by poor, disadvantaged and tribal groups; (ii) effective non-communicable disease interventions scaled up throughout the state; (iii) improved health outcomes, access and quality of service delivery through strengthened oversight of the public sector health systems and greater engagement of non-governmental sector; and (iv) increased effectiveness of public sector hospital services, primarily at district and sub-district levels.” The project objective in the AF Financing Agreement (page 5) was the same (except for trivial differences in grammar). The AF extended the project Closing Date and the end date for expected outcomes by three years. The Results Framework was revised to reflect the new PDO focus on consolidating and continuing the achievements made and to roll out NCD programs state-wide, with expected outcomes defined as follows: (i.) At least 23% of complicated maternal admissions at certified project CEmONCs (state- wide) will be for SC/ST. patients. (ii.) Effective functioning of CEmONCs (state-wide) as measured by % of complicated admissions and no increase in maternal and neonatal case fatality rates. (iii.) Scale-up of cancer cervix screening and cardio vascular disease prevention and control based on a comprehensive assessments of the pilots. (iv.) Improved access to health care as measured by in-patient utilization of services by the poorest 40% of the population. (v.) Improved quality of care as measured by (i) bed occupancy rate, (ii) number of diagnostic services performed, and (iii) number of night time caesarians at CEmONCs. (vi.) Patient satisfaction (perceived quality of care) as measured by patient satisfaction surveys. (vii.) Strengthened state-level capacity of pharmaceuticals and medical supplies procurement, repair and maintenance of medical equipment. This table explains continuities and changes in the outcome indicators: Original AF Comment on continuity and changes Indicator Indicator (a) (iv) Same indicator, worded slightly differently (b) (iii) Indicator for NCD pilots replaced by indicator for scaling up programs across the state (c) (v) The revised indicator on improved quality of hospital care dropped reference to private hospitals, and specified 3 things to be measured (d) (i), (ii) This revised indicator for ST/SC access to quality care for complicated deliveries includes neonatal and maternal mortality outcome measures, and includes baseline data for access to care. (vi) New indicator (vii) New indicator to explicitly measure aspects of quality 4 1.4. Main Beneficiaries The primary target groups expected to benefit most from the project were “poor, disadvantaged, and tribal groups” – people in scheduled castes and scheduled tribes, or living in tribal, hill, remote and underserved areas, whose access to health care services would be increased. All patients using Tamil Nadu’s secondary hospitals would benefit from improved services as a result of hospital refurbishment and upgrading, equipment repairs and maintenance. The whole population of Tamil Nadu would benefit from improved quality of health services, and women and infants and their families would benefit from the expected improvement in maternal and neonatal mortality. The AF would additionally benefit adults in the state through interventions to reduce NCD risks, and screen and provide treatment for cervical and breast cancer and cardio- vascular diseases. Secondary beneficiaries: Health care providers and other staff working in health facilities were also expected to benefit. Strengthening the Health Management System (HMIS) would benefit the Department of Health and Family Welfare (DoHFW), hospital administrators, medical professionals and other staff of hospitals and other health care facilities by digitizing health records and reporting, enabling more efficient referral and management of patients, and providing health facility information that could be used for better budgeting, planning and accountability. DoHFW staff would benefit from training activities. Improved health care waste management at health care facilities would enhance safety for patients and service providers. Public-private partnerships (PPPs) with non-governmental organizations (NGOs) to provide health services in tribal and remote areas would benefit patients, and develop the capacity of NGO partners. Private sector service providers would benefit from government contracts for house-keeping services and health care waste collection, transportation and disposal. The provision of ambulances would benefit women in labor, victims of road traffic accidents and others requiring emergency transport to a hospital. Strengthening the state health sector would enable better implementation of ongoing and planned centrally sponsored health programs in the state, including the Reproductive and Child Health Project, National AIDS Control Project, and Revised National Tuberculosis Control Project that were funded by the WB. Finally, learning from NCD pilots and innovations in HMIS would inform state policy and could be replicated in other states of India. 1.5. Original Components (as approved) Component 1: Increasing Access to and Utilization of Services (US$43.79 million). This component would:  Reduce maternal and neonatal mortality by establishing at least 2 CEmONCs in each district, first in disadvantaged districts, equipped with a trained complement of clinical and paramedical staff and the equipment, supplies and drugs needed to provide treatment for all types of obstetric and neonatal emergencies. Contracts would be signed with NGOs to provide emergency transport services and facilitate referral.  Improve tribal health by strengthening existing primary and secondary health services in tribal areas through PPPs with NGOs and contracting NGOs to provide mobile clinical 5 services in 12 identified districts; giving grants to NGO hospitals to provide in-patient services; and training a cadre of village level tribal health volunteers.4  Facilitate use of hospitals by poor and disadvantaged people and stimulate demand for services through: (i) community mobilization by NGOs and outreach workers, (ii) behavior change strategies to promote health, (iii) counseling centers run by NGOs and local self-help groups to guide patients seeking hospital services, and (iv) training health personnel in inter-personal communication to improve provider behavior. Component 2: Developing Effective Models to Combat Non-Communicable Diseases and Accidents (US$5.65 million). The component aimed to develop effective ways to reduce NCDs and road traffic accidents, undertake pilots and evaluate their impact so as to inform state policy and future NCD programs. The activities supported under this component were:  Health promotion activities for preventing NCDs by reducing exposure to risk factors, such as behavior change communication (BCC), interventions in communities, schools and workplaces, and setting up life-style counseling centers.  Two NCD pilots, each in two districts: Pilot 1 screened for hypertension and provided medications and advice on modifying risk factors such as diet, sedentary lifestyle, and smoking in 2 districts. Pilot 2 assessed the costs and benefits of cervical cancer screening and treatment. In each case, one of the pilot districts was relatively more industrialized, so that the impact of urbanization on NCD risk factors could be analyzed. Rigorous evaluation of the pilots provided information for decisions on whether and how to scale- up across the state.  Traffic injury prevention and treatment interventions in coordination with relevant Departments (e.g. Transport and Police). Component 3: Building Capacity for Oversight and Management of the Health System (US$25.61 million). The activities were designed to achieve four things:  Improve monitoring and evaluation by strengthening the health management information system (HMIS) to report regularly on quality of care indicators, utilization rates at health care facilities and hospital activity indicators. Establish a computerized system at all levels to track patient, service and management information, network all hospitals to track referred cases and monitor outcome of programs. Provide feedback to service providers and program managers for follow up and continuity of care.  Improve Quality of Care by mainstreaming continuous quality improvement practices, developing and implementing quality indicators, establishing Quality Improvement Circles in health facilities, developing protocols for improved management of key health conditions, and helping GoTN implement a stronger system for overseeing health facilities.  Strengthen health care waste management through implementing guidelines on proper segregation, color-coding, transportation, and disposal of hospital solid wastes; set up 4 Public-private partnerships (PPP) were a relatively novel approach, and included u se of public funds to “purchase” basic services for the poor from NGOs and private health providers, collaboration with the private sector for a range of professional services by “contracting in” to government health facilities, and encouraging NGOs in remote tribal districts to operate government facilities to ensure outreach of health services to disadvantaged populations. 6 PPPs with NGOs for waste transportation and treatment; develop training manuals, train trainers, and monitor progress.  Build capacity for developing and implementing health strategies, by establishing a Strategic Planning Unit to act as a think tank and conduct studies on important health systems issues; setting up a PPP “wing” in the GoTN to manage and monitor PPP contracts; conducting a health insurance pilot, and strengthening Project Management’s capacity for monitoring and undertaking procurement. Component 4: Improving the Effectiveness and Efficiency of the Public Sector to Deliver Essential Services (US$50.90 million). The main activities were to:  Refurbish and upgrade secondary care facilities including assuring basic amenities such as water and electricity.  Repair equipment and implement a good maintenance system through Tamil Nadu Medical Services Corporation (TNSMC), equipment suppliers and hospital officials.  Establish and implement staffing norms and train government staff in human resource planning and development; conduct activities to improve staff morale and courtesy to patients; introduce accreditation of health facilities and performance appraisal to help improve workforce efficiency.  Enhance management of public facilities by setting up twinning arrangements between hospitals, giving recognition to high performing hospital administrators, and testing new ways to enable hospitals to improve their performance. 1.6. Revised Components AF of US$117 million was approved on April 29, 2010 to enable the GoTN to consolidate and continue successful project activities, and scale up selected NCD programs state-wide, based on the results of the NCD pilots. The AF continued to support three of the original four components, but the activities under each were enhanced and expanded or fine-tuned (see detailed description of components below). The second component changed from “Developing effective models to combat NCDs and accidents” to “NCD prevention and Control”, supporting implementation of NCD programs across Tamil Nadu, building on the successful pilots carried out under the original project. The additional funding allocated to each component was as follows: Component 1: Increasing Access to and Utilization of Service (AF of US$44.79 million, totaling US$88.58 million allocation for this component).  Reducing Maternal/ Neonatal Mortality: support effective ongoing provision of obstetric and neonatal services by the 80 CEmONCs established under the project, train doctors and nurses, finance contractual staff salaries at CEmONCs for two years (subsequently to be financed by GoTN), construct and equip higher maternity referral institutions at 8 medical colleges, design and provide Information, Education and Communication (IEC) materials, and broadcast and disseminate information.  Improving Tribal Health: implement the Tribal Development Plan (TDP) in all identified tribal areas (12 districts) to increase access to health care, and strengthen existing primary and secondary services in tribal areas through PPPs. The AF supported (i) provision of additional vehicles, equipment, operating costs, and TV/DVD sets for mobile out-reach health services based on need; (ii) sickle cell anemia screening at three tribal hospitals; 7 (iii) tribal counseling services; (iv) a bed grant scheme for inpatient care for tribal populations at selected hospitals; (v) performance-based payments, training and incremental costs for village health volunteers implementing tribal activities; (vi) IEC activities; and (vii) monitoring and evaluation of the TDP.  Facilitating use of hospitals by the poor and the disadvantaged: (i) retrain patient counselors; (ii) provide patient counseling services; (iii) provide 200 additional ambulances and mortuary vans; (iv) contract NGOs to provide mortuary van services; (v) strengthen laboratories at selected hospitals; and (vi) finance and monitor housekeeping services at selected hospitals. Component 2: NCD Prevention and Control (AF of US$22.01 million, totaling US$27.66 million for this component). This component omitted the traffic injury prevention activities planned under the original component, continued support for health promotion, and added support to implement NCD screening and treatment programs across the state. The activities to be funded were as follows:  Health promotion activities to prevent NCDs, training teachers and peer educators for school-based activities (through the Education Department), interventions in workplaces and community-based interventions through the Rural Development Department and by NGOs.  NCD Interventions – on the basis of results of successful pilots, scale-up NCD interventions state-wide: (i) provide necessary equipment at identified primary and secondary level facilities, train doctors and nurses, fund honoraria for village link volunteers/community resource persons supporting cervical cancer screening and breast- cancer detection; (ii) provide necessary equipment and training for medical officers, nurses, and laboratory technicians at identified primary and secondary level facilities for cardio vascular disease (CVD) screening (including diabetes), and finance two years of salaries for contractual nurses based on needs; (iii) IEC posters, stickers, flip charts, information boards, broadcasting and dissemination for the scaled-up NCD interventions; and (iv) monitoring and evaluation of NCD interventions. Component 3: Building Capacity for Health System Oversight and Management (AF of US$33.80 million, totaling US$59.41million). The new and revised activities funded under this component were as follows:  Strengthen M&E Capacity in DoHFW – provide software, IT services and equipment to roll out Phase II of the computerized Hospital Management System (HMS) in the remaining 222 Project hospitals (total of 270 hospitals), selected Medical Colleges (tertiary level hospitals) and attached hospitals.  Improve Quality of Care – support continuous monitoring of quality of care, provide training in management and rational use of drugs for hospital and PHC staff, and enhance capacity for the hospital accreditation process within DoHFW.  Strengthen Health Care Waste Management (HCWM) – expand training on infection control and waste management to all health personnel at primary, secondary and tertiary levels of healthcare, and carry out an impact evaluation of the implementation of the Environment Management Plan.  Capacity Building for Strategy Development and Implementation – expand the Directorate of Medical Services Annex building by adding two floors; train doctors and 8 Tamil Nadu Health Systems Project (TNHSP) staff, finance administrative costs and additional staff for the TNHSP Society to enable it to scale up project activities, studies, evaluations, monitoring, dissemination of project lessons learned and achievements, convene an International Health Conference in 2010; and establish a data resource center.  The community based health insurance pilot was dropped because of its likely negligible impact on the project objective, especially given the new TN Chief Minister’s Health Insurance Scheme. Instead, the project provided complete administration and management support to the health insurance scheme, rolled out with technical assistance from the Bank in 2012. Component 4: Improving Effectiveness and Efficiency of Public Sector to Deliver Essential Services (AF of US$30.18 million, totaling US$81.08 million). This component supported new and revised activities under two of the original four sub-components:  Equipment Rationalization and Strengthening of Equipment and Pharmaceuticals Management – provide essential equipment (Intensive Care Unit (ICU), x-ray and poison treatment centers); strengthen logistics and procurement of pharmaceuticals and equipment; and strengthen repair and maintenance system in the TNMSC (including financing salaries of biomedical engineers).  Human Resource Planning and Development -- finance additional contractual staff (doctors, nurses, pharmacists, laboratory technicians, radiographers, auxiliary nurse midwifes (ANMs), hospital workers, sanitary workers, dental surgeons and cooks) in project hospitals for the first two years in accordance with established staffing norms in order to improve overall efficiency and performance. No civil works to rationalize secondary care facilities were included, as all priority works had been completed. 1.7. Other significant changes  In January 2005, $20 million equivalent (at prevailing XDR exchange rate) from the original Credit was cancelled for reallocation to the Emergency Tsunami Reconstruction Project in response to severe damage caused by the tsunami of December 26, 2004. It was understood that subject to satisfactory implementation of TNHSP, the Bank would provide AF to fill the financing gap created by this cancellation.  Minor changes in the Results Framework (noted in data sheet section F) were agreed with GoTN and approved by the Country Director (CD) on May 18, 2007 (Management Letter and Aide Memoire, 5/18/2007). This was not processed as formal restructuring. The changes reflected recommendations from a Bank-wide health portfolio review in 2005-2006, and added newly available baseline data. The details of the main indicators were aligned better with core activities being supported by the project. Specifically, “careful monitoring of the effectiveness of other NCD prevention activities” was dropped from PDO indicator 2 to keep the focus on monitoring and evaluation of the NCD intervention pilots; “implementation of a regulation/accreditation system to improve quality of care in private sector hospitals” was dropped because accreditation was to focus on public (not private) hospitals.  Traffic injury prevention was restricted to surveys of helmet and seatbelt use, instead of the originally planned state-wide BCC activities, to avoid duplication of effort under a Bank financed transport project. Although sales and use of helmets were rising, there was 9 strident public opposition to the new helmet use law. GoTN succumbed to public pressure and backed away from stringently enforcing the law on an unwilling population. As noted in section 1.6 above, this activity was dropped in the 2010 AF, and the project focused on NCD activities where there was better traction.  The Community based health insurance pilot in the original project was dropped – an AM in 2007 noted that it was likely to have only a marginal impact on the PDO, especially in light of the state-wide Chief Minister’s Health Insurance Scheme that was rolled-out with technical assistance from the Bank in 2008, and to which TNHSP provided complete administration and management support.  Closing Date extensions: The Closing Date was extended by three years to September 30, 2013 as part of the AF. After the Bank removed the time limit of 3 years for AF, the Closing Date was extended two more times – first by one year to September 30, 2014, and then by 11.5 months to September 15, 2015. These extensions were needed to enable full completion of innovative interventions that had taken longer than expected to start up, some activities that had been delayed by back-to-back state and municipal elections in 2011 and national elections in 2014, periodic delays when the procurement workload was especially heavy, and to ensure full scale up of the NCD interventions throughout the state and of the HMS/College Management System (CMS) and University Automation System (UAS) to tertiary level public health facilities. The second extension was also to ensure that the TNHSP could consolidate the project gains and do a thorough hand- over to DoHFW and the National Health Mission (NHM).  The Level 2 restructuring (approved August 7, 2014) streamlined the Results Framework by selecting the most directly relevant outcome and intermediate results indicators (in addition to the Closing Date extension noted above). 2. Key Factors Affecting Implementation and Outcomes 2.1. Project Preparation, Design and Quality at Entry Background. The project was prepared in a little over two years (2002 to 2004)5, a reasonable time for the India portfolio, especially given the scope, cutting edge reforms, innovative approaches, and thoroughness of preparation. The GoTN’s Health Policy (2003) laid out a road map for the next two decades toward reducing IMR to 15 per 1,000 and MMR to 50 per 100,000 live births, improving the health status of the general population and especially low-income communities and families, starting to address key non-communicable diseases, while sustaining vigorous efforts to control communicable diseases, and strengthening first referral hospital services (district and sub-district hospitals) as a priority. The PDO and design of the four components focused fully on these priorities. The project indicators were well aligned with the objectives and design of the project components. Project design was thoughtful, clear and straightforward. Although the project included a large number of activities at all levels of the health system from communities to tertiary care hospitals, the design was very “tidy” and coherent. Each of the four components included three 5 Project identification was in October 2002, Project Concept Review in March 2003, appraisal in June 2004, negotiations on November 3, 2004, Board approval on December 16, 2004, and effectiveness on January 27, 2005. 10 or four mutually supporting sets of activities. Activities were clearly identified, with a tightly linked results chain that gave the project strong clarity of purpose and design. Management and institutional arrangements were well-specified, capacity gaps were identified by thorough assessments documented in PAD annexes, and early activities included to address them. There were no co-financiers, and only two safeguard policies were triggered and well addressed. GoTN was fully committed and provided strong leadership throughout. GoTN established a project preparation team led by a senior officer of the Indian Administrative Service (IAS) with members from all levels of the health system, academics, and representatives from the DANIDA funded primary healthcare project. GoTN organized workshops starting in early 2002 to discuss possible project design, content and the results framework. Extensive detailed analyses during preparation informed the project design. These included:  Burden of disease study on all causes of mortality and morbidity;  Analysis of regional imbalances and required interventions that identified districts in need of certain interventions;  Criteria were established to identify four districts including two tribal districts for investments in the first year, targeting tribal communities and disadvantaged groups with the worst health indicators;  In addition to a social assessment study, several supplementary studies provided information on ways to increase access to quality health care for SC/STs and other underserved groups;  District mapping of public and private hospitals;  An analysis of facility planning needs in pilot districts and a health facility survey identified needs for strengthening facilities;  An environmental assessment included waste management practices and patterns in a sample of health facilities to inform needed improvements;  Study on drug prescription and dispensing practices;  Service norms and associated staffing and equipment norms were developed for rationalizing services;  Public and private service providers were mapped;  Private health care providers in 15 of the (then) 29 districts were enumerated, a qualitative study done on the private sector including informal service providers, and a study on practices and attitudes of informal rural medical practitioners in Tamil Nadu was completed;  An institutional assessment identified needs for strengthening capacity of GoTN for PPPs, anticipated staff availability and skill mix issues that might arise in implementing the revised service norms and planned how to resolve them, systems for personnel management were developed, and a mechanism for stakeholder/community participation put in place to enable feedback from communities on the quality of care;  A public-private partnership policy framework and terms of reference (TORs) for a proposed PPP oversight unit were prepared;  An Environmental Action Plan; and  A Tribal Development Plan. Preparation for the AF included development of a detailed Governance Accountability Action Plan (GAAP), as required in the region and as part of the recommendations of the Detailed Implementation Review (DIR). 11 Stakeholders6 were deeply engaged in project design and preparation, with extensive use of participatory processes. Development partners provided input in their areas of expertise; for example, DANIDA was consulted for technical input on the HMIS. Project preparation included extensive consultation with NGOs providing health outreach services in tribal areas, and with private providers on their experience of partnering with GoTN to provide services (such as emergency transport and reproductive and child health (RCH) services). GoTN made special efforts to increase participation and enhanced ownership of the Social Assessment by actively engaging staff at all levels of the DoHFW, beneficiaries (individuals and communities), donor agencies, NGOs, community organizations, local authorities, the private sector and academic institutions in consultations. The PAD detailed the extensive use of participatory processes in project preparation including for the Tribal Development Plan, researching how best to increase demand for services and address the special health needs of the tribal population. The project design reflects careful thought about behaviors, behavior change, and incentives. In addition to activities to expand and improve supply of services, the project also included efforts to increase demand for services, with activities to inform underserved groups and encourage them to access services. Behavior change (BCC) interventions for providers aimed for greater responsiveness to poor. Noting that complex “soft” investments may get less attention from implementers, the project team used a phased approach which began with both infrastructure and non-infrastructure (“soft”) inputs, but required the “soft” investments to be completed before the next phase of infrastructure investment could begin. QER and other project reviews: The Bank-wide Quality Assurance Group (QAG) did not conduct an assessment. Project design benefitted from the recommendations of a quality enhancement review (QER) arranged by the Health, Nutrition and Population anchor in 2002 shortly before appraisal. The QER panel were “impressed by the scope and range of preparation work… and … many positive aspects of the project. The Panel is confident that, if the points discussed are addressed, a project of good quality at entry will result.” The appraisal package included the QER report, and the PAD shows how thoughtfully the recommendations were taken on board.7 A Quality Assessment of the Lending Portfolio (QALP-1) in 2008 concluded that: “the project was well designed. The design built on the state's successful track record in health as well as the lessons learned from nine other Bank-financed state health system strengthening projects in India. The panel was pleased to note the strong focus on the poor, the involvement of NGOs, and the attention to the growing problem of NCDs through well-designed pilot programs. The project rightly focuses on emergency maternal and neonatal care including the need for an effective transportation system. While innovative, the panel did not consider the project to have been over-ambitious.” (QALP-1, p4) The QALP judged the project’s FM design to be very good. 6 Stakeholders included members from all levels of the health system, academics, the public and private sectors, civil society, NGOs, academics, and DANIDA. 7 In additional to suggestions on improving explanations in the PAD, the panel recommended: rethinking some of the indicators; greater clarity about how activities in areas such as personnel, information and financing could help integrate centrally funded programs and state health services; more attention to ensuring adequate levels of staffing in facilities used predominantly by the poor; policy dialogue and attention to the state budget for health and financing mechanisms to reduce out-of-pocket payment for health; considering simplifying the project implementation arrangements, and stronger economic analysis. 12 The project design incorporated the lessons learned from a systematic review in 2002 of all Health Systems Development Projects implemented in India since 1995 that considered future directions for health projects in India. The TNHSP was the first State health system project designed after this review, and deliberately and explicitly reflected its findings and recommendations – facilitated by having the same Task Leader as the review. Six key recommendations/lessons are clearly reflected in the project design: (i) project objectives should focus on health outcomes among the poor, and special efforts are needed to reach the poorest: several studies and participatory activities during preparation measured access to and utilization of health services by the poor and sought their input on how to increase their access and use, the project activities target least-served areas and populations with poor health; (ii) new ways to enhance health outcomes need to be explored: the project included pilot testing of innovations in NCD care on a reasonably large scale with rigorous evaluation to assess their effectiveness; (iii) deepen public-private partnerships (PPP) by going beyond contracting out “hoteling” functions which do not impact clinical care: the project contracted with private and NGO providers to deliver clinical services particularly in underserved areas and enhanced GoTN capacity in PPP; (iv) special attention should be paid to sector planning which was often weak at state level: the project set up a strategic planning unit to function as a policy advisory body and think tank; and to improving the management of public hospitals: the project included measures to enhance public hospital management including twinning with well-performing private hospitals; (v) implementation of non-infrastructure “soft” investments typically gets too little attention and is often a weakness of project performance: the TNHSP team thought carefully about how to phase implementation and incentivize completion of “soft” investments, as noted above; (vi) centrally sponsored health schemes (CSSs) and programs and state-financed health services should be integrated better: the project’s many activities to strengthen the state health system would enable better CSSs service delivery by improving provider skills and availability, equipment and supplies, the health information system, state level health planning, etc.. Risk and mitigation measures. Risks and mitigation measures were appropriately identified. The risk of inadequate budgetary allocations was discussed early with GoTN, assurances were received, and this was regularly monitored and did not become a problem. Capacity for implementation was assessed and additional staffing, training and improved management systems included in the project. New procedures were discussed in detail during preparation, and groundwork completed during preparation. Detailed preparation for procurement was completed well before project effectiveness (including all Terms of reference, bid documents, technical specifications, Requests for Proposals and the procurement plan), to try to avoid early delays. 2.2. Implementation Especially given the complexity, innovations and long time period, implementation was very good. As rated in the Implementation Status and Results (ISRs) reports, implementation was satisfactory or moderately satisfactory during the initial five years of the project (before AF) except for two ISRs in 2006 that rated Implementation Progress (IP) moderately unsatisfactory. After AF approval, project implementation was consistently satisfactory (5 years and 9 months). Ratings for Development Objectives (DO) were satisfactory in 20 of 23 ISRs and moderately satisfactory in the other three. Annex 2 lists the outputs achieved under each project component. 13 Factors contributing to consistently strong project implementation:  Consistently strong commitment of the Government of Tamil Nadu (GoTN), irrespective of which of the two main political parties were in power. Successive governments in TN have given consistent high priority to health (and other social sectors), and to implementing the 2003 health policy. Health has been “above politics” – shifts in political power have not affected the emphasis on health or the continuity in policy and its implementation. Successive governments have retained the 2013 health policy, and built on the actions of previous governments, continuing to strengthen service delivery to improve health outcomes. The GoTN assigned experienced, very high caliber senior officials to manage the project and to staff the Project Management Unit and other key posts. The continuity and low turnover among staff (from 2007) was especially important given the scope and complexity of the project (and a sharp contrast, for example, to the 9 project directors in 2 years in a health project in another state in India). GTN’s complete commitment to ensuring the success of the project comes through clearly in ISRs and AMs, and is noted in the QALP-1.  Strong mutual respect and trust between the Bank team and GoTN. The AMs, numerous other project documents, and information from people involved in the project attest to the productive professional relationship between the Bank and GoTN. This enabled frank discussion and constructive joint problem-solving when needed. It is noteworthy that this relationship was not at all disturbed by tensions related to the Detailed Implementation Review (DIR) of five health projects in India 2006-2007.  As a state-level project, it was not affected by the DIR tensions and tendency to centralize and tightly control fiduciary functions that affected several national projects. The State’s independent management enabled the project to continue its focus on strengthening the state health system, including fiduciary aspects, without disruption (although the project was required to comply with enhanced reporting requirements on procurement and financial management for several years after the DIR).  The project was fully integrated within Government structures at all levels. Project activities were an integral part of the DoHFW’s work and activities. This further strengthened project ownership throughout the DoHFW, boosted commitment and implementation performance, and enhanced the likelihood of sustainability after project completion. All project activities were mainstreamed and their full financing absorbed into the state health budget or nationally-funded health programs during the final years of the project.  Use of a phased approach for most project activities to learn and adapt before scaling up in all districts in the state. Most activities followed this approach -- training, CEmONCs, HMIS, NCD interventions, health care waste management, rationalization of health care facilities, and improving equipment maintenance and repairs. The phased approach to infrastructure improvements, whereby system reforms and “soft” investments had to be completed before the second phase of infrastructure upgrading could begin, was an effective incentive for successful and timely implementation of activities that often lag.  Well-functioning routine monitoring and information flows and feedback triggered clear actions to continuously improve performance. (Details are provided below, in the section on M&E utilization.) 14  Extensive independent assessment and validation as important input into decisions. The project commissioned numerous independent assessments of activities by academics and other experts, that provided data and unbiased views on strengths and weaknesses of implementation processes and outcomes. The project team used the findings constructively to make improvements and decisions on whether to scale up, drop, or change activities.  A strong focus on capacity building and skill enhancement through training and retraining. The training covered activities essential for efficiently delivering health services across all public facilities. It included specialized skills for doctors and staff nurses in CEmONCs; hospital administration and management skills for administrators, senior medical officers and nursing superintendents; a range of skill training for clinical, paramedical and laboratory technicians needed for NCD screening and treatment; use of the HMS/HMIS; quality of care and accreditation interventions; infection control and health care waste management; poison treatment, and rational use of medicines for all relevant health professionals; and training in proper equipment use (ventilator, dialysis, echocardiogram) as needed. Continuous quality assessments helped identify training and retraining needs.  Continuity and a strong Bank team. There was one task team leader from project identification (2002) to approval of the AF (2010), and two until project closing (2010-2015). The Bank team was mostly based in the Delhi office, so all the necessary skills (operations., financial management, procurement, environment, social safeguards, information technology, and health care expertise) were readily available during regular supervision missions, and to respond quickly as needed in between missions. The Operations Officer who joined the team in 2007 was exceptionally effective in providing continuous supportive supervision, and ensuring continuity and “institutional memory” for the team including during changes in the task team leader. The Bank’s Lead Health Specialist was also located in Delhi until the final months of the project. The team had excellent working relationships with the PMU and DoHFW. Supervision was systematic, detailed and regular, and the MTR was on schedule, well-planned, intensive and detailed. There was strong follow-up between missions and quick resolution of items identified for action. The QALP panel rated supervision inputs and process as HS, and all other aspects S, noting that: “The Bank task team was proactive and dealt in a timely and creative way with the hurdles encountered. The panel was especially impressed with actions such as the team visit to Hyderabad to review a successful model of emergency transportation that was subsequently adopted for this project.”  Candor in project assessment ratings. The project ratings were realistic throughout the project, and the team did not hesitate to recognize potential or actual problems and forcefully bring them to the PMU and DoHFW in a timely manner (the IP rating was downgraded to moderately unsatisfactory in two ISRs in 2006). Early detection of potential problems and candid discussions throughout the project contributed to keeping the project on track. The QALP panel “began with the view that some of the ISR ratings were slightly optimistic…, but concluded that the team, overall, was justified in its ratings”, rating PDO ratings as HS and IP ratings as S, noting that the team’s ratings sometimes seemed premature to the sector manager but were validated by subsequent implementation.  Provision of Additional Financing and recognition of good project performance. The intention was always to process AF to replace the $20 million reallocated to an Emergency Tsunami Reconstruction Project in 2005, if progress was satisfactory. The Bank agreed to a 15 much larger AF operation to continue successful activities and scale-up well-performing ones, notably the piloted NCD screening, prevention and treatment; the HMIS; and maternal and neonatal health services. Justifiable pride in the project’s accomplishments and reputation as one of the best performing projects, and the associated “Hawthorne effect” likely contributed to continued strong implementation performance.  Project extensions enabled the HMIS roll-out to be completed and even expanded beyond the originally intended scope. This is especially impressive in the light of numerous failed HMIS efforts elsewhere. Extensions also provided enough time for smooth institutionalization and absorption of activities by the DoHFW, carefully and well-informed assessment and planning for future financing by the government, and for all planned assessments to be completed. Factors that were outside the control of the government and caused implementation difficulties.  The massive tsunami that struck eleven coastal districts of Tamil Nadu two weeks after project approval in December 2004 caused large scale destruction. Diversion of GoTN’s attention, efforts, and resources to disaster relief, recovery and reconstruction activities delayed project implementation at the start, which would have been an intense period of activity. Government attention was diverted entirely to managing the damage caused by tsunami for at least the first half of 2005.  Delays caused by the freeze on all procurement activity for the 45-60 days before all national, state and municipal elections under the “Model Code of Conduct”. The Model Code of Conduct caused procurement activities to be put on hold – sometimes for several months, delaying project implementation. The project was very proactive in requesting exemptions to the freeze, but these were not always granted, including in 2010 and 2014. Factors that caused implementation difficulties included:  Turnover of Project Directors in 2005-2006. The turnover of Project Directors (senior officers from the Indian Administrative Service) in the first two years of project implementation slowed project progress. There was however stability thereafter.  Ineffective financial management for 15 months. The long-vacant position of Financial Advisor and Chief Accounting Officer (FA&CAO) early in implementation resulted in inadequate attention to financial management, notably a delay in submitting the 2006-07 audit report, delays and inadequate follow up in settling advances drawn for training etc., and inadequate project financial oversight over NGOs. These problems were resolved after GoTN posted a well-qualified professional to the project in early 2008.  Slow initial procurement affected disbursement. Delays in procurement actions in the first 18 months of the project (especially for civil works and baseline studies) slowed initial project implementation and disbursements. This was exacerbated by the procurement freeze required by the Model Code of Conduct before the elections in 2005 and early 2006. Procurement issues were subsequently resolved as a result of proactive actions by the PMU. The delayed start of the project in 2005 due to the tsunami, and the procurement issues noted above slowed disbursements particularly during the first 2 years of the project (2005-2006). Weekly audio follow-up meetings between the Bank team in Delhi and the PMU in Chennai starting December 2006 led to an improved pace of procurement and disbursements. 16  Difficulties with contracts to NGOs for providing emergency medical transportation. The MTR concluded that the NGO contracting and contract management process was not working well, and needed more oversight and monitoring and action to ensure strong performance and achievement of the intended results. It recommended that the PMU consider hiring an external agency to take over this task, and use performance based funding with clearly defined selection and monitoring criteria, and robust financial control systems. The GoTN contracted Emergency Management and Research Initiative (EMRI) who were operating very successfully in Andhra Pradesh, to provide these services, which proved successful.  The complexity and time needed to design, test, trouble-shoot, implement and make good use of a new HMIS were underestimated. Conceptualizing, designing, and back-end work to set up an HMIS/HMS system – especially one that links hundreds of facilities across the state and multiple levels of the health system, replaces many legacy systems, introduces new work flow and technologies, and relies on adequate server capacity and extensive end-user behavior change – was an enormous task. After a slow start and significant implementation challenges, the MTR team worried that there might not be enough time to fully roll out the system, and persuaded GoTN to pilot the system in five hospitals before proceeding.  Turn-over of specialist doctors at CEmONCs. Despite the proactive efforts by DoHFW, loss of specialists – particularly anesthetists – has left some CEmONCs short of the agreed staffing during some periods. The agreed process of regular recertification of CEmONCs (every six months) kept a spotlight on staffing relative to norms. This will require constant monitoring and proactive action by DoHFW. 2.3. Monitoring and Evaluation (M&E) Design, Implementation and Utilization M&E design. Overall, the M&E design had many strengths, and reflected careful and logical thought, consistent with “best practice thinking” about health systems. The “Flagship Framework” on Health Systems Strengthening (developed by the Bank’s training institute with Harvard University and other experts a few years before the project) defines three ultimate goals of a health system: to improve the health of the population; to provide “financial protection” (that is, to ensure that health care does not cause financial harm); and to provide patient satisfaction. These three “ultimate outcomes” require a health system to provide good access, quality, and efficiency, which the framework calls the “intermediate outcomes” by which to judge a health system’s effectiveness. The project indicators cover all of these 6 outcomes. Financial protection is measured less well than the other 5, in the numbers of people who benefitted from the free ambulance and mortuary transport services, and the bed-grant scheme.8 The original project Results Framework (RF) in the PAD was comprehensive and logical, developed in consultation with key stakeholders, and with input from the QER and other experts. It reflected a clear results chain, with four appropriate outcome indicators to measure PDO 8 Financial protection was not a central explicit focus of the project -- activities to overcome financial barriers were seen as increasing utilization/access. A community-based health insurance pilot added at the suggestion of the QER panel was later dropped because it was judged likely to have little impact on achieving the PDO, especially in light of the roll-out of the Special Minister’s health insurance scheme. 17 progress – one for each component, and 22 fairly well-chosen intermediate outcome indicators (IOIs) – one or two for each main activity. The PAD Annex 3 lists data sources, frequency, and clear responsibility for data collection for all indicators. The indicators for increased access and use of services, especially by the poorest and Tribal groups, and for maternal and neo-natal mortality rates (adjusted for risk) were very important for assessing the core impact of the project. Excellent systems were put in place as part of project design for routinely collecting data on quality of care, utilization rates at health care facilities, and hospital activities, and reviewing, analyzing, agreeing on actions and then continuously following up. This enabled remarkable quality and impact improvements in a very short time. The project design also included major investments to improve monitoring and evaluation capacity through a new, integrated, system-wide HMIS to replace manual paper-based, time- consuming reports that involved minimal feedback or basis for action. The new HMIS was designed to track patient, service and management information, network all hospitals to track referred cases, monitor changes in health outcomes, and provide feedback to service providers and program managers for improved follow up and continuity of care. M&E design also included an impressive number of independent evaluations of selected project activities that assessed the impact of innovations – these included surveys and studies on NCDs, patient satisfaction, and health services available to and used by tribal populations. However, M&E design had some imperfections, despite the team following advice from the QER. At appraisal, baseline data were available for only a few of the indicators, and most targets were vague instead of SMART.9 Many targets were arbitrary placeholders (“10% increase”) pending collection of baseline data, but relatively few were replaced later with carefully chosen targets, as had been intended. As noted by the sector manager in ISR#6 (after baseline data had been collected), targets that aimed only for an undefined increase or to maintain the baseline value were disappointing in their lack of ambition. They probably reflect risk-aversion, and/or inability to decide what might be feasible in the absence of evidence and experience on which to draw. But this is a weakness in the project M&E design that was not well-addressed despite creditable efforts by the team to improve the results framework during project implementation. Despite the faults that can be found with the M&E design, it was stronger than in most projects, especially considering how little experience in health system strengthening projects the team was able to draw on a decade ago. The 2008 QALP noted that: “The task team, by its own admission, struggled with getting the results framework right, not an easy task in a project of this type. To its credit, the [Task Team] employed many experts and continue to revisit, refine and improve the framework” – which was done in 2007, at AF, and again in 2014. The 2007 RF revisions made well-considered improvements: over-broad and vague parts of two of the IOs (e.g. “careful monitoring of the effectiveness of other NCD prevention activities”) were dropped, base-line data were added, and indicator definitions refined. However, some “Quality of Care” indicators measured efficiency and not quality (bed occupancy, number of surgeries) or were ambiguous (night-time C-sections are only a valid quality indicator if they are emergency and not elective procedures). Even the number of diagnostics tests in itself may not 9 Good practice requires targets that are Specific, Measureable, Agreed/Achievable/Assignable (clearly defined responsibility), Relevant/Realistic, and Time-bound. 18 indicate quality. In its zeal to monitor all of the many project activities, and probably over- compensating for slow implementation in the first 3 years, the AF added an excessive number of indicators – for several years the PMU diligently reported on 72 items. (To their great credit, the documents prepared for each mission were impressively complete and detailed, with thorough attention paid by both Bank and project teams to each item.) The decision to drop 38 IOIs and supplementary indicators as the project neared its end, to focus on core activities and impact measures, was understandable. However, instead of formally dropping the indicators, the Bank could have agreed with GoTN that reporting was no longer required on the activities already completed, and retained September 2014 as the target date for activities on which no further detailed reporting was warranted. M&E implementation. The planned M&E was fully and well implemented. Data reporting to supervision missions and in ISRs was impressive: comprehensive project status reports including data for the RF and all project activities were provided every six months throughout the project. The range of data sources used to monitor and report regularly on progress and results included: routine health system records, routine project data, on-line monthly reports from the hospitals on a set of 20 indicators (Institutional Services Monitoring Report or ISMR), facility surveys, baseline and endline surveys, and many detailed studies of selected project activities. Data were used well to monitor progress, check that activities were achieving their desired results, and decide where corrective actions were needed (details below). The PMU took good advantage of expertise in the state, for example, partnering with the Christian Medical College in Vellore, the Indian Institute of Technology, and the National Institute of Epidemiology (NIE) in Chennai to evaluate pilots and other activities. Three weaknesses are noted: (i) delays in baseline surveys – 15 months for the NCD baseline studies for the pilots and for patient satisfaction surveys, (ii) changes in methodology in repeat surveys of patient satisfaction that make trends difficult to assess; and (iii) mistaken entry of data on the poorest 40% in the results reporting for SC/ST. This latter was noted when preparing the 2014 restructuring; detailed scrutiny found mistakes in the analysis of NSSO data in a 2007 consultant report. The Bank commissioned a careful new analysis (see Annex 10) to correct the data for the indicator on access and utilization of services by the poorest 40% and ST/SC groups. An impressive aspect of M&E implementation was the successful comprehensive Health Management Information System (HMIS), despite its challenging and increased scope during the project. Rolled out in a phased manner from December 2008 onwards, the HMIS comprises (i) a Hospital Management System (HMS) which automates reporting on clinical activities in public health care facilities; (ii) a Management Information System (MIS) which is an online reporting platform for clinical and ancillary support services, national health programs and administrative information for all public health facilities; (iii) the College Management System (CMS) to capture data from government medical colleges; (iv) the University Automation System (UAS) for data from the Tamil Nadu Dr. MGR Medical University; and (v) customized web-sites for 20 government medical colleges. By July 2015, the HMIS was fully functional in 264 secondary care hospitals and at an advanced stage of implementation in the state’s 50 tertiary care hospitals; the MIS had integrated 1,889 primary health centers, 264 secondary care hospitals, and 50 19 tertiary care hospitals; and the CMS was operational in 20 government medical colleges. All PHCs were reporting through the HMIS effectively, with all reports flowing to the Directorate of Public Health (DPH), and all data from secondary care hospitals flowing to the Directorate of Medical & Rural Health Services (DM&RHS). Over 165 million patient visits, 62.8 million laboratory requests, 81.6 million pharmacy dispensations and 4.1 million in-patient visits had been recorded in the system. The new HMIS provides quick access to information of all important aspects of the health system -- hospital activity and efficiency indicators (in-patient and out-patient data, referrals, waste management, quality of care, morbidity/mortality), financial management information, and human resources. In the final years of the project, project monitoring data were provided exclusively from the HMIS. The HMIS system is a major project achievement, and has received national and international awards. A State Health Data Resource Center (SHDRC) was set up to collate, mine, and run higher order analytics on data from over 20 Directorates of the Health Department. The SHDRC provides easy to use dashboards for various levels of administrators and managers in the health department. Its mandate is to drive and enable evidence-based planning, budgeting, management, forecasting, monitoring and reviews by the DoHFW. The Center is managed by a consulting firm, contracted (in 2015) to run the Center for two years, and then hand it over to the state, but continue to maintain and support the activities of the Center until March 2021. M&E utilization. The PMU was effective in using the data from all sources to make improvements during project implementation. A few of many possible examples follow:  During the NCD pilots, an evaluation by the National Institute of Epidemiology (NIE) found that patients were being lost in follow up for treatment, in response the PMU decided to issue 30 day supplies of medicines for hypertension and diabetes so that patients did not need to visit health care facilities more frequently, and set up an online tracking system for patients tested positive during screening. In response to the evaluation survey findings that (i) shortages of staff nurses were affecting screening and treatment under the NCD programs, and (ii) inadequate skills among health professionals for the NCD interventions, the PMU sought approval from the State Empowered Committee to recruit nursing staff on contract, and conducted periodic training programs to remedy specific gaps in knowledge and skills. Assessments of IEC activities led to changes in the messages and methods used, and also monitored the extent of changes in awareness and knowledge. The results of the pilot evaluations were carefully incorporated into the design of the scaled up NCD programs.  Early in implementation, the PMU developed quality and utilization indicators to measure hospital performance. All public hospitals reported these data monthly (Integrated Services Monitoring Report). The project used the data to grade hospitals A to D every month; hospitals with C and D grades were followed up to assess constraints and agree actions to improve service delivery. Quality Circles of Excellence were set up in hospitals to track progress, and develop and implement improvement actions. This proved be effective in improving performance and quality of care at the hospitals.  The project and NHM and DoHFW instituted a practice of monthly reviews (by video/audio) of every maternal death in which senior medical officials and relevant health facility staff 20 discuss the causes and actions to prevent future similar situations. This contributed to the substantial fall in maternal deaths in the state.  The HMIS system assigns a unique patient identification number (PIN) to track all health services provided to each patient, and make the patient medical record available at all points of care. When the number of PINs began to exceed the estimated number of patients using the public health system, it was realized that some patients were registering multiple times (after losing their PIN). A concerted intensive state-wide campaign explained how the PIN was used and the benefits of having a unique PIN, as a result of which patients made sure to keep and use their PIN on each encounter (pasting them in notebooks, keeping them on a small laminated card, etc.). Despite the extensive use of data generated and reported for the project and by key project activities, there is still unrealized potential to use the data to improve efficiency, quality, and allocation of resources. For example, detailed data on the actual use of pharmaceuticals and medical supplies could enable more accurate projections of need and trigger re-supply. The established of the SHDRC is intended to realize this potential, which could make Tamil Nadu a global leader in this area. 2.4. Safeguard and Fiduciary Compliance The project was classified as a Category B and triggered two safeguard policies: OP/BP/GP 4.01 Environmental Assessment and OD 4.20 Indigenous Peoples. Both policies were handled well and rated satisfactory in all ISRs. There was full compliance with all Bank requirements. Environmental aspects. Improving management of health care waste and fully institutionalizing the activities across all programs and facilities in the state under the project were exemplary. A sound comprehensive Health Care Waste Management (HCWM) Plan was developed. It was implemented in a phased manner -- first as a pilot in 2006, and based on satisfactory pilot implementation, from 2008 it was scaled up steadily in 449 health facilities including secondary care, tertiary care and ESI hospitals, and thirty-bedded PHCs. At AF, health waste management was integrated with infection control, in line with emergent good practice, referred to as Infection Control and Waste Management (ICWM). HCWM/ICWM activities were proactively supervised by a Bank specialist. Implementation and adequacy were assessed at various stages including near the end of the project (2014). Over 49,500 health personnel from 449 public health institutions were trained and retrained in health care waste identification, collection, segregation, disinfection, and disposal, through a network of 9 Regional Training Centers established by TNHSP. Supervision missions found adequate availability of color coded bins and bags, trolleys, needle destroyers, protective gear, consumables, handbooks on infection control and biomedical waste management, IEC materials and training modules. PPPs were established with 30 Common Treatment Facilities (CTFs) where waste was collected, disinfected and disposed of. From 2013, the cost of implementation of the HCWM plan was financed through the NHM, with the project financing training only. Findings of a comprehensive end-line assessment in 2014 included the following: (i) all hospitals were implementing Infection Control and Waste Management (ICWM) and had access to CTFs, (ii) 95% of respondents had been trained in ICWM, 60% mentioned need for additional refresher 21 training and training for new recruits, 89% were fully satisfied with the quality, relevance and method of training, (iii) best practice of labelling bins was observed in 60% of facilities, (iv) 92% facilities had storage room for biomedical waste, (v) sharps disposal “hub cutters” to replace needle destroyers were available in 75% of hospitals,10 (vi) 78% of hospitals had an infection control officer for monitoring ICWM processes, and 80% had infection control committees. The assessment provided reassurance of well-implemented ICWM and offered minor suggestions for improvements (e.g. on-line training, better reporting of needle-stick injuries). Indigenous peoples. Consistent with the project’s objective of improving health care outcomes among vulnerable groups including women, ST/SC groups, the poor and populations in remote areas, a Social Assessment was conducted with good participation of key stakeholders. This informed preparation of a Tribal Development Plan (TDP) also done in a highly consultative manner (October 2003) to develop appropriate and carefully chosen interventions to increase access to health care in tribal areas. There were some challenges in implementing the TDP. NGOs varied in their willingness to partner with Government, which was eased by Project efforts to engage with NGOs regularly through consultations and meetings in the field (facilitated by the Bank). Training/capacity building of NGOs/field workers could have been improved. It took a long time for proper guidelines to be finalized and then communicated to the NGOs and field staff. Turnover of field staff of NGOs was a problem, and repeated capacity building was needed. The NGO consultants who were supposed to do field supervision of the various NGO activities were never fully on board, so adequate monitoring of various NGO activities remained a challenge. While the NHM has taken over various programs for tribal populations, effective absorption of the activities will require close collaboration between the DoHFW and the NHM. Despite the challenges in implementation, a 2014 end-line assessment of four of the five activities carried out under the project found strong results. A survey of the targeted population found high levels of use and satisfaction with the services and their quality. Financial management (FM). Overall, financial management was satisfactory. The financial management arrangements for the project were completely mainstreamed within the regular government funds flow and accounting systems and procedures. Twenty of the 23 ISRs rated FM in the satisfactory range (S/MS). Three ISRs (May and October 2007 and April 2008) rated financial management performance moderately unsatisfactory, largely due to slow disbursements. Slow disbursements during the initial years were mainly due to delays: in procurement actions, in releases from the Treasury, in the appointment of the Financial Advisor/Chief Accounting Officer, and in the submission of monthly financial reports from the TNMSC and PWD. These issues were satisfactorily addressed. The Bank agreed to the GoTN’s proposal to change the funds flow mechanism by creating a Society at the state level (an independent legal entity) for implementing all project activities other than civil works, equipment and goods. Delays in settlement of advances drawn on training etc. were also 10 To prevent accidental needle sticks, used needles need to be safely disposed of, and re-use prevented. “Hub cutter” needle-syringe disposal devices cut up the entire device, so the used needle does not have to be removed. 22 addressed. The PMU also took actions to strengthen control and monitoring of NGO contracts by holding regular annual performance reviews before renewal of contracts, reviewing the cost elements and building in an institutional fee to NGOs. All audit reports and IFRs were submitted but in some cases with a small delay. The Comptroller and Auditor General (CAG) conducted external audits per terms of reference agreed with the Bank, Department of Economic Affairs and CAG. Very few financial statements of the PMU were qualitied and in one instance, because of a special opinion, an accountability flag was triggered. All issues were addressed to the satisfaction of the auditors and their observations were answered and resolved. There were no unresolved audit objections. For the AF, a Governance and Accountability Action Plan (GAAP) was prepared, and as part of the GAAP, Interim Unaudited Financial Reports (IUFRs) and internal and external audit reports were hosted on the website of the project. For the AF, it was agreed to shift to report-based disbursements. Procurement. Procurement activities were under the overall direction of the Project Director. The PMU was directly responsible for procuring consultancy services, and coordinated other procurements. The GoTN Public Works Department (PWD) was the implementing agency for civil works under the overall control of PMU. The Tamil Nadu Medical Services Corporation (TNMSC), as the GoTN Procurement Agent, procured all equipment and goods. Capacity to handle procurement was assessed as adequate and the procurement risk as average. A procurement plan for the first 18 months was agreed prior to project approval. For the AF, the Electronics Corporation of Tamil Nadu (ELCOT) was the procurement agent for information technology (IT) hardware and associated supplies/services for HMIS. The AF assessed the overall procurement risk as substantial. As noted above, project implementation was slowed at the start by delays in procurement actions by PWD and TNMSC, lack of interest by potential bidders in the first 18 months of the project and in 2008, and slow decisions in processing two consulting services. The massive Tsunami that struck Tamil Nadu a few days after project approval, causing huge destruction and diversion of Government’s resources and attention to recovery efforts, greatly contributed to the initial delays. Procurement issues were subsequently resolved. The triggering of the Model Code of Conduct by the announcement of elections also put on hold decisions on procurement actions at critical times of project implementation. Despite these issues, the PMU’s strong team, GoTN’s commitment to the project’s success, and the Bank’s regular intensive implementation support throughout the project ensured satisfactory completion of all procurement activities. The Bank conducted regular ex-post procurement reviews and the PMU took actions as needed in a timely manner to address issued raised. Overall procurement performance was moderately satisfactory. 2.5. Post-completion Operation/Next Phase The GoTN is keen to continue its partnership with the World Bank, and is developing a proposal for technical and financial assistance for a second Health Systems Development Project. It is greatly to the credit of the GoTN that no immediate follow-up operation was needed to sustain the project activities, which had all been fully mainstreamed. Transition planning was an integral part of the project design, and GoTN ensured that programs continued without interruption, and with adequate financing. Well before project closing, the GoTN started to absorb project activities into the work and budget of the NHM and the DoHFW (e.g., mobile 23 outreach, counselling services, the bed grant scheme, heath waste management, emergency transport, HMIS,) with only essential inputs (technical support and monitoring and evaluation) continuing from the project. By the end of the project, all major activities funded by IDA under the project had been taken on by the DoHFW Directorates for continued implementation as regular departmental activities. All contract staff (female NCD staff nurses, bio-medical engineers, IT coordinators) and other additional staff approved for the programs were retained and transferred to the Directorates so that programs remained fully staffed. GoTN and NHM are fully funding these activities. Project assets have been handed over to the Directorates. The few project programs such as State Health Data Resource Centre, 108 Emergency Ambulance Services, and Free Mortuary Van Services not integrated into regular DoHFW operations are continuing, implemented by the TNHS Society, funded by the GoTN, and housed/located in the Directorate of Medical Services (DMS) Annex. GoTN is continuing the innovations started and supported under the project including ongoing public private partnership program contracts. ICWM is a good example of the integrated mainstreaming of project activities. ICWM is now implemented and monitored by the Directorates of Medical and Rural Health services, Public Health and Preventive Medicine, and Medical education. To ensure sustainable capacity for ICWM training, the project strengthened 9 Regional Training Centers. Institutionalization and strengthening of HCWM in public health institutions under the project was exemplary, and ICWM activities were integrated across all vertical programs in the state. 3. Assessment of Outcomes Given the change in the PDO under the AF in 2010, two project phases are assessed, before and after the AF: 2005 to April 2010, and May 2010 to closing in September 2015. The main difference was the progression from piloting NCD interventions in the original project, to scaling them across the state under the AF. The project performed well in all three outcome aspects— relevance, meeting and exceeding objectives, and efficiency. 3.1. Relevance of Objectives, Design and Implementation The project’s relevance is High in all aspects – objectives, design and implementation, for both phases. Relevance of Objectives – (rating: High) The project objective of significantly improving the effectiveness of the health system in Tamil Nadu responded fully to the state’s needs and policy priorities, and to the Bank’s assistance strategy, both at appraisal and now. It was ahead of the strong global shift in emphasis to strengthening health systems (HSS). HSS was advocated in the Bank’s Strategy for Health, Nutrition and Population Results (2007) which also noted the importance of M&E systems (p. 6), and concern that the increasing burden of NCDs would strain countries’ health systems. The project focus on improving effective health services delivery, and access, utilization, and health outcomes for all, especially marginalized groups, anticipated today’s global focus and commitment to Universal Health Coverage. 24 The GoTN’s Health Policy (2003) identified strengthening hospital services and quality (especially at district and sub-district levels) and preventive health as priorities, aimed for ambitious reductions in IMR and MMR, to improve the health status of the general population with an emphasis on poor and the disadvantaged, and to address non-communicable diseases. The project fully reflects these, which remain current health policy. The GoTN DoHFW Policy Note on Health 2015-16 provides a comprehensive update on health outcomes and services in the state, including sections explicitly on the project, because the project is fully part of GoTN health policy. The Policy as set out in 2003 remains in force. The project objectives remained Highly relevant to successive Country Strategies for India (2009-2012 and 2005-200811), and to the Bank’s current Country Partnership Strategy (CPS) for India (2013-2017). The CPS focuses on using the Bank’s financing, knowledge, advisory services and technical assistance in catalytic and transformative ways to strengthen health delivery systems, improve access to services for excluded segments of the population, reduce rates of maternal and infant mortality, address the growing burden of non-communicable diseases, and improve delivery systems by strengthening accountability and M&E systems (p. 20-21, 27-28, CPS Summary). In lending to more advanced states, the CPS focus is on innovative and transformative activities to test second generation approaches which can then be applied to low-income states that often have limited capacity (p 30, 33). This was one of the explicit justifications for the project’s AF. Relevance of Design and implementation – (rating: High) The design of the original project and AF was highly relevant and remained so throughout implementation. The PDO clearly defined four sets of activities through which the overall goal would be achieved, and one component was devoted to each of the four. Each component included three or four intermediate results that were logically linked, with clearly defined activities for each. The activities explicitly addressed constraints on access, utilization and quality of health services in Tamil Nadu, with a dedicated set of activities to serve tribal groups. The indicators kept the focus on the poorest 40% of the population (a key target group for the Bank’s current “twin goals”). The PAD clearly explains how the project activities relate to the identified constraints. For example, data on the main causes of most maternal and neo-natal deaths informed the decision to provide free emergency transport and upgrade facilities to enable them to provide 24 hour emergency obstetric and neo-natal care. Data and studies documented Tamil Nadu’s growing burden of NCDs, so the project design included a dedicated component to 11 The project objectives were highly relevant to the third pillar of the Bank’s Country Strategy for India (2009- 2012, p. 2, 14, 16), consistent with the themes of India’s 11th plan (2007-2012). The first of 5 health targets in India’s 11th five year plan was to reduce the IMR to 28 and MMR to 100. The vision of the 11 th Plan was “to …. ensure broad based improvement in the quality of life….especially of the poor, Scheduled Castes and Scheduled Tribes, other Backward classes….create access to essential services in health…. especially for the poor… and good governance.” The project objectives were also highly relevant to the Bank’s Country Strategy for India (2005-2008) “…to reduce the health risks of the poor – by improving health outcomes including reductions in maternal and infant mortality, by improving the overall health system of the states, by focusing on the access to and quality of health services for the poor, by breaking new ground in forging public-private partnerships, and by reorienting health facilities to ensure service for the poor, to reallocating public resources to priority areas for the poor, and to improving governance and service delivery” (p. 31, 38 -39, CAS). 25 improve NCD prevention and better enable the health system to detect and treat important NCDs. Selection of activities was informed by current data on cost-effectiveness and “best buys” in health. The project design focus on partnering with the private/NGO sector through PPPs was well aligned with the Government’s approach and Bank strategy. It was a sensible approach given the limited capacity of the government’s own health service delivery network, and the very high use of private sector health services by the population at all income levels. The services that were delivered through PPPs with private/NGO providers were all things for which the public sector did not have a comparative advantage (emergency transportation, service delivery to tribal populations and in remote areas, disposal of health waste, counselling services, diagnostic laboratory tests). One of many examples of good project design was that the project made the capital investments (e.g. procured the ambulances) and then “contracted out” operation and maintenance to NGO/private partners (many of which would not have been able to finance the needed investments). All project investments in infrastructure and equipment were informed by careful assessments of existing situations and needs to achieve the desired reforms and improvements. Investments included all necessary inputs for the expected output and outcome: equipment, drugs and supplies, staff increases, training in skills and protocols, supportive supervision and quality assurance, as well as demand-side activities. Moreover, these mutually reinforcing inputs appear to have been well phased, sequenced and coordinated, which is an even greater design and implementation accomplishment. Reforms and “soft” activities such as training, behavior change, and new quality assurance mechanisms were thoughtfully timed relative to “hardware” investments in buildings and equipment. The phased approach facilitated effective implementation. Pilots and first phases were large enough in scale to test innovative and complex interventions, learn from them, adjust design details, and then appropriately scale up. Incorporating the new HMIS system as part of the project was important to being able to monitor and evaluate the effectiveness of the health system, and hence a key part of the project design. The phased approach to its development and implementation – and to other innovative or complex activities – enabled design details to be refined and improved during implementation. This approach, as well as the quality of the project design, and its forward-looking focus on strengthening the health system to be able to cope with emerging as well as current health care needs, resulted in very little need for changes in the project scope or design, despite its long implementation period of more than 10 years. The MTR in 2007, the QALP in 2008, and the AF processing in 2010 all gave the objectives, design and implementation thorough scrutiny, and all concluded that the relevance was strong. The change at AF from piloting NCD programs to scaling up state-wide was a natural progression. The few (minor) activities that were dropped were in response to changed circumstances, to avoid duplication of effort, and a clear-eyed judgement on their likely contribution to the PDO. Institutional and implementation arrangements were based on two sound principles: (1) placing project management responsibility within the DoHFW, given its responsibility for the state’s health sector, and (2) the full use of the different health system actors for project implementation. PMU staff were deputed from different health Directorates, and consultants were recruited only 26 when necessary because the DoHFW or other state agency did not have the specialized skills needed, or capacity to take on additional tasks. Project extensions were fully justified and provided sufficient time for critical institutional capacity building and mainstreaming of project activities, helping ensure a high likelihood of sustainability of the Bank’s investments. 3.2. Achievement of Project Development Objectives (rating: Phase 1 – Substantial; Phase 2 – High) As noted in the section on M&E design, the project indicators enable a fairly complete assessment of changes in the effectiveness of the health system. The impact of the project on the well-being of Tamil Nadu’s population is assessed through the three “ultimate outcomes” by which to measure health system performance – health outcomes, financial protection, and patient satisfaction. The indicators include patient satisfaction; for health outcomes the important indicators for maternal and infant mortality; and some (although inadequate) measure of the extent to which the project improved financial protection, through numbers who benefitted from various free services. The indicators also include various measures of the three intermediate outcomes by which to measure health system performance – access, quality and efficiency. Summary of Project Achievements against Results Indicator Targets Phase 1: 2005 - March 2010 Phase 2: April 2010 - 2015 PDO Intermediate PDO Intermediate Surpassed 2 1 4 16 Achieved 3 10 5 14 Partially Achieved 0 0 112 113 Not Met 0 114 0 114 Data not available 3 3 0 0 Total indicators 8 15 10 32 % surpassed and/or achieved 100% 92% 90% 94% (indicators with available data) Phase 1: effectiveness on December 16, 2004 through AF approval in March 2010 Phase 2: AF (when the PDO was revised) to project closing. Most project indicator targets were met and many surpassed (summarized in the table above and detailed in Data Sheet Table F). The only PDO indicator not fully met was the percentage of caesarean sections (C-sections) that were among ST/SC women. Given the large increase in C- sections from 15% to 26% of all deliveries in public facilities, the shortfall from this target is not considered a problem.15 The indicator on use of helmets is relevant for traffic accident fatalities 12 93% of target for “C-section deliveries among SC/ST mothers at secondary level CEmONCs”. 13 “Availability of staff according to norms at all project facilities” was met for 79% of project health facilities (211 out of 267) - an excellent improvement on 8% (22 facilities) in 2005. An IEC unit to coordinate all activities across the state was not set up in DoHFW. 14 Improved helmet use was not achieved (dropped at AF). 15 An independently conducted study (financed by the project) confirmed that C-sections done at CEmONCs were medically indicated in response to complications. Women with pregnancy complications were increasingly referred to CEmONCs – as intended. Treatment for pregnancy complications in Government facilities increased from 46% in DLHS3 to 64% in DLHS4, and treatment for post-delivery complication rose from 48% to 59%. Thus the CEmONCs were dealing with more complicated cases, and their rate of C-sections was within internationally accepted norms for both SC/ST women and other women. 27 but is not a measure of the effectiveness of the health system but of police enforcement of traffic safety regulations and related state policy. Assessed using the project indicators only, the project’s achievement of its objectives was much better than satisfactory. An assessment of whether the project made Tamil Nadu’s health system more effective follows, drawing on other data in addition to project indicators to assess each of the six aspects of health system performance. The original PDO in the PAD explicitly included “both private and public” parts of the health system. Most project activities were in the public sector, and project indicators measure impact on public sector effectiveness, as well as that of the NGOs and other private sector service providers contracted under the PPPs and informal health care providers (including traditional medicine practitioners) trained under the project. However, this excludes many private facilities and health care providers. The GoTN (like the rest of India) has almost no oversight over private sector health care providers. Changing this would have required a difficult, major policy shift and new enforcement capacity, and was not the intention of the project. The project could have had an indirect impact on the effectiveness of the private health sector if, by improving the quality and availability of services provided by the public sector, private providers improved their own effectiveness in order to compete. There could be a direct impact through a demonstration effect if private providers adopted treatment protocols or other good practices developed under the project, and also as a result of providers moving from the public sector into the private sector after their skills and capacity had been improved by project activities. No measures are available of the effectiveness of only the private sector, but all data from the NSSO and other household surveys (for example on health outcomes) reflect the effectiveness of the whole health system, both private and public. Ultimate Performance Measure 1 - Health Outcomes. Rating: High (both phases) The state NFHS surveys 3 and 4 (2005-06 and 2015-16) and the 2010 Census show marked improvement in TN’s infant mortality per 1,000 live births from 30 in 2005 to 24 in 2010 and 21 in 2015 (nearly half the India national rate of 40 in 2015), and in under-five mortality which fell from 35 in 2005 to 27 in 2010, with no additional decrease in 2015. Tamil Nadu had by far the largest decline in IMR of all states in India in the decade before 2010: 46%, a full ten percentage points more than the states with the next-largest falls in IMR (IMR fell 29% nationally). Tamil Nadu also has one of the smallest disparities between male and female under-five and infant mortality in India. There were also improvements in all measures of child nutrition status (see Annex 10). The maternal mortality ratio (MMR) fell steadily from 134 in 2003, to 97 in 2009-11, 79 in 2012-14, and 68 in 2015, less than half of the all-India rate of 167 per 100,000 births. Although population-wide data are not available for trends in NCD outcomes, the available data on risk factors and from evaluations commissioned by the project suggest that a positive impact is likely to have been achieved. Between 2005 and 2015, there was a notable decline in the percent of adult men using tobacco from 40% to 32%, and from 2.8% to 2.2% among women, and high percentages (40% of women and 23% of men) who used tobacco at the time of the 2015 NFHS reported having tried to stop during the past 12 months. Data on other NCD risk factors were collected for the first time in the 2015 NFHS, so no trend data are available. 28 The NCD screening and treatment programs achieved extensive state-wide population coverage: 77% of people over 30 years of age were screened for hypertension. Evaluations commissioned by the project found substantial improvements in diastolic and systolic blood pressure among patients who were regularly followed up. Diabetes screening covered 61% of the population over 30 years of age. Of women aged 30-60 years, 71% were screened for cervical cancer and 86% for breast cancer. This resulted in a dramatic increase in case detection and – despite some loss to follow-up – in treatment for these diseases. Although no data are available on treatment outcomes, it is safe to assume a significant gain in disability-adjusted life-years (DALYs - a combined measure of prevented deaths and illness). Tamil Nadu has one of the highest burdens of road-traffic fatalities and injuries in India; the large increase in the percentage of calls for ambulances after road traffic accidents that were served from 66% in 2009-2010 to 87% in 2014/2015 (exceeding the project target of 79.4%), as well as the investments in hospital capacity to provide emergency care, are very likely to have improved outcomes for traffic accident victims. There certainly are other factors that would have contributed to improved health outcomes in the absence of the project. During the project period, fertility levels continued to decline, literacy among women and men to improve, electricity reached most of the 11% of households in the state who had not had it in 2005, households using clean energy for cooking rose from 31% to 73% (removing a major risk factor for NCDs), and households using improved sanitation rose from only 22% to 52%. Much more sophisticated analytic work would be needed than is possible for this report to try to disentangle the impact of these and other relevant factors from the impact of the project activities on the improvements in health outcomes in Tamil Nadu over the course of the project. But the project’s contribution was clearly substantial, given the strong declines in the rates of neonatal and maternal mortality in project facilities, and the fact that by 2015, 67% of all institutional deliveries took place in public facilities (see next paragraph). Ultimate Performance Measure 2 - Patient Satisfaction. Rating: High (both phases) An important measure of the improved effectiveness of the public health system is the extent to which people seek care in public facilities, rather than from private sector providers. The NFHS-4 shows a large shift from 2005 to 2015 in the percent of institutional deliveries that took place in public rather than private facilities, from 48% to 67%, and an increase from 75% to 86% in children aged 12-23 months who received most of their vaccinations in public rather than private facilities. The NSSO records a small increase in the percent of hospitalizations that were in public hospitals among patients in the lowest two income quintiles, from 51% in 2004 to 54% in 2014, but a fall from 40% to 35% for all hospitalizations in the state – all accounted for by a fall in urban areas from 37% to 29%. Patient satisfaction surveys were done in 2006, 2010 and 2014. The project team thinks that the fairly high satisfaction scores at baseline (2006) reflect low expectations: Likert Scores on a 1-5 scale were: (a) overall satisfaction: 3.99 (in-patients), 3.95 (out-patients), (b) satisfaction with patient amenities: 3.74 (in-patients), and 3.72 (out-patients), and (c) satisfaction with cleanliness: 3.51 (in-patients), and 3.7 (out-patients). 29 The 2010 survey used a different methodology, complicating comparisons, but satisfaction was higher. It found: (a) 91% of out-patients and 92% of in-patients were satisfied with the facility, (b) 89% of out-patients and 91% of in-patients satisfied with cleanliness, (c) 96% of in-patients satisfied with the admission process, (d) 79% of out-patients and 70% of in-patients likely to return to the same hospital, 76% of out-patients and 82% of in-patients would recommend the facility to a friend or family. Waiting times were short: 80% of patients waited no more than 20 minutes to access any services at OPD; 60% of in-patients perceived waiting time at emergency registration and access to doctors in emergency as short; 71% of in-patients were assigned a bed immediately. Indicators for the quality of care were good: 82% of out-patients and 71% of in- patients said that doctors asked questions to understand their history, 85% of out-patients and 92% of in-patients said they had adequate time with the doctor, and 70%/85% were satisfied with their discharge summary and explanations. By the 2014 survey, when patient expectations were considerably higher, overall satisfaction scores were 3.92 (in-patient) and 3.87 (out-patient), 98% reported that the out-patient department and waiting area were clean and hygienic; 97% of in-patients said facilities such as labor and ward rooms were clean and hygienic; and 79% of patients were satisfied with the facility. Waiting times set high standards: 84% of patients were satisfied with the 4 minute wait for out- patient care and 75% of in-patients found 10 minutes of registration time acceptable. Satisfaction scores improved for infrastructure, communication and behavior of hospital staff, quality of treatment, cleanliness, crowding, the discharge process and outcome of treatment, and also for outpatient registration. The improved satisfaction with treatment quality and staff interactions are critical aspects that encourage patients to return to a facility. An end-line study among tribal groups who had been hospitalized under the bed-grant scheme found that almost all were satisfied with the facilities, 93% received medicines, 88% considered services to be of good quality. The project clearly made considerable and successful efforts to provide health care services that met the needs of the populations, with high standards for access (services within 30 minutes of everyone, free emergency transport, 24/7 availability of many services, ensuring that drugs and other medical supplies were always available), mobile services for tribal and remote populations that brought a doctor, nurse and medical technician and vehicle equipped with basic laboratory services to their doorsteps regularly every 7-14 days, and training providers in respectful care. Ultimate Performance Measure 3 - Financial Protection. Rating: Substantial (both phases) A key measure of health system performance is that care does not cause financial hardship. This is part of the definition of Universal Health Coverage, a goal fully embraced by the global health community and India (and most other nations). At the start of the project, Tamil Nadu, like other Indian states, had very high out-of-pocket (OOP) spending on health care, and a low share of all health expenditure financed by the government. Public expenditure on health in Tamil Nadu was less than US$3 per capita per year in 2005; health’s share of the state budget had fallen from 7.5% in the mid-1980s to only 4.6% in 2005. During the first three years of the project, per capita spending on health more than doubled from Rs. 227 to Rs. 472 (Figure 2 below, from AM 08/2015), all of which came from the state budget (central transfers for health were fairly 30 constant from 2006-07 to 2009-10, as indicated in the upper (blue) section in Figure 2). The GoTN absorbed all the recurrent costs by the end of the project, notably by regularizing nearly 3,000 staff nurses hired for the NCD programs and CEmONC units. The state budget share for health did not rise during the project, but robust economic growth increased the health budget substantially by an average of nearly 8% annually. Per capita Total Public Expenditure on Health Per capita State Expenditure on Health Source: NIPFP, 2012. In 2005, only 4% of households in Tamil Nadu had any member covered by health insurance or a health scheme; in 2015 this was over 64%. Although not funded by the project, the PMU was responsible for administering the state insurance scheme that achieved this. The project directly reduced OOP by providing free emergency transport services and free mortuary transport to over a million people. The bed-grant scheme provided free in-patient care for nearly 12,000 tribal patients between 2007/08 and 2013/14. Data are not fully comparable, but the average patient spending per hospitalization in public sector facilities across India of 6120 rupees (NSSO 2014) was many multiples of the amounts reported in the Tamil Nadu 2014 patient satisfaction survey: patients interviewed after receiving care at facilities reported having spent just under 200 rupees on average, and respondents to a household survey reported having spent less than 100 rupees for care (IPSOS, End-line evaluation - Patient Satisfaction Survey, March 2015). While scanty, the available data suggest a considerable improvement in financial protection in Tamil Nadu, which, combined with increased access to care especially among the poorest 40% and the tribal populations, and greatly improved quality of care, indicates substantially improved well-being for Tamil Nadu’s population as a result of the project. Having assessed the 3 “ultimate outcome” indicators of the performance of the Tamil Nadu heath system, we now look at the 3 “intermediate outcome” indicators: access, quality and efficiency. Intermediate Performance Measure 1 - Access and Utilization. Rating: Phase 1 – Substantial, Phase 2 - High Access to health care – especially the project focus of maternal and infant care services, was greatly improved, particularly among the poorest 40% of the population and tribal groups. Most 31 respondents to an end-line study among tribal groups said that the bed-grant scheme motivated more tribal community mothers to deliver in facilities, and reduced self-medication/visits to a traditional healer. Regular visits by well-equipped, well-staffed mobile vans provided door-step access to doctors and nurses. The project paid attention both to supply (which enables access) and to demand (which results in utilization). The first project phase provided emergency obstetric and neonatal care services within no more than an hour of travel time for every woman in the state; the second phase improved the access standard to no more than 30 minutes. Combined with IEC/BCC interventions to encourage appropriate care-seeking, and counsellors to guide and advise patients in using of services, the project impact on service utilization was high. The graph below shows increases in numbers of maternal admissions, complicated maternal and neonatal admissions, and deliveries, from 2006 to 2015. The fall in neonatal admissions is a highly desirable outcome, likely a result of improved quality of care at delivery and improved follow-up. The percent of births in facilities rose from 88% to 99%, with almost all births (99.3%) assisted by a trained attendant by the end of the project. There were surprising decreases in utilization of individual antenatal services (Annex 10 indicators 32-35) and immunization of children under 2 (indicators 50-55). The low incidence of vaccine-preventable illnesses in the state during the decade resulting from the earlier impressively complete coverage may have made immunization seem less important, and this also likely reflects some shift in emphasis in service provision. The percent of children who were taken to a health facility if they had diarrhea rose from 62% to 73%, and from 76% to 82% for children with symptoms of acute respiratory infection (NFHS). The roll-out state-wide of breast and cervical cancer, hypertension and diabetes screening and referral for treatment reached more than 65% of the targeted age-groups, and identified 3.65 million positive cases, most of which are unlikely to have been detected without the project. Overall, utilization of care increased markedly in TN. The NSSOs in 2004 and 2014 show an increase among people who reported any ailment in the past 15 days and received any medical care from 81% to 97%, and especially large increases for rural residents from 78% to 97%, and 32 for the lowest 40% from 76% to 98%. The increase for urban residents was from 87% to 98%, and for the upper 60% income groups from 84 to 97%. The negligible differentials between the lowers and highest income groups, and rural/urban residents are very unusual, and show the impressive success of the project in achieving its aim of improving services “particularly for poor, disadvantaged and tribal groups”. Although the percent of ST who received any medical care fell slightly from 96% to 93%, the percent who reported any ailment rose from only 0.8% to above 10%, much closer to the general population norm, indicating a dramatic change in recognition of need for care, and an overall very substantial increase in use of care. The SC population treatment rate rose from 71% in 2004 (well below other groups) to 97% in 2014, completely closing the gap in treatment utilization. Hospitalization rates were 34% higher in 2014 than in 2004; the 27% increase among the poorest 40% compared to a 14% increase among the richest 60% narrowed but did not eliminate the disparity that existed before the project (see Annex 9, table 4). Intermediate Performance Measure 2 - Quality. Rating: High (both phases) Most available indicators show large improvements in the quality of care. The percentage of mothers who had full antenatal care (at least four visits, took folic acid for at least 100 days while pregnant, and received a tetanus shot) rose from 27% to 45% for their most recent pregnancy. Another strong improvement in quality of care at the primary care level is that the percentage of children with diarrhea who received oral rehydration salts almost doubled from 32% to 62% between 2005 and 2015 (NFHS). The risk-adjusted maternal case fatality rate at CEmONCs fell from 19.6 in 2006-07 to 4.6 in 2014-15, and risk-adjusted neonatal mortality fell from 5.24 to 3.98, indicating greatly improved case management and outcomes despite the much higher proportion of high-risk maternal clients. Referral of mothers from CEmONC centers decreased from 15% of total maternal admissions in 2004-05 to 4% in 2014-15, indicating better case management and availability of comprehensive care. The increase in caesarian sections (C-sections) from 15% to 26% in public facilities, although well below the increase in private facilities from 33% to 51% (2005 to 2015, NFHS-4 and NFHS-3) was worrying. Experts consider a population rate above 10% unlikely to improve maternal or infant mortality; rates above 15% often indicate medically unnecessary C-sections and are not recommended.16 The project commissioned an independent review that found that the increase was a result of the increase in complicated pregnancies (from 46% to 64% between the DLHS3 and 4) and post-delivery complications (from 48% to 59%) being treated at Government health facilities, especially CEmONCs, and not the result of medically unnecessary elective C- sections. The increase in use of government facilities for complications of pregnancy and childbirth suggests increased confidence in their ability to provide this care, an indirect indicator of improved quality. The project provided vital inputs for achieving higher quality care: renovations at 2,330 facilities, ensuring availability of running water, electricity, sanitation and proper health waste 16 WHO Statement on caesarian section rates, April 2015. 33 disposal and infection control, additional staff, extensive training for staff, protocols, equipment, improved maintenance and repair of equipment, and supplies of drugs and other consumables. The project put in place rigorous quality assurance and quality improvement mechanisms: monthly reporting on 80 indicators by all EmONCs and on 20 indicators by all public hospitals used to assign quality grades that triggered immediate remedial action to address substandard grades, in concert with Quality Circles of Excellence in 267 secondary hospitals. The accreditation of 12 public hospitals by the National Accreditation Board for Hospitals, with another 46 preparing for accreditation, is a major and path-breaking project achievement – these were the first public hospitals in the country to undergo the rigorous process. Extensive independent evaluations of all pilots and many project activities by universities and the National Institute of Epidemiology to identify ways to improve programs were an integral part of the project’s commitment to continuous improvements in the quality and effectiveness of the health system. The summary of the stakeholders meeting held at the end of the project (see Annex 7) contains recommendations for improvements, showing that this mindset persisted. Intermediate Performance Measure 3 - Efficiency of the health system. Rating: High (both phases) There are two aspects to efficiency: allocative efficiency, or spending money on the “right” things, and technical efficiency, the rate at which inputs are transformed into outputs or outcomes, or “doing things the right way”. The project explicitly aimed to improve allocative efficiency in the health system by strengthening services at the primary and secondary levels. Over the life of the project, Tamil Nadu’s overall health budget became more cost-effective by focusing more on primary care, increasing its budget share relative to secondary and tertiary care by more than 15 percentage points, while the tertiary share fell by more than 10 percentage points. The project funded services to prevent and treat conditions that account for well over half the disease burden in Tamil Nadu, another indicator of the project’s likely positive impact on allocative efficiency of the health system. The available global literature on cost-effectiveness, summarized in the definitive Disease Control Priorities17 project (DCP), suggests that the innovative state-wide programs to screen for four major NCDs/risk factors and provide treatment and life-style advice to reduce NCD disease risk, are likely to be cost-effective. For example, cardiovascular disease management (e.g., screening and treatment for ischemic heart disease) is among the more cost-effective interventions available, costing approximately $1000 per disability-adjusted life year (DALY) averted. Drug treatments for acute episodes such as heart attack and congestive heart failure are even more cost-effective. Treatment costs for the more treatable cancers covered by the project (breast and cervical cancer) have ratios of $1,300-$6,200 per year of life saved, compared to $53,000 - $163,000 for less treatable cancers. Thus, the project appropriately targeted the more cost-effective cancers and other NCD interventions. In addition, “improved quality of comprehensive emergency obstetric care” and “neonatal packages” are both identified as 17 Laxminarayan, R., et al. (2006). “Advancement of global health: Key messages from the Disease Control Priorities project”. Lancet 367: 1193-1208. 34 neglected low-cost opportunities in the Disease Control Priorities project, with very favorable costs per DALY averted. The quality improvements and increased utilization of the health system that the project helped to achieve would have increased the technical efficiency of the system. The better the quality of health care, the greater the impact on health outcomes, the lower the cost per unit of health gained, and the greater the system “outputs” (of health, patient satisfaction and financial protection) for a given level of inputs. The total project cost was small compared to the total budget of the health system in TN, rising from 2.6% in the first year to a maximum of 7.6% in 2010, and steadily decreasing to only 0.1% in the final year. This is a modest cost for the measured improvements in quality, access and utilization. Total health spending in Tamil Nadu in 2015 was about US$17 per person, with relatively good health outcomes compared to other states and countries with similar levels of spending. The HMIS system also improved the health system’s efficiency by saving time for patients and providers at each visit, and enabling better continuity and quality of care by making patients’ health records available to providers at every point of contact, and by automatically prompting providers with relevant clinical protocols and treatment options. The immediate availability of system-wide data, presented in dashboards that are designed to support quality improvements and data-supported decisions offer potential for continued improvements in care and system efficiency well beyond the project life. Can the outcomes and impact be attributed to the activities supported by the project, or might other factors be responsible? An assessment of what might have happened without the project is difficult. The GoTN was strongly committed to improving the health system and health outcomes, and it seems likely that some actions would have been taken without the project. However, it is clear that the Bank’s strong technical advice throughout the project, and the accountability that a Bank-funded project brings – especially one that is supervised with such regularity, energy, proximity and attention to detail, provided important support for the project’s accomplishments. In fact, the project was developed at a time when the tensions around the DIR deterred many states from wanting to work with the Bank, but the GoTN was clear that they wanted and needed the expertise and advice of the Bank in seeking to strengthen their health system. A possible counterfactual might be the extent to which Tamil Nadu’s health system improved compared to health systems in other states or countries, looking at “difference in differences” over the project time period. But there are (at least) two major difficulties: the choice of appropriate comparator states is not obvious, and it would be extremely challenging to control sufficiently for confounding factors to be able to draw useful conclusions. There are other factors that likely contributed to the observed improvements in health outcomes and utilization of health care: gains in literacy and per capita incomes, and a (small) rise in the age of marriage and child-bearing and continued drop in fertility (see Annex 10), anti-tobacco activities in response to the FCTC, the activities of the Transport Project to make roads safer and reduce road traffic accidents. However, none of these external factors would have improved the other measures of the effectiveness of the health system discussed above. The well-chosen and 35 focused investments by the project clearly resulted in increased quality, access and efficiency of TN’s health care system. There is compelling evidence that the project’s achievement of its development objective was at least Substantial for the first phase, and High overall and for the second phase, looking both at key project indicators and other data. 3.3. Efficiency (rating: Phase 1 – Substantial; Phase 2 - High) The comments above on the project’s impact on the efficiency of the health system in TN and the discussion of the relevance of the project’s objectives and design have already made the case for the project’s allocative efficiency (spending on the right things). The project was exemplary in deciding what to fund. The choices made in designing the project were informed by rigorous data gathering and problem analysis for every component and sub-component. The best available technical advice was solicited – including from intended beneficiaries. For example, the QALP team noted that: “The Tribal Plan was developed based on consultations with various NGOs working on tribal health issues, tribals and their ‘sangams’, and field visits to tribal areas, and also various government departments including the Health, Tribal Welfare and Forest Departments. Both primary and secondary data were collected and analyzed to provide the basis for the chosen interventions. The TDP includes a broad range of interventions to address the issues identified and provide quality health care to tribal populations in Tamil Nadu. The TDP emphasizes the integration and strengthening of existing health interventions like the RCH project, the RNTCP, the NLEP and other on-going government welfare schemes.” Well-justified decisions were taken to drop activities that duplicated other efforts (e.g. the road safety component), were judged unlikely to contribute to achieving the project objectives (the proposed pilot community insurance scheme), or for which the policy environment and low enforcement capacity made successful implementation unlikely (regulation of the private sector). The project was an “early adopter” – even a trail-blazer – in its decision to pilot and roll-out state-wide NCD interventions. The estimated costs of treatment and lost productivity caused by cancers, heart disease and other NCDs suggests high rates of return on well-chosen NCD program investments.18 The screening and treatment protocols, technical decisions and reporting formats developed under the project have had a very strong influence on national policies and in other states in India. In addition to being strongly justified by allocative efficiency, the costs relative to outputs and outcomes indicate that the NCD interventions have strong technical efficiency as well. For a total project cost of just US$19m, the NCD program provided hypertension screening for 29 million people, diabetes screening for 23 million people, breast cancer screening for over 12 million women, and cervical cancer screening for 10 million women. This detected approximately 3 million new hypertension cases, 1 million people with diabetes, 350,000 cervical cancer and 153,000 breast cancer cases. It would require only a very 18 Bloom et al, 2014, “Economics of NCDs in India: The costs and returns on investment of interventions to promote healthy living and prevent, treat and manage NCDs”, (World Economic Forum, Harvard School of Public Health), estimates that between 2012 and 2030, India will incur very high costs of $2.17 trillion from cardio-vascular disease, $0.25 trillion from cancers, and $0.15 trillion from diabetes. 36 tiny fraction of these 78.5 million people to gain even one additional year of life as a result of screening and treatment for the benefits to far exceed the costs (see Annex 3 Economic and Financial Analysis for a more detailed estimate). The study by Bloom et al cited in footnote 18 estimated the return on investment of the project NCD program at well above 15%. An economic analysis of the NCD pilots provided important justification for the decision to scale the programs up and expand their coverage by adding screening for diabetes and hypertension, and low-cost cervical cancer screening in addition to breast cancer screening. The ISRs and Aide-Memoires show close attention to efficiency, discussing ways to reduce costs and improve efficiency and effectiveness. Monitoring data and rigorous evaluations were used to make improvements. For example, at the suggestion of the WB, the project analyzed and compared the costs per patient of the three NGOs contracted to provide care under the Bed Grant scheme. Significant variation was found across the three service providers, which triggered negotiations to rationalize the reimbursed costs and align them with the reimbursement rates used by the new state-wide health insurance scheme. The project monitored the number of patients served by each of the Tribal patient counsellors, assessed who was using the services (mostly poor and often illiterate people) and that satisfaction with the services was high, to ensure that the costs were well-justified. Improvements were constantly made – for example, a high risk antenatal screening program was tested in two districts in 2014, and scaled up to 20 more districts when good outcomes were achieved. In one year (June 2014- June 2015), 89,000 pregnant women were screened in 1674 “camps”, 45,000 (about half) were identified as at risk for complications, and 21,700 were referred early to CEmONCs for better management of possible complications. There were 25 fewer maternal deaths over the year compared to the previous year, a very substantial reduction. Changes in the way the emergency transport services were managed and IEC to increase their use reduced the average operating costs per trip from Rs 2,551 to Rs 1,096 from 2008/09 to 2014/15. The project also increased implementation efficiency by working through government partners (departments of education and labor), schools and workplaces, and NGOs including the TN Women’s Development Corporation and Gandhigram Rural Institute. Disappointing initial results from contracting NGOs to operate emergency ambulance services led the project to seek and adopt the much more efficient arrangement with the EMRI. The investment in the new HMIS was specifically justified by its potential to enable more efficient services, and will pay for itself many times over if it is instrumental in achieving even small (e.g., 1%) system efficiencies. The project extensions provided the time needed to fully complete (and expand the intended scope) of the HMIS development. Although disbursements were much slower than expected for most of the first phase of the project, disbursements accelerated from 2009, and from 2010 they tracked the revised expected disbursement profile closely (see data sheet section I). Project efficiency is rated Substantial for the first phase and High for the second phase. 3.4. Justification of Overall Outcome Rating A summary table of the ratings for each phase of the project (before and after the 2010 AF) is provided below. Consistent with the rating guidelines, the overall outcome of phase one is 37 satisfactory (high relevance, and substantial achievement of PDO and efficiency). The overall outcome for the second phase is highly satisfactory, given high ratings for relevance, efficiency and achievement of objectives. The project disbursed US$88.59 million – 42.65% of the total $210.05 million prior to the Additional Financing, and $121.5 million – 57.35% in the second phase. The weighted overall outcome rating is assessed as Highly Satisfactory, since the weighted rating score is 5.57, and should be rounded up rather than down (see second table) Phase 1 (2004-2010) Phase 2 (2010-2015) Relevance High High Objective H H Design H H Implementation S H PDO Achievement (Efficacy) Substantial High Efficiency Substantial High Overall Outcome Satisfactory (5) Highly Satisfactory (6) Phase 1 Phase 2 Overall 1 Rating S HS 2 Rating value 5 6 3 Total disbursed (US$ million) 88.59 121.50 4 (% = total disbursed/final disbursed amount) 42.65% 57.35% 5 Weigh value (2 x 4) =5x42.65 =6x57.35 6 Final Outcome Rating 2.13 3.44 5.57=HS Note: HU (1); U (2); MU (3); MS (4); S (5); HS (6) 3.5. Overarching Themes, Other Outcomes and Impacts (a) Poverty, Gender Aspects, and Social Development The project central focus on poverty and vulnerability in improving access to, utilization of, and improved quality and efficacy of public health services particularly by poor, disadvantaged and tribal groups has been well covered already. Loss of productivity and health care costs are a key cause of poverty and worsened impoverishment – the improved effectiveness of the health system and better access to free care would have reduced both. The Tribal Development Plan and interventions targeted to tribal groups were a core part of the project. In addition to specific interventions to strengthen service delivery (ensuring adequate medical staff in PHCs, Health Service Centers and government hospitals in tribal areas, IEC activities and counsellors in PHCs and general hospitals in tribal areas to encourage use of health services, increased services access through PPPs with NGOs to provide regular mobile outreach services, the bed grant scheme to provide free hospitalization in selected private hospitals – as well as free care provided in TN’s public hospitals), the project also addressed a specific health need of the tribal population – sickle cell anemia, and arranged for pregnant tribal women to stay at PHCs prior to delivery to encourage institutional deliveries. The project also gave priority to the poorest regions and communities with the worst health outcomes in selecting facilities to be the first to get EmONCs. Criteria for selecting secondary hospitals for upgrading included health indicators (IMR and MMR), and those that served populations below the poverty line, and in tribal areas. 38 The benefits to the poorest 40% population and ST/SC groups surpassed project targets, with 1.84 million tribal people living in remote rural areas treated through mobile outreach health services, 11,889 people receiving free hospitalization under the bed grant scheme, and 1.936 million patients using counseling services in tribal PHCs and hospitals. The 30 four-wheel drive vehicles to transport patients over difficult terrain increased uptake of services by tribal groups from 16,000 in 2013-2014 to nearly 27,000 in 2014-2015. IEC activities included use of traditional media such as street plays that brought messages into poor communities, in addition to use of mass media and print materials that have much lower penetration in poor communities. Interventions to reduce maternal mortality, cervical and breast cancer obviously benefit women primarily, and women also benefitted from all other project interventions. In addition, project preparation included development of a Gender Plan to ensure that all components were sensitive to the specific needs, constraints, and situation of women. (b) Institutional Change/Strengthening The project emphasis on institutional development in all activities has been noted above – notably extensive training of health care staff at all levels in clinical, managerial and process skills; and setting up the new SHDRC, Strategic Planning Cell, PPP Unit, and system for inventory control, maintenance and repair of all medical equipment. The project was instrumental in overcoming GoTN wariness of working with the private sector, demonstrating the benefits and efficiencies that could be gained through careful contracting. As experience and skill were gained in contracting, the project began to move from fixed-cost to performance-based contracts to ensure better value for money and incentivize contracted partners. The staff and functions of the various cells in the PMU have been absorbed into the relevant Directorates of the DoHFW without any loss of the expertise and capacity developed under the project, and project activities smoothly transferred. The additional nurses contracted under the project have also been added to the state regular payroll. Accreditation of 12 hospitals helped build hands-on capacity in the Directorates of DoHFW on all quality dimensions of health service delivery, and has enabled the state to begin the process towards accreditation of another 46 hospitals. (c) Other Unintended Outcomes and Impacts (positive or negative) Although the demonstration effect of the project was intended, the extent of its influence went beyond expectations. In addition to the other state governments, the project has also been visited by USAID, JICA and the Bill and Melinda Gates Foundation (among others), and has had a very strong influence on national level policies. The extent to which the processes and evaluations have been documented and disseminated has also been beyond expectations, covering policy, administration, financial, operational and management aspects. Another unplanned benefit was the introduction of a unique patient identification number (PIN) that is able to be integrated with the identifier provided by the Gol Aadhaar program, and also the extension of the HMIS to include medical colleges and the Tamil Nadu Dr MGR Medical University. Third, the project provided complete administration and management support for rolling out the state-wide Chief Minister’s Health Insurance Scheme after 2008. This scheme had not been envisaged when the project was being developed. 39 3.6. Summary of Findings of Beneficiary Survey and/or Stakeholders Workshops The GoTN held a workshop of stakeholders in Chennai on August 28 and 29, 2015 to solicit stakeholders’ views on the project’s performance, capture and disseminate experiences, discuss innovative interventions under the project, discuss challenges and recommend actions to help address them. Stakeholders included officials from the GoI, GoTN, Governments of Kerela and Uttar Pradesh, former and current Project Directors, NIE, Indian Council of Medical Research, academic institutes, Institute of Public Health, officials of various DoHFW Directorates, project hospitals, Medical Colleges, consultants, NGOs, PPP providers, civil society, and project staff, as well as eight WB staff. Topics covered project interventions and support to maternal and child health, HMIS, health care quality, health care waste management, surveys and studies, PPPs, the Tamil Nadu Chief Minister’s Comprehensive Health Insurance Scheme, tribal health, and NCDs. Findings are presented in Annex 6. In summary, there was consensus on the strong achievements made under the project in reducing maternal and infant mortality, improving tribal health, implementing NCD interventions, strengthening monitoring and evaluation including the HMIS, improving quality of care and HCWM, and strengthening secondary care hospitals, and making progress on adequate staffing of public health facilities across the state. There was a shared conviction that the project has helped improve the functioning of Tamil Nadu’s health system, and brought together and helped develop a talented group of officials who are now working in the DoHFW, and will be able to sustain and continue the project impact. A continuing theme throughout the workshop was the commitment to continue all activities implemented under the project. Appreciation was expressed for the World Bank’s expertise and rigorous implementation support throughout the project, complementing Bank financing to help put in place a robust health sector system and to scale successful ideas. 4. Assessment of Risk to Development Outcome Rating: Negligible risk that the PDO will not be maintained, given strong commitment at all levels in Tamil Nadu to build on the successes and lessons learned under the project, the strong M&E system, and the availability of financing from the NHM and state budget to continue activities. The justifications for the negligible risk assessment are as follows: (i) continued strong commitment of the GoTN to and its full ownership of the project development objective and activities including public-private partnership programs, outsourcing contracts and other innovations started under the project. (ii) Project activities are fully mainstreamed and integrated into the work programs and budget of Tamil Nadu’s DoHFW. All programs and activities have been handed over to the Directorates of DOHFW for continued implementation as regular departmental activities. All contract staff and other new staff sanctioned under the project have been transferred to the Directorates along with the programs. GoTN and NHM are fully funding these activities. The State Health Data Resource Centre, 108 Emergency Ambulance Services, and Free Mortuary Van Services have been retained and funded by GoTN through the TNHS Society. The GoTN is acting fully on its commitment to provide state budgetary funding for any activities not financed by the NHM. (iii) The project administrative structure, formalized as the TNHS Society, together with the DoHFW, has built a strong consistent track record in implementation performance. (iv) There has been consistent demonstrated strong capacity in 40 monitoring and evaluating project programs, and a culture of evidence-based decision making. Data are being collected routinely and scrutinized to continue this. (v) TN is justly proud of being the first state or among the first in India to start and implement innovative activities such as the HMIS, a large scale NCD program, tapping into the resources of the private sector and NGOs to help deliver carefully chosen priority health services through PPPs and out-sourcing, focusing on quality of services simultaneously with infrastructure investments, and developing and applying more realistic staffing needed to deliver defined services to clear standards, starting with the CEmONCs. The strong sense of achievement and commitment bode well for the future sustainability of the programs put in place under the project. It should however be noted that full success of (i) the NCD interventions will depend heavily on adequate follow up for confirmatory diagnosis and appropriate treatment of patients who test positive during screening, and (ii) the CEmONCs in delivering effective 24x7 maternal and neo- natal health services will depend on continued adequate staffing of specialists and staff nurses. 5. Assessment of Bank and Borrower Performance 5.1. Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory. As discussed earlier (Sections 2.1 and 3.1), project objectives and design were strongly relevant and remained relevant, and were fully aligned with government and Bank priorities. The project benefitted from extensive analytic work including an organizational review of Tamil Nadu’s DoHFW and a quality enhance review in 2002 of the Bank’s experience in India since 1995 with health systems development projects to inform future directions, and from workshops in India to disseminate and discuss the review findings. Lessons were incorporated, as noted in the PAD, p.8-9, for example on the need for special efforts to reach the poorest and to measure their access to and utilization of health services; to proactively explore possibilities for experimenting with news ways to improve the public health sector, including opportunities to partner with private sector providers in underserved areas; and to pay attention to strategic planning and management in order to strengthen hospital management. The Bank team engaged fully with important relevant stakeholders in developing the project. Implementation arrangements were appropriate, and, apart from some baseline surveys, the project was fully ready for implementation by effectiveness, including a detailed procurement plan for the first 18 months of planned activities. Risks were sensibly identified and well mitigated through project design and preparation. Some of the delays in awarding contracts in the first year of the project indicate that procurement processing capacity of the state’s PWD for civil works and of TNMSC for goods was overestimated; but this was appropriately and quickly addressed. Although there were some weaknesses in selection of indicators, other aspects of M&E design were clear and comprehensive, notably the development of a computerized HMIS system to replace manual reporting and make real-time rich data available for decision-making, and independent evaluations of numerous project activities. Environmental and social safeguards aspects were adequately covered including a sound HCWM plan, a Tribal Development Plan and a social assessment, which were fully implemented. (b) Quality of Supervision 41 Rating: Highly Satisfactory The project was supervised by a strong team with one task team leader from the start of project preparation in 2002 until after approval of the AF in 2010, and two task leaders in the remaining five years of the project. After the DIR, project supervision budgets were supposed to increase. The project did not in fact receive the 30% increase it was supposed to get, but still managed to provide highly satisfactory supervision. An important factor was that the project team was mostly based in Delhi, comprising all necessary skills including an IT specialist (for the HMIS component). The relationship with the PMU and DoHFW was professional and strong. Supervision was systematic, detailed, and every six months, including a carefully prepared and rigorous MTR; field visits to hospitals were frequent. Aide Memoires were comprehensive, detailed, issue- and action oriented; they included the status of results achievements, and benchmarks. Project ratings were realistic. Potential and emerging problems were recognized early, discussed candidly, and forcefully brought to the PMU or higher state authority’s attention as needed. The Bank team in Delhi held weekly audio meetings with Chennai to follow-up issues and support the PMU in resolving bottlenecks, and was diligent in monitoring fiduciary and safeguard aspects including implementation of the Tribal Development Plan and the Health Care Waste Management Plan. The Bank team and PMU jointly paid close attention to the project’s development effectiveness, and the Bank offered technical advice as needed. The Sector Manager and Country Management Unit paid close attention to the project. Satisfactory project implementation (and highly satisfactory for several project activities) particularly since 2007 justified Bank approval of AF in 2010 to replace the $20 million that had been released for dealing with the Tsunami aftermath, plus almost $100 million in new financing to implement the NCD activities across the state and extend well-performing components. The extensions of the closing date were well justified, and enabled full disbursement and completion and expansion of planned activities. (c) Justification of Rating for Overall Bank Performance Rating: Highly Satisfactory. With a rating of satisfactory for preparation and highly satisfactory for supervision, overall Bank performance is rated as highly satisfactory in line with the overall outcome rating of Highly Satisfactory. 5.2. Borrower Performance (a) Government Performance Rating: Highly Satisfactory. GOI supported the Government of Tamil Nadu at all stages of the project preparation and implementation including its endorsement of GoTN’s request for the AF. GoTN’s ownership and commitment to the overall project objective was consistently strong, reflected in its decisions to establish CEmONCs able to provide 24x7 maternal and neonatal health services, try using PPPs to deliver health services to low-income communities including SC/ST populations in remote and tribal areas, to pilot test innovative approaches to NCDs on a reasonably large scale and subject them to rigorous evaluation before scaling up, and establish the first fully computerized HMIS in India. The project was fully integrated into Government structures at all levels. To ensure sustainability, the GoTN began absorbing project activities (well before project closing) into the work and budget of the NHM and the DoHFW (e.g., mobile outreach, counselling services, sickle cell anemia, bed grant scheme, heath waste management, HMIS, emergency ambulance transport), with only essential inputs (technical support and 42 monitoring and evaluation) continuing to be funded by the project. As needed, the government approved recruitment of doctors and staff nurses on contract for CEmONCs and NCD interventions to ensure adequate staffing to deliver services, and subsequently absorbed them into the civil service cadre. The Government is continuing all project activities with financing from the NHM and state budget. The GoTN showed unwavering strong support and commitment to the project during thirteen years of preparation, implementation and transition, and managed to push the project to do much more than originally envisaged and complete all activities. (b) Implementing Agency or Agencies Performance Rating: Satisfactory. The PMU had overall responsibility for managing the project with support from PWD for civil works and TNMSC for procurement of equipment and maintenance. GoTN appointed a Senior Officer from the IAS as Project Director of TNHSP. The PMU team was highly experienced and successfully managed implementation of the many project activities in different technical areas including new areas of NCD interventions and HMIS -- large undertakings in any context. It maintained a strong focus on capacity building and skill enhancement and built strong professional training capacity for the public health sector. The PMU worked diligently with the Bank, PWD and TNMSC staff to resolve the issues that delayed procurement and implementation during the first 18 months of the project. It proactively adjusted interventions during implementation to address bottlenecks or improve efficiency or impact. The PMU took a strong lead on actions to improve quality of care in hospitals, championed the system of grading, and regularly followed up agreed actions with hospitals graded C and D. It was proactive in preparing proposals for the State Empowered Committee chaired by the Chief Secretary to obtain Government Orders to proceed with project activities when necessary (such as requests for exemptions to procurement freezes prior to elections). The PMU consulted fully and regularly with key stakeholders and worked closely with the DoHFW Directorates. On financial management, audit reports and IUFRs were submitted regularly but with some small delays. Disbursements were slow in 2005-2006, but picked up pace from later in 2006; one important action was establishing a TNHSP society to ensure a smooth flow of funds for all activities except civil works and major equipment and goods procurements. Minor shortcomings (slow disbursements in the first two years), delayed procurement actions in PWD and TNHSP and small delays in the submission of audit reports and IUFRs) did not impact the timely and smooth implementation of the project activities. (c) Justification of Rating for Overall Borrower Performance Rating: Highly Satisfactory, combining the ratings of highly satisfactory for government performance and satisfactory rating for implementing agency performance given the overall outcome rating of highly satisfactory. 6. Lessons Learned Key Lessons  Careful strategies, including skillful sequencing, can help deal with the complexities of health system strengthening. Strengthening a health system is a complex undertaking, and requires appropriate balancing between physical investments and reforms, careful phasing 43 and sequencing, ensuring well-trained personnel, supplies, governance, and a long-term horizon for institutional capacity building. A health system is as strong as its weakest link, so there is need to consider the whole chain of care, and identify bottlenecks on which to focus. For example, the potential health gain of an excellent fast-responding emergency transport service is lost if the quality and capacity for care upon arrival at the hospitals is not at least as good (and hopefully better) than the care that the ambulance and its crew are equipped to provide. If patients are being successfully stabilized in the ambulance, but then die waiting for care or for want of capacity to provide appropriate care at the hospital, then the investment in the ambulance service is wasted (at last for that patient). However, the temptation to try and implement everything at once should be resisted. Phasing enables data to be analyzed and to inform decisions, and time to learn from and incorporate lessons from pilots and their evaluations. Skillful sequencing of physical upgrades that are relatively easy to implement (civil works to ensure running water and good sanitation, and fully equipping facilities) with the incentive of further upgrades if “soft” investments19 are successfully implemented, can add strong motivation for reforms that require behavior change. A highly supportive environment, incentives, and shared belief in their purpose are also needed.  Thoughtful, nimble adjustments are needed throughout implementation, learning along the way and resolving issues as they arise. It is not possible to anticipate all details when designing new programs, and very important to put in place good mechanisms and shared commitment at all levels for making continuous improvements. Regularly measuring and monitoring performance of CEmONCs and of tribal health interventions including PPPs and then actively using the information to make adjustments in the interventions and PPP contracts during implementation were essential to the project’s success in reducing maternal and neonatal mortality, improving tribal health, and facilitating the use of hospitals by the poor and disadvantaged groups. The experience of designing and implementing the HMIS pilot is another good example. The complexity and time required were underestimated, and the work would not have been able to be completed within only five years. Working productively with the technical agency hired to design and help implement the system required intensive and frequent interactions with the government and users, and a significant amount of “hand holding” and mutual trouble-shooting. It took time to establish good reporting formats that would be easy to use. Careful behavior change support was needed for hospital staff at all levels to transition to the new ways of reporting, including to allay the perceived threats and insecurity, and to overcome the belief that both paper and electronic reporting were needed, by demonstrating the reliability and robustness of the electronic system before gradually phasing out paper reporting.  The design of the NCD component offers lessons in successful use of well-evaluated pilots to make difficult choices and set priorities. The GoTN initially wanted to address all NCDs and provide a wide range of curative services. Careful and evidence-based discussions on what was technically possible, especially within the staff and other constraints of the system, as well as what was most cost-effective, helped reach agreement on a limited initial scope for TN’s NCD program. The Bank brought in international expertise from CDC and India’s leading national expert (Dr Srinath Reddy) who is also a highly respected global 19 “Soft” investments refer to new procedures and processes, and other reforms that require behavioral change, as opposed to “hardware” investments in infrastructure and equipment. 44 expert, to work with the TN team. The result was well-focused pilots to test the feasibility and impact of screening and interventions to manage hypertension and detect and treat cervical cancer, with a strong focus on prevention, early detection and disease management. The pilot protocols were developed through extensive consensus discussions with national experts. A rigorous evaluation was built into the pilot design, with input from international and national experts. The data collected throughout the pilot, and impact and process evaluations were all carefully scrutinized before deciding whether, how, and how fast to proceed with scaling up. Valuable lessons were learned from the pilots and incorporated into the scale-up design. One of the most difficult challenges was effective follow up of people who tested positive in screening to ensure they receive appropriate treatment and education in life-style changes to help manage and prevent further complications. Careful additional assessment was needed to understand the systemic and behavioral reasons for high loss-to- follow up, and how best to address them. This is absolutely central to the success of a screening program. Additional Lessons  Well-designed partnerships with the private sector/NGOs through PPPs for delivering health care services and outsourcing carefully selected services such a diagnostic tests, cleaning and laundry, can improve efficiency and services, and make health care more accessible for hard-to-reach populations. There is usually more than one way to deliver services, and new potential partners can be attracted to service areas where they have not operated before. When initial contracts with local NGOs to operate emergency medical transport did not yield the desired results, the GoTN found a very different approach in partnering with the EMRI that was operating a successful ambulance service in another state. EMRI proved willing to partner with the project and expand its operations into TN. Careful monitoring and willingness to acknowledge that the initial arrangement (with NGOs) was not working well, and to try a different solution, were important. Another lesson is that contract terms and approaches may usefully be changed over time, as the contractual parties become more familiar with each other, and with the contractual process, and the activities. For example, the initial contract with EMRI (and other partners) were lump-sum contracts, but the project is slowly embracing performance-based contracts that increase the incentive of the contracted partner to improve efficiency and utilization.  Infection control and health waste management are better addressed in a systematic, sector-wide, state-wide way rather than a smaller-scale project-specific approach. The approach adopted under the project was efficient, and enabled the Ministry to rely on the municipal authorities for regular quality assurance of the private sector disposal facilities. It ensured that the whole health sector in Tamil Nadu would benefit.  IEC and BCC activities were a well-integrated part of the design of programs and components, and carefully considered both supply of services, and demand. The project made skillful use of Information, Education and Communication (IEC) and Behavior Change Communication (BCC) activities, especially to boost demand and use of services that were being set up and expanded, and to encourage and enable expanded use of services by SC/TC members. Counselling sought to reduce loss to follow-up in NCD screening and treatment programs. The project made thoughtful and strategic use of IEC and BCC. 45  Resistance to being evaluated can be overcome by demonstrated usefulness of good evaluations. There are many reasons why implementers may not welcome evaluations. Delay in selecting and contracting the consultants for the NCD evaluation was partly the result of resistance from a skeptical key official. It took patience and persistence to get the evaluation underway. During project implementation, the usefulness and value of independent evaluations was clearly demonstrated, and came to be highly valued for being able to answer important questions about the impact of programs and activities, and as the basis for well-informed decisions. The project funded numerous evaluations of specific programs and activities (for example, to discover whether the increased rate of C-sections was medically warranted or not).  A “perfect storm” of mutually reinforcing factors all contributed to the project’s outstanding success. The project was very well designed, implementation was flexible and evidence-driven, both client and Bank teams had continuity, and high sustained commitment and competence. The project also had a strong champion in the Department of Health with a deep commitment to its goals. In addition to health being treated as “above politics” and being given consistent priority by successive ruling parties, Tamil Nadu’s civil service has a reputation for seriousness of purpose and “getting the job done well”. Close, collegial, supportive supervision helped identify early problems and bottlenecks, and to work out solutions. In the initial period when the project was not doing well, weekly phone-calls between the Bank and TN team helped resolve issues. The close supervision also enabled the Bank team to effectively and diplomatically be a “broker” when needed between the TN team and consultants or external evaluators. The stability in the Bank team was also a clear signal that the Bank was fully committed to the best interests of the state. Frequent changes in Bank teams give a negative signal to the client, undermine trust that is achieved over time, and can generate resistance. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies (b) Cofinanciers (c) Other partners and stakeholders (e.g. NGOs/private sector/civil society) 46 Annex 1. Project Costs and Financing (a.) Project Cost by Component (in USD Million equivalent) Appraisal Estimate Total Final Actual Actual as Components – Original Estimate20 (USD Percentage of (USD millions) (USD millions) millions) Appraised (a) (b) (c) (c/b) x100 Increasing Access to and 43.79 81.60 82.97 101.68% Utilization of Services Developing Effective Models to 5.65 26.72 26.68 99.85% Combat Non-Communicable Diseases and Accidents (Revised to “NCD Prevention and Control” at Additional Financing) Building Capacity for Oversight 25.61 55.41 60.86 109.84% and Management of Health System Maximizing Efficiency of the 50.90 73.00 73.28 100.38% Public Sector to Deliver Essential Services Total Baseline Cost 125.95 236.73 243.79 Contingencies 5.64 5.64 Total Project Cost 131.59 242.37 243.79 102.98% Total Financing Required 131.59 242.37 243.79 (b.)Financing Appraisal Estimate Original minus $21 cancelled Actual as Appraisal Actual due to Tsunami plus Percentage of Source of Funds Estimate (USD millions) Additional Financing Appraised (USD millions) (c) (USD millions) (c/b) x100 (a) (b) Borrower 20.76 33.84 33.70 99.58% International Development 110.83 208.53 210.09 100.75% Association (IDA) Total Financing 131.59 242.37 243.79 102.98% Disbursements: (i) Up to April 29, 2010 (date of Approval of Additional Financing) = US$ 88.59 million (42.17% of the total disbursed) (ii) From April 30, 2010 to August 7, 2014 (From AFs approval to the date of change in the Results Framework/ Restructuring): US$ 98.18 million (46.73%) (iii) From August 8, 2014 to final disbursements: US$ 23.32 million (11.1%) 20 To simplify presentation, this column shows the final estimated cost: the appraisal estimate, less the $21 million ($20 million at 2005 exchange rate) cancelled on June 30, 2005 to use to help finance the Emergency Tsunami Reconstruction Project in 2005, plus the Additional Financing approved on April 29, 2010 47 Annex 2. Outputs by Component TNHSP completed almost all planned activities and exceeded targets for many. This Annex summarizes the main outputs delivered, compared to what was planned under each component. (Final actual disbursements for each component are noted in the component heading.) Component 1: Increasing Access to and Utilization of Services (USD 82.97 millions) Planned Accomplished Sub-component 1: Reducing Maternal and Neonatal Mortality. Establish at least two  Established and strengthened 75 CEmONCs (including 20 CEmONCs in each of CEmONCs in Medical Colleges) and 50 EmONCs in the state to the 32 districts, able to provide definitive treatment and improved quality of care, 24x7 treat obstetric and for all obstetric and neonatal emergencies. In addition, neonatal emergencies, strengthened 8 identified medical college CEmONCs with including C sections. extensive civil works and inputs. First ones to be in  Instituted a mechanism for regular recertification of disadvantaged districts. CEmONCs using established criteria to ensure adequacy of resources at the facility and quality of care for provision of 24 hour emergency obstetric and new born care services. Four rounds of re-certification of CEmONCs were taken. Equip CEmONCs with  Steadily staffed CEmONCs with doctors and specialists per treatment protocols, norms (2 OBGYN, 2 pediatricians, 1 anesthetist)-75 CEmONCs trained staff, had 4 OBGYN, 2 pediatricians and 2 anesthetists, and 50 equipment, supplies EmONCs had 2 OBGYN, 2 pediatricians and 1 anesthetist and drugs needed. (9/30/2013) against the target of 80 CEmONCs.  562 staff nurses were recruited for CEmONCs and their salaries paid through project for the first two years, thereafter, the GoTN absorbed the staff nurses into existing cadres and financed their salaries from budget.  1,419 doctors and 3,342 nursing staff were trained in skills for operationalization of CEmONCs during Oct 2010 – March 2014 (target was 1,068 doctors and 1,334 nurses by 9/30/2013).  (from 2010: 37,468 doctors and nurses trained in CEmONC skills, medical equipment use and NCDs)  Prepared and disseminated several guidelines such as for “Blood transfusion to obstetric cases” to the obstetricians, blood bank medical officers of the CEmONC hospitals; guidelines and protocols for high risk pregnancies to all secondary care hospitals in the state.  Provided several rounds of technical training (labor skills, new born resuscitation techniques, ultra-sonogram) to specialists and nurses (skilled birth attendants), for managing maternal and neonatal emergencies at CEmONCs. 48 Provide emergency  Provided free emergency transportation for mothers to transport services to CEmONCs, and to inter-facility transfers if mothers had to be reduce delayed referral. referred to tertiary care centers for complications. Train personnel who  Established a high risk antenatal screening program in two attend deliveries in districts of the state in July 2014, later scaled-up to 20 districts homes and primary based on the appreciable outcomes achieved. The program care facilities to involved screening of all ante-natal mothers (mobilized by recognize obstetric village health nurses) by a team led by obstetrician and medical emergencies, also acute officers of the PHCs in a camp mode. Identified high risk respiratory infections, mothers were referred to CEmONCs in advance of their delivery and to identify and date for better management of complications and to reduce track high-risk infants. maternal deaths. Program included extensive training of all relevant providers.  Print, outdoor and electronic IEC materials were developed and successfully deployed to improve uptake of services at CEmONCs. IEC materials are stored on TNHSP website and available for future use by the NHM and DoHFW Improve quality of care  Established a process to analyze every maternal death in to prevent maternal secondary care and tertiary institutions once a month, by a team deaths (sub-component of OG specialists led by the concerned Nodal Officer, NHM and 2(ii) is also relevant) an expert at the NHM, through video conferencing, and used findings to improve the quality of care in hospitals.  Project convened regular reviews of progress of delivering maternal and neonatal health services -- monthly performance reviews through ISMRs and CEMONC Center reports, monthly review of services by Joint Director of Health Services, Quarterly reviews by DM&RHS, and state level review by Secretary Health.  Carried out baseline and end-line assessment of CEmONCs. Sub-Component 2: Improving Tribal Health. Strengthen existing  Trained/retrained NGO partners on delivery of quality health primary & secondary care services to tribal groups. services in tribal  Introduced mobile outreach services in partnership with 12 areas through PPPs NGOs (PPO model) in 13 districts to enhance service with experienced accessibility (outpatient, maternal and child health (MCH) and NGOs (e.g., to provide laboratory services) for tribal groups, typically living in remote key staff for vacancies rural areas. Twenty vehicles, equipped with basic laboratory in selected PHC/HSCs, services and necessary medicines, were staffed by a medical reimburse in-patient doctor, staff nurse and a lab technician visited difficult to access services provided by areas once in 7-14 days per a fixed schedule. 2007-14, 1.84 NGOs, train village million tribal patients were treated. Availability of services at level tribal health door step was main factor that motivated assessment survey workers, provide respondents to use mobile outreach van services. 49 mobile clinical services  Supplied 30 four-wheel drive vehicles for emergency in 12 identified areas). ambulance service for difficult terrain to increase uptake of (Note: very detailed services by tribal groups. plans were developed  Established a Sickle Cell Anemia (SCA) screening and and summarized in the treatment program (a genetic disease specific among tribal PAD, Annex 10) population in Nilgiris district) in three NGO run hospitals in Nilgiris and Coimbatore districts. A standard management protocol for diagnosis and treatment of SCA was followed. From 2008/09-2013/14, 21,900 persons were screened; 5,158 people given secondary confirmation tests; 252 persons with confirmed SCA enrolled in a treatment program. All patients received genetic counselling, more than 90% received services (blood test, body check-up etc). Genetic, premarital, and antenatal counselling was provided to those in SCA endemic areas to ensure long-lasting health-seeking behavior. Impact: About 63% of the respondents (end line study) reported reduced instances of critical illness because of this intervention.  Instituted a bed grant scheme to provide free in-patient care for tribal patients (3 NGO hospitals provide a range of services including complicated deliveries including C sections and pediatric services). Between 2007/08 and 2013/14, 11,889 in- patients received care. An analysis of per patient cost of the scheme was carried out in order to rationalize the charges in line with the reimbursements authorized under the statewide insurance program. An end-line study found that almost all patients were satisfied with the facilities, many had used the scheme for delivery related services or general ailments, 93% received medicines, 88% considered services to be good quality, most agreed that the program motivated more mothers from tribal community to deliver in health facilities, and that the scheme reduced self-medication/visit to traditional healer.  Launched a program in 4 PHCs to enable pregnant tribal women to stay at PHCs for up to a week prior to their expected date of delivery to encourage institutional deliveries. The program was handed over to the National Health Mission in 2010, which scaled up to over 20 PHCs in remote villages. IEC strategies will be  Print, outdoor and electronic IEC materials were developed and directed at behavior deployed on preventive and promotive health behaviors for tribal change so (i) those who populations. are underserved  A campaign using traditional modes of communication was demand better services deployed to inform tribal and disadvantaged populations and and are better able to encourage use of public health services. manage their own health care; and (ii)  Trained/retrained NGO partners for delivery of quality health more responsive care services to tribal groups. 50 behavior by service  Established Tribal Patient Counselling facilities (32 in 2008- providers. 2009 and 42 in 2013-2014) at 42 district, taluk and non-taluk hospitals as well as in selected PHCs in partnership with local NGOs, to improve health seeking behaviors of tribal communities in Government hospitals, assist tribal patients to navigate through health facilities, better understand doctors’ advice and prescriptions and for preventive and promotive health care. Between 2008/09 and 2013/14, about 1.936 million patients availed the services of these counsellors. Feedback: 2/3 satisfied and 1/3 highly satisfied with counsellor’s service. Comments: all fully achieved.  Carried out mid- and end-line assessments of all four above schemes through PPPs to draw lessons for further improvement.  Uptake of services by tribal groups increased from 16,379 tribal beneficiaries in 2013-14 to 26,915 in 2014-15.  Handed over these initiatives to the NHM and GoTN for sustainability. Sub-component 3: Facilitating Use of Hospitals by the Poor and Disadvantaged (i) community  Added 108 Emergency ambulances (currently 730 ambulances mobilization by NGOs in operation). Since 2008-09, use of 108 ambulances for and outreach workers; maternity cases increased from 20.2% to 26.3% and for (ii) well-designed cardiovascular cases from 5.4% to 6.1%). Added 30 four-wheel behavior change drive vehicles for difficult terrain (June 2014) (target was 200 strategies for health more ambulances) promotion;  Print, outdoor and electronic IEC materials were developed (iii) counseling centers and successfully deployed to encourage uptake of emergency run by NGOs and local transportation services. A campaign using traditional modes of self-help groups to communication was deployed to inform tribal and provide information disadvantaged populations and encourage use of public health patient rights, services. (noted above also) availability of services, and legitimate charges  Provided 63 mortuary vans to provide a Free Hearse Service for services. (FHS), operated through a PPP with the Indian Red Cross (iv) interpersonal Society, supporting poor families in their time of need (in three communication training months (April – June 2015), 21,505 deceased were transported for health personnel to by the FHS. improve provider  Established 185 Patient counseling centers through NGOs in behavior. all CEmONCs and selected non-CEmONC hospitals to facilitate access to information by the poor and disadvantaged patients. This activity was discontinued in the Additional Financing phase. 51 Component 2: Developing Effective Models to Combat Non-Communicable Diseases and Accidents (USD 26.68 million) Sub-component 1 : Supporting Health Promotion Help develop the evidence-  Project used data on smoking, NCDs, TN’s very high rate base to advocate policy of traffic accidents to advocate for policy change. change (tax reform, policies,  GoTN adopted several policies on tobacco including enforcement) by analyzing restriction on sales to youth, advertising and smoking available data, and restrictions near schools. commissioning special studies (Dropped in 2010) Mass media BCC on  Designed and carried out a community-based BCC for smoking cessation, healthy CVD prevention. diets and exercise,  Carried out community based interventions, leveraging (Mass media dropped 2010) women’s Self Help Groups (SHGs) on preventing, screening and treatment select NCDs. Over 250,476 (97% Community-based of the target) women’s self-help groups through Tamil interventions for enabling Nadu Corporation for Development for Women environments and targeting (TNCDW) oriented in risk factors for CVDs and specific groups such as encouraged to avail screening and comply with treatment women if screened positive for risk factors or disease conditions.  Carried out health promotion activities in 16,369 School-based health government and aided schools in the state with the Sarva promotion Shiksha Abhiyan (Department of Education). (establish health-related school policies, provide safe water and sanitation, skills-based approach to health, hygiene and nutrition, and healthier school meals  From 2010, expanded health promotion activities for prevention of CVD in schools in collaboration with education department, completed IEC prevention activities in worksites Workplace-based health promotion  Workplace based health promotion activities conducted at 400 worksites with the Department of Labor. (smoke- free workplaces, programs to help employees quit smoking, workplace exercise and healthier food available in the cafeteria etc.) life-style counseling centers  adults screened for hypertension were counseled for life- to help control cardio- style modification vascular risk factors in district hospitals through PPPs with (Dropped in 2010) experienced NGOs to provide advice, particularly to poor and disadvantaged patients, on risk factor management 52 Sub-component 2: Pilot Testing Clinic-Based NCD Control Pilot 1: screening and Carried out two NCD pilots on cervical cancer and treatment of hypertension cardiovascular diseases (CVD) in two districts each – cancer using anti-hypertensive cervix February 2007-September 2010, and hypertension July medications. 2007 -- September 2010 -- to serve as examples for state-wide Pilot 2: assess costs and roll-out after evaluations of the pilots. benefits of universal cervical The NCD pilots were the first of their kind in the region. cancer screening and Health professionals were trained in the skills they needed to treatment. carry out the pilot interventions NGO partners will provide agreed services such as community-based Cervical cancer pilot: implemented a sensitization and mobilization, health mobilization program, among 30-60 year old women, promotion and follow-up of established functional screening center at PHCs and general registered patients. hospitals in the pilot districts, women tested positive during Implementation will involve cervical cancer screening were referred for treatment training staff and private providers, hiring additional Hypertension pilot: patients diagnosed with hypertension staff, providing extra were provided treatment and followed-up, adults screened for medications as per “stepped- hypertension were counseled for life-style modification up care” protocol), etc. Operational research – collect The National Institute of Epidemiology, Chennai, evaluated and analyze data on cost, field the pilots for hypertension and cervix screening during 2008- effectiveness of risk factor 10, and concurrent evaluation of the clinic based management; operational screening/treatment program of four diseases (cancer of issues such as adherence, and cervix, breast cancer, hypertension, and diabetes) in all challenges of implementing districts since 2011, as well as end line evaluation of school the intervention in different and community based interventions. Lessons learned from the settings (more/less pilots informed design of program for scaling up NCD industrialized) interventions state-wide. AF: Scale up NCD Programs throughout Tamil Nadu in two phases (16 districts each). AF: Scale up throughout the  Scaled up hypertension and cervical cancer pilot, added state screening and treatment diabetes and breast cancer - provided functional screening of specific NCDs (based on services for screening of cancer of cervix, breast cancer, the pilots) hypertension, and diabetes free of cost at 1,753 PHCs, 270 GHs, 23 Government medical college hospitals, ESI dispensaries and hospitals, and 100 selected municipal health facilities in the state.  Provided reagents, consumables, drugs, and necessary equipment for implementing the NCD program.  Recruited 2,344 NCD staff nurses in health centers to facilitate the NCD screening program. Trained them to counsel patients on accessing screening, complying with advice and medication and ensuring follow-up care. The 53 salaries of the staff nurses were reimbursed by the project. The GoTN has absorbed the NCD staff nurses in regular cadres and their salaries are now being paid through domestic budget and NHM funding.  Provided skills training to 1,155 clinical staff comprising female medical officers and staff nurses of private empaneled hospitals for clinic based interventions (October, 2014 – March 2015); and trained 190,567 persons affiliated to SHGs, statisticians affiliated to the Integrated Disease Surveillance Program (IDSP) and staff nurses in NCD online screen use.  Based on the lessons from the CVD pilot, made improvements in health care delivery such as in dispensing medication supplies for longer durations.  Several print, outdoor and electronic IEC materials were developed and successfully deployed on preventive and promotive health behaviors with respect to NCDs. The mass media campaigns were very well received as indicated in concurrent evaluations conducted by NIE.  Screened 77.4% of over 30 years of age persons in Tamil Nadu for hypertension, and 71.3% of the women in the age group of 30-60 years for cancer cervix. During July 2012 and September 2015: (i) screened over 29.03 million individuals for hypertension, 2.972 million were positive and put on treatment (a positivity rate of 9.62%); (ii) screened 23 million individuals for Diabetes Mellitus, 0.958 million positive were put on treatment (a positivity rate of 4.17%); (iii) screened 10.3 million women for cancer cervix, 0.353 million positive and availed higher level diagnostics and treatment (a positivity rate of 3.45%); and (iv) screened 12.50 million women for breast cancer, 153,330 women positive, availed higher diagnostics and treatment (a positivity rate of 1.23%).  Provided life style counseling to all individuals screened  Higher order diagnostic, medical, pharmaceutical and surgical interventions required for breast cancer and cervix cancer services linked with the package of services available under the Chief Ministers Comprehensive Health Insurance Scheme for individuals with a certified annual family income of less than INR 72,000. (FP) Monitor and evaluate the  An external quality assurance program was established for NCD interventions all laboratories of health facilities running the NCD screening program with Christian Medical College, Vellore to ensure high quality of diagnostics. Contract is 54 performance-based, with service norms for quality and turn-around time.  Evaluations of the clinic based screening/treatment program of four diseases (cancer of cervix, breast cancer, hypertension, and diabetes) done in all districts since 2011, also an end line evaluation of school and community based interventions. Sub-component 3: Traffic Injury Prevention and Treatment Additional support for health Project advocated with state to pass new laws on helmet and promotion to encourage seat-belt use, with police on enforcement, and with on-going helmet use, obey traffic rules, WB-financed Transport Project to visibly mark all spots not drink and drive. where traffic accident fatalities occur. Strengthen emergency More than 200 additional ambulances brought into operation, transport through and response time for ambulances fell well below the partnerships with NGOs to standard of under 30 minutes. Ambulances stationed at police place fully equipped stations, all health facilities, and other strategic places near ambulances at accident-prone accident “black spots”. spots on identified highways. Provide training and Extensive training program established and all cadres equipment for paramedics in trained, with continuous training as needed, including in accident relief for better immediate care protocols, trauma care, etc. immediate care (“golden Emergency services in TN are operated by a private agency - hour”) The Emergency Management and Research Initiative Develop standard treatment (EMRI), which had started in AP. Its demonstrated and protocols for trauma care, documented impact in TN enabled it to expand to 15 other emergency and poison states in India. Initially, EMRI bore 5% of the costs of the management, to improve services provided, now it has a lump-sum contract that covers trauma management in public full costs. and private facilities in Poison centers set up in all 32 districts, protocols developed, accident-prone areas. training done. The Strategic Planning Unit Monthly helmet-use data collected and analyzed, but not used will analyze and use data for to advocate for enforcement. The state backed away from advocacy for policy changes enforcing the helmet law in the face of strident public and better enforcement of resistance. Project decided to drop this sub-component – traffic rules and regulations. initial traction lost. Component 3: Building Capacity for Oversight and Management of the Health System (USD 60.86 million) Sub-Component 1: Strengthening Monitoring and Evaluation Strengthen the HMIS  HMIS software developed, hardware installed, and full Develop a new computerized system rolled out in phases (December 2008 onwards) HMIS system, pilot it, and HMIS comprises (i) HMS which automates data on all install it across the state. clinical activities public health care facilities; (ii) MIS online reporting platform for clinical and ancillary support 55 Provide training at all levels services, national health programs and administrative on how to use and maintain information for all public health facilities; (iii) CMS data the system. on academic activities of government medical colleges; (iv) UAS for data from Dr. MGR Medical University; and (v) customized websites for government medical colleges.  Established a central helpdesk with adequate staffing.  DoHFW appointed IT coordinators in all districts, and e- core teams in hospitals to solve IT issues.  GoTN issued government orders for (i) implementation, sustainability and usage, (ii) responsibility of end users, (iii) budgetary provisions for maintenance and support, (iv) removal of manual records, (v) creation of new posts at district and state level to support ICT interventions, (vi) instructing Heads of Departments and Directorates to use data from HMIS for purpose monitoring, review and analysis, and (vii) formation of a dedicated team at the Directorate.  HMIS full function: HMS in 264 secondary care hospitals; MIS in 274 GHs, 70 municipal dispensaries, 67 medical colleges and 1,889 PHCs; CMS in 20 government medical colleges; and UAS in TN Dr. MGR Medical University. HMS for DME institutions was also completed.  Monthly reporting of hospital-level data on service utilization on-going through the ISMRs - all 264 secondary level hospitals report on 20 Quality of Care Indicators monthly using HMIS.  GoTN integrated HMIS with e-TAAL (Electronic Transaction and Analysis Layer), completed transaction count of HMIS is reflected in national Govt. e-TAAL site.  SHDRC set up as a central repository of data, collate, mine and run higher order analytics on data from 20+ Directorates, and provide easy to use dashboards for various levels of administrators and managers in the health department, to drive and complement evidence- based planning, budgeting, and management, forecasting, monitoring and review in DoHFW. Carry out independent  Baseline, mid-line and end line patient satisfaction and evaluations of selected project quality of care surveys were done, and actions taken to activities to assess address gaps in services. innovations including surveys  Evaluations of NCDs completed (see above) on NCDs, patient satisfaction and out-of-pocket spending.  OOP survey not done because data are collected by NSSO  Comprehensive assessment of infection control and waste management systems in public health sector done, 56 covering policy and operating environment, efficiency and effectiveness of training through Regional training Centers, improvements in knowledge of health personnel, and impact of practices of bio-medical waste management in public health facilities of TN Sub-Component 2: Improving Quality of Care (a) Develop and implement  All 267 health institutes report monthly on 20 efficiency, quality indicators, performance and quality of care indicators, and are ranked (b) monitor quality of A, B, C and D. C and D scores trigger additional support services through base-line, and guidance to Chief Medical Officers of to bring the mid-term and endline surveys hospital back on track in performance, efficiency and (c) Establish Quality quality of care. Improvement Circles in  Quality Circles of Excellence (QCE) established and health facilities to track institutionalized in 267 secondary health facilities in the progress on indicators, state with representation from all cadres of personnel in monitor implementation of the health institution. They hold monthly meetings, maternal death audit, medical review data and issues affecting quality, and discuss their audit and prescription audit; resolution. (d) Develop protocols to improve management of key  Developed protocols for improved management of key problems (e.g. hypertension, health service delivery activities supported under the smoking cessation, cervical project, and trained health service staff at all levels in the cancer, hemorrhage, use of protocols and manuals. emergency care, etc.) and  Trained health personnel of secondary hospitals to train providers in use of improve quality of care, including in hospital protocols and manuals, and management, rational use of drugs and skills-based rational drug use. training: 80 CMOs in hospital management; 739 staff in (e) Develop hospital hospital administration; 1,692 staff in Quality Indicators; inspection checklists, set up 1,915 in rational use of drugs. panels of experts to do  Trained and retrained 398,285 health personnel (October regular inspections and 2010-June 2015). provide feedback.  12 hospitals accredited (3 full and 8 progressive level accreditation, and 1 entry level accreditation); more hospitals completed final assessment and awaiting decision/feedback from Quality Council of India.  Strengthened clinical laboratory services in secondary care hospitals and medical college hospitals by (a) initial PPP with private agency, to provide services at regional laboratories in 5 District Head Quarters Hospitals, later transferred the activity to the Directorate of Medical and Rural Health Services (b) supplied necessary equipment for laboratories in all secondary care hospitals in TN. Provide to hospitals: (i) basic Mentioned in the PAD under this component, but addressed amenities; (ii) equipment and under Component 4, sub-component 1. See below. inputs; (iii) train technical and 57 Managerial staff at all levels Strengthen regulation of  Strengthened capacity of DoHFW to manage and support public and private hospitals the accreditation process. building on the existing regulatory system: help The PAD noted that “Regulation of the private sector would GOTN implement stronger take longer than that in the public sector and would depend on oversight system for both certain factors beyond the scope of the Project.” private and public facilities. Sub-component 3: Strengthening Healthcare Waste Management  HCWM plan implemented by 449 public health Install hospital waste institutions, consistent with the GoI’s Biomedical Waste management system in all (Management and Handling) (Second Amendment) Rules, hospitals per GO1guidelines 2000 which details good practices to be followed and all roles and responsibilities for effective disposal of health care waste. Implementation was phased. First, a pilot was done in 11 hospitals in 2 districts over 2 years, and the experience evaluated independently, gaps identified and corrected. Phase two scaled up in 449 public health institutions (270 secondary care, 41 tertiary care, 130 thirty bedded PHCs and 8 ESI hospitals).  Over 49,500 health personnel were trained and re-trained, Train health personnel at all in identification, collection, segregation, disinfection, and levels of facilities in disposal of health care waste and maintenance of records, healthcare waste management through a network of 11 Regional Training Centers in medical colleges (7 government and 4 private but only 2 private were active at the end), which were strengthened for sustainability of training.  Training complemented by behavior change campaign. Implement guidelines on  Established effective coordination with municipal bodies proper segregation and color- and the State Pollution Control Board, which performs coding, transport, and annual quality assurance inspections of all treatment disposal. (PPPs with NGOs) facilities.  Established PPPs with 30 Common Treatment Facilities (CTF) for collection, disinfection and disposal of waste in secondary level institutions. Till 2013, project provided all hospitals with consumables, equipment and personal protective gear for ICWM, including per bed or per kilogram cost for disposal of health care waste to CTFs. Since 2013, the flexifund of National Health Mission has paid for ICWM in all institutions, including payments to CTF. ICWM implementation and monitoring handed over to the Directorate of Medical and Rural Health Services and Directorate of Public health and Preventive Medicine. 58 Comment: management of health care waste institutionalized and strengthened in public health institutions, (a key element of Bank’s mandatory Environmental Management Plan), contributing to safer, and more effective health care. Sub-component 4: Building Capacity for Strategy Development and Implementation Set up Strategic Planning  Established a functional Strategic Planning Cell (SPC) unit to identify issues, as a think tank for GoTN, and to lead significant aspects generate and evaluate options of various health delivery interventions supported under to improve HS performance TNHSP. Before project closed, SPC/TNHSP prepared (eg strategic planning, HR, proposals to Government for issue of Government Orders commission operational and for handing over TNHSP programs to the Directorates of policy research, study key the DoHFW and for sanction of budgets. issues for effective  Tamil Nadu Medical Code was revised by SPC. implementation of project activities to achieve  Conducted end line evaluation studies on CEmONCs and desired output tribal health activities, quality of care, patient satisfaction, infection control and waste management. Establish PPP wing in GoTN  A PPP Wing was established by SPC. to promote inclusive  Evaluation studies of the PPPs were carried out for: 2 partnerships with the private pilot PPPs for health care waste treatment; 2 PPPs for bed sector in provision of grant schemes, 2 PPPs for sickle cell anemia and several healthcare, especially in hard PPPs for mobile van outreach programs in tribal areas; to reach areas with low access PPP for emergency transport (ambulance); and several to government health care PPPs for the provision of patient counsellors at services, and in sectors where CEmONCs and non-CEmONCs facilities co-ordination is essential. The  Number of performance based contracts delivering health wing would also manage and care services increased during the AF by 37 (target was 9 monitor all PPP contracts PPP contracts) -- 5 contracts for Regional Diagnostic signed under the project. Labs, 2 Housekeeping services contracts and 30 CTF contracts for handling hospital waste. Conduct a Health Insurance  Pilot not done because new Chief Ministers Health Pilot to explore feasibility of Insurance Scheme was set up and rolled out widely with providing community-based full administrative support from the project. health insurance on a  International Conference on Health Systems Financing reasonable scale (May 6 – 10, 2010) held in Chennai with participation from GoI, GoTN, other states across India, international and bilateral development partners and others to discuss implementation issues, share international experiences to promote good practices, and guide national and state policies for improving health systems. Strengthen Project  Added staff to PMU Management Unit to enable  PMU monitored and reviewed regularly project activities PMU to track progress and including ISMR and grading of hospitals, outsourcing of carry out project procurement housekeeping services, quality of care activities, poison and financial management treatment centers. activities. 59 To be a useful pioneer  TNHSP Stakeholders Workshop (August 28-29, 2015) in project, experience must be Chennai took account of the project’s performance, shared disseminated experiences, discussed post-project plans.  Developed and disseminated publications on TNHSP activities and innovations: Training Manual on Quality of Clinical Care Indicators, Handbook on Infection Control and Biomedical Waste Management, Health Management Information system, Documentation and Dissemination of a Best Practice, and Standard Treatment Guidelines Manual. Project staff also participated and contributed to inter- and intra-state workshops.  Many delegations from other states, and development partners (Bill and Melinda Gates Foundation, JICA, USAID etc) have come to see the programs set up under the project, and emulated them elsewhere. Component 4: Improving the Effectiveness and Efficiency of the Public Sector to Deliver Essential Services (USD 73.28 million) Sub-component 1 : Rationalization of Secondary Care Facilities Refurbish and upgrade  Completed all planned civil works (35 Phase I secondary secondary care hospitals to care hospitals; 190 hospitals in Phase 2, maternity blocks ensure functioning basic in 8 government medical colleges, and the Annex building amenities (water, sanitation of the Directorate of Medical Services). and electricity), and ability to  Civil works in 8 CEmONCs in medical colleges and 12 provide care per new service hospitals undergoing NABH accreditation were norms for each grade of completed under the AF. hospital. Provide equipment required  Provided essential hospital equipment required to deliver to deliver services per norms. services per norms for secondary level hospitals. In total, constructed, renovated and/or equipped 2,330 health facilities (1,889 PHCs, 274 GH, 100 medical dispensaries and 67 medical college hospitals (throughout project) Sub-component 2: Rationalizing of Equipment Undertake one-time repair, Completed. Electronic inventory of about 100,000 pieces of after assessing inventory and equipment in hospitals under various departments of DoHFW repair needs. enables more efficient equipment management. Implement a good  Established a system to track and improve utilization, maintenance system similar to repair and maintenance of equipment in health care Andra Pradesh (through facilities. Annual maintenance contracts for complex TNSMC, equipment expensive equipment, regional workshops manned by in- suppliers, and local hospital house biomedical engineers and technicians to service officials). moderately complex equipment and to manage Annual Maintenance Contracts (AMCs) and other ad hoc 60 contractors, and assist hospitals in training users in basic maintenance and care of equipment.  Recruited and trained 48 biomedical engineers to maintain and manage medical equipment. They liaise closely with TNMSC, TNHSP and equipment manufacturers and suppliers, and manage equipment repairs. AF: Strengthen equipment  TNMSC’s capacity for managing pharmaceuticals and and pharmaceuticals medical supplies using World Bank procurement management procedures was strengthened under the project. The new HMIS provides real time data on pharmaceuticals use and inventory, which could enables better management of pharmaceuticals. Sub-component 3: Human Resource Planning and Development Establish and implement new  Revised and rationalized manpower service norms. staffing norms, conduct  Recruited 1,212 NCD staff nurses under contracts for extensive training of Phase II hospitals (IDA financed), 562 staff nurses for government staff, including CEmONCs and 1,132 NCD staff nurses for Phase 1 management training for districts (GoTN financed) during Additional Financing, in hospital administrators. accordance with established staffing norms to improve overall efficiency and performance. GoTN has taken on salaries of additional staff from own resources and financing of NHM.  Trained and re-trained 398,285 health professionals (during Additional Financing, Oct 2010-June 2015) to enhance capacity of the public health system and to enhance skills and improve quality of care in all areas supported by TNHSP. These included senior medical officers and administrators, doctors, nurses, clinical, paramedical and laboratory technicians, health personnel in bio-medical waste management, medical assistants, ANMs, counsellors, pharmacy staff, HMIS staff and administrative assistants. Most training was done by reputable organizations recruited from outside. Substantial training activities were also undertaken during the original project (2004-2010) to improve skills and knowledge of health personnel. Carry out activities to  Trained service providers in interpersonal communication improve staff morale and to encourage team effort and role clarity and recognition. courtesy to patients and set up incentive measures. Subcomponent 4: Enhancing Management of Public Facilities (dropped at AF) Twin hospitals with well-  Established a performance grading system of hospitals to known private hospitals. identify and resolve bottlenecks and improve performance Provide incentives to hospital in hospitals experiencing difficulties/ performing poorly. 61 administrators for high performance. Where twinning is not  Introduced PPPs to provide and manage health services feasible (no private hospitals including operation of ambulances financed under the available), consider/test other project, free hearse service, housekeeping services, ways to improve performance laboratory diagnostic services, tribal mobile outreach of lagging public facilities services, screening and treatment of sickle cell anemia, tested, including: eg counseling services for tribal patients, and a bed grant improvement budget fund, scheme. PPP contract with NGOs to  Outsourced Housekeeping Services in 48 secondary care operate the facility, recruit a hospitals to a competitively selected vendor (cleaning, hospital manager or sanitation, security, assistance in electrical, plumbing, consultant on a performance- catering, cooking, laundry, gardening and carpentry linked contract. services). The contract was handed over to the Directorate of Medical and Rural Health Services. 62 Annex 3. Economic and Financial Analysis The economic and financial analysis of the project in the PAD consisted of a qualitative discussion of the major components, pointing out that the CEmONCs could save lives at low cost; that NCDs imposed a significant economic burden through lost productivity and thus programs to address NCDs could reduce DALYs lost in a cost-effective manner; and infrastructure investments would be pro-poor. There was also a brief discussion of project implications for future recurrent spending, the affordability of Borrower funding to the project, and financial sustainability assessed by placing project costs in the context of the overall health budget. No economic rate of return (ERR) or net present value (NPV) was calculated. The economic analysis of this ICR updates and extends this analysis. It assesses the major project components through an economic lens, including the following: (1) recurrent costs and budgetary implications; (2) cost-effectiveness considerations; (3) cost-benefit considerations; (4) efficiency considerations; (5) equity considerations. The key message is that the available evidence suggests there were significant economic returns from the project and no major red flags related to economic impact. The economic rationale for public spending in the health sector should be noted at the outset. There are many issues, including insurance market failures, market power among the providers of medical care, externalities associated with some health goods, newer behavioral economic theories that emphasize under-utilization of care, and equity considerations. All are cited as reasons for government intervention. These factors help explain why over 80 percent of health spending in high-income countries is typically public (i.e., financed through general taxes or social health insurance). In India the share is just half this amount, but can be expected to trend upwards over time as it pursues an increasingly MIC agenda. During the life of the project, Tamil Nadu enjoyed strong economic growth, even faster than the robust 7.5% average annual Indian average over the same period. While this trend slowed during 2011-13, it picked up momentum again in 2014. There are some fiscal challenges, however, with a deficit of 2.7% of state GDP, and a rising debt to SGDP ratio (although still moderate at about 20 percent). Budgetary implications of recurrent costs The project’s components included both capital investments (e.g., hospital improvements) and programmatic initiatives (e.g., NCD screening) that imply ongoing recurrent costs that will endure long after project completion. An important question is whether these costs can be absorbed in GoTN’s regular health budget. Counterpart funding amounted to 14% of total project costs, very close to the expected level in the PAD. This was less than 0.5% of the Tamil Nadu health budget over the course of the project, and was therefore easily manageable. While the health budget increased significantly in absolute terms, Tamil Nadu’s budget share for health stayed remarkably constant over the project 63 life, with the share almost identical in 2015-16 as 2005-06 (4.5% vs. 4.6%). The PAD made note of a declining trend in state health spending in the years preceding project preparation; while the stability of the health budget share over 10 years is welcome, a higher allocation would have been preferable given that most LMICs allocate about 6-8% of their budgets to health. However, with the strong economic growth in the state over the project years, even a constant budget share provided a substantial increase in the health budget. During implementation, total project costs reached a peak of 7.6% of the overall health budget (this was in 2010/11), slightly below the 8.9% peak forecast in the PAD. Over the final three years, it averaged just 1%, suggesting there will be no major handover issues with respect to sustainability. This is shown in the table below. One of the more important sources of recurrent costs arising from the project was the regularization of nearly 3000 staff nurses from the NCD and CEmONC programs. This was an important and commendable step, and an important lesson for other projects for ensuring that achievements will be sustained. Although detailed salary information is not available, the total cost of these nurses should be far less than 1% of the total health budget. In brief, as expected at the time of the PAD, the budgetary implications of recurrent costs arising from TNHSP are relatively small and do not represent a concern going forward. Table: Project expenditure as a share of total health expenditure Year 2004-05 2005-06 2006-07 2007-08 2008-09 2009-10 2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 Share 0.0% 2.6% 2.5% 5.3% 5.2% 5.6% 7.6% 3.1% 3.1% 1.5% 0.7% 0.1% Cost-effectiveness considerations Cost-effectiveness evidence can help identify “best buys” for achieving health improvements within a fixed budget. There is a large international literature on the cost-effectiveness of health interventions that is broadly applicable to Tamil Nadu, even if local studies are not always available. The project supported many activities – both general and disease-specific – with varying degrees of cost-effectiveness. Among high-burden diseases addressed by TNHSP, global evidence drawn from the Disease Control Priorities21 project (DCP) suggests that the chosen interventions were cost-effective. For example, cardiovascular disease management (e.g., screening and treatment for ischemic heart disease) is among the more cost-effective interventions available, with a cost per disability- adjusted life year (DALY) averted of approximately $1000. Drug treatments for more acute episodes such as heart attack and congestive heart failure are even more cost-effective, and also benefited from the project activities. Treatment costs for the more treatable cancers covered by TNHSP (i.e., breast and cervical) have ratios between $1300-6200 per year of life saved, compared to $53,000 to $163,000 for less treatable forms. Thus, the project appropriately 21 Laxminarayan, R., et al. (2006). “Advancement of global health: Key messages from the Disease Control Priorities project”. Lancet 367: 1193-1208. 64 targeted the more cost-effective cancer interventions. In addition, “improved quality of comprehensive emergency obstetric care” and “neonatal packages” were both identified as neglected low-cost opportunities in the Disease Control Priorities project, with very favorable costs per DALY averted. Taking a broader perspective, over the life of the project Tamil Nadu’s overall health budget became more cost-effective by focusing more on primary care, with its budget share relative to secondary and tertiary increasing by more than 15 percentage points, while the tertiary share declined by more than 10 percentage points. Cost-benefit considerations A health project’s economic benefits can be estimated in two ways. Improved health outcomes can contribute to a healthier workforce that raises economic growth and productivity. However the economic literature emphasizes that the intrinsic (direct) value of a healthier population (as proxied by rough estimates of willingness to pay for better health) is much more important than the instrumental (indirect) value that is achieved by way of higher economic output. Consider first the benefit in higher growth and productivity. A recent study by the Harvard School of Public Health (HSPH) for the World Economic Forum22 estimated that the economic loss from cardiovascular disease, diabetes, and cancer across India between 2012 and 2030 would be about US$2.5 trillion. It also analyzed a number of specific interventions, including the pilot phase of the TNHSP NCD program. It estimated that the return on investment of the program was well in excess of 15%. A cost-benefit ratio can also be estimated by converting health gains achieved by a project or intervention into monetary terms based on the value of health. Although this exercise may sit uncomfortably with some, it can be useful for policy purposes, and typically serves to underline the very high value attached to better health. The standard economic approach for quantifying the benefit of better health in monetary terms is based on the concept of the “value of statistical life” (or life-year). Studies from around the world suggest that the value of a statistical life-year is at least five times higher than GDP per capita, which translates into about $11000 in Tamil Nadu. With this value, and if project spending was on average about $20m per year, then the project would only have to achieve an average of 2000 additional life years annually to “break even”. This threshold is very feasible, given that the program generated an estimated 3 million new hypertension cases identified through screening, 1m diabetes cases, 350,000 cervical cancer and 153,000 breast cancer cases detected. All this was achieved at a project cost of only US$19m. Moreover, within the period 2010-2014, there was a 16% decline in maternal mortality and a 12% decline in neo-natal mortality at those medical colleges that were part of the TNHSP CEmONC intervention. Thus, even if only 1% of those put on treatment attained one additional year of life as a result, the benefits would substantially exceed the costs. In brief, the project appears to have achieved a very favorable cost-benefit ratio. This would be consistent with an existing literature, most advanced in the US, which has found benefit-cost 22 Bloom, D.E., et al. (2014). Economics of NCDs in India: The costs and returns on investment of interventions to promote healthy living and prevent, treat, and manage NCDs. World Economic Forum, HSPH. 65 ratios of greater than 6 to 1 for anti-hypertensive therapy and medical management of coronary heart disease, and greater than 1 to 1 for breast cancer treatment.23 More generally, there is a large literature suggesting very high rates of return from health spending due to the very high value that people attach to longer, healthier lives. Efficiency considerations A project can contribute to efficiency if it helps to achieve the same health gains at lower cost (or, equivalently, greater health benefits for the same cost). Starting with a broad perspective, there is little evidence that the overall health system in Tamil Nadu is especially wasteful. Health accounts for about 4.5% of total government spending, which is relatively low compared to many countries (LMICs are usually in the 6-8% range). And while costs are not high, health outcomes are good, suggesting good value for money is being achieved. The Tamil Nadu health system also achieves a good balance between primary, secondary, and tertiary care spending, and indeed over the project life, the share spent on primary care increased at the expense of tertiary care. It is also difficult to identify areas where project achievements could have been realized more cheaply. The major investments were generally made at the appropriate level of care – for example, CEmONCs were not and should not be developed at the primary care level, whereas NCD screening should be and was done to a significant extent at lower level facilities. Moreover, the funded services are helping to address conditions that represent well over half of the disease burden in Tamil Nadu, so resources were not being misdirected to low-priority interventions. In qualitative terms, numerous project activities are likely to have achieved efficiency gains. Many did so by making investments in the quality of care to strengthen the link between outputs and outcomes – for example, training 400,000 health care workers, accreditation reforms, a system to track the utilization, repair and maintenance of equipment, and so on. The project’s PPP initiatives – for example, for housekeeping and laboratory services, are also likely to have generated better value for money than previous arrangements. However, concrete data on these gains are not readily available. As reported in the project indicators, there was a modest improvement in the bed occupancy rate. More concretely, improved efficiency was an important objective of the HMIS, and specific efficiency indicators were developed as part of the HMIS. It is difficult to quantify the efficiency impact of HMIS in monetary terms, but based on the success of using ISMR to improve quality of care, the potential is clearly there. The total cost of HMIS over the project was about $30m, slightly more than 2% of the current Tamil Nadu health budget. Thus the HMIS would easily pay for itself many times over if it can be leveraged to achieve even small (e.g., 1%) efficiency gains on an annual basis. 23 Rosen A. et al. (2007). “The Value of Coronary Heart Disease Care for the Elderly: 1987-2002”. Health Affairs 26(1): 111-23. 66 Equity considerations An improvement in outcomes for the poor and vulnerable was an explicit goal of the project, as expressed in the PDO. As noted in project indicators, the hospitalization rate of the bottom 40% increased by about 25%, from 33.4 to 41.9 per 1000. In addition, utilization of CEmONCs by SC/STs also increased, as per the relevant project indicator. The sub-component on tribal health only represented a project cost of about $1.8m, and had a positive impact on access to care within that population. Ambulance services also were intended to have a pro-poor orientation. 67 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a.) Task Team members Responsibility/ Names Title Unit Specialty Lending Preeti Kudesia Senior Public Health specialist SASHD TTL Mohan Gopalakrishnan Financial Management Specialist SARFM Financial Management S. K. Bahl Sr. Procurement Specialist SARPS Procurement Ruma Tavorath Environment Specialist SASES Environment Safeguards V. Vemuru Social Development Specialist SASES Social Development Snehashish Rai Chowdhury Operations Officer SASHD Operational Aspects Benjamin Loevinsohn Sr. Public Health Specialist SASHD Public Health Isabella Anna Danel Sr. Public Health Specialist LCSHH Public Health Sara Gonzalez-Flaveli Sr. Counsel LEGMS Legal Philip Beauregard Sr. Counsel LEGMS Legal Shreelata Rao Seshadri Consultant – Social Development Social Development David Porter Consultant – Biomedical Engineer Biomedical Engineering Subhash Chakravarty Consultant - Architect Architecture Nirupama Sarma Consultant – Health Promotion Health Nina Anand Program Assistant SASHD Administration Mohammad Khalid Khan Program Assistant SASHD Administration Supervision/ICR Bushra Binte Alam Senior Health Specialist GHNDR Task Team Leader Sangeeta Carol Pinto Operations Officer GHNDR Operations Officer Ramesh Govindaraj Lead Health Specialist GHNDR Health Specialist Owen K Smith Senior Economist GHNDR Economist Ajay Ram Dass Program Assistant SACIN Administration Arvind Prasad Mantha Financial Management Specialist GGODR Financial Management Atin Kumar Rastogi Procurement Specialist GGODR Procurement Rohit Gawri IT Analyst, Client Services ITSCR Information Systems Sundararajan Srinivasa Senior HNP Specialist GHNDR Task Team Leader Gopalan Preeti Kudesia Senior Health Specialist GHNDR Task Team Leader Vikram Sundara Rajan Senior Health Specialist GHNDR Health Specialist Maria Gracheva Senior Operations Officer GHNDR Additional Financing Sushil Kumar Bahl Senior Procurement Specialist SARPS-HIS Procurement Shanker Lal Senior Procurement Specialist GGODR Procurement Senapati Balagopal Procurement Specialist GGODR Procurement Sr. Financial Management Mohan Gopalakrishnan GGODR Financial Management Specialist Shashank Ojha Senior e-Government Specialist GTIDR Information Systems SASHN - Non-communicable Michael Maurice Engelgau Sr. Public Health Specialist HIS diseases Ruma Tavorath Senior Environmental Specialist GENDR Environment Safeguards Subhash Chakravarty Consultant Architecture 68 Peter A. Berman Consultant GHNDR Health Economist Maneesha Gupta E T Consultant ISGEG-HIS Information Systems Benjamin P. Loevinsohn Lead Public Health Specialist GHNDR Health Specialist Shyama Nagarajan Health Specialist SASHN-HIS Health Specialist Shreelata Rao-Seshadri Consultant GHNDR Social Development (b.) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including travel No. of staff weeks and consultant costs) Lending FY02 0.4 3.16 FY03 31.12 140.94 FY04 47.21 166.95 Total: 78.73 311.05 Supervision/ICR FY05 38.35 111.18 FY06 25.4 96.19 FY07 35.36 119.86 FY08 29.18 145.60 FY09 27.38 129.47 FY10 33.09 121.35 FY11 42.04 197.42 FY12 28.75 144.39 FY13 26.04 109.62 FY14 20.27 68.48 FY15 26.91 113.65 FY16 9.47 39.07 Total: 342.24 1,396.33 69 Annex 5. Beneficiary Survey Results See text discussion of Patient Satisfaction. 70 Annex 6. Stakeholder Workshop Report and Results Inaugural remarks  Advantage of the World Bank projects is that they bring in global experience, best practices, procurement guidelines, and monitoring indicators, put in place a very robust system and facilitate upscaling of ideas.  Due to the TNHSP and other Government programs, the number of people seeking health care in both urban and rural areas is higher in Tamil Nadu. Up-take of quality of care activities in TNHSP was very good.  Government expressed appreciation of the World Bank for its whole hearted and constant support for ensuring success of the project and that it becomes a model for other states in India to follow.  Government is committed to continuing all project activities by the DoHFW. Interventions in maternal and child health  Intervention to track high risk mothers (first as pilot in 2 districts and then up-scaled to 18 districts) resulted in a marked reduction in MMR in those districts.  During planning for CEmONC services the GIS mapping helped to identify hospitals where major deaths were taking place.  Provision of human resources proved to be a far greater challenge relative to the provision of infrastructure.  TNHSP contributed to CEmONCs through construction of maternity blocks, supporting certification, monitoring and evaluation and through training programs.  The maternal mortality rate in Tamil Nadu has plummeted to half from 2007-08 to 2014- 15 reflecting the success of intervention.  Concern that “at high risk mother camps” take doctors away from secondary and tertiary care facilities and it leads to shortage of doctors at those centers.  Fluctuations in specialist doctors and in posting staff nurses at CEmONCs are a significant impediment to service delivery.  Revised training methodology encompassing consolidated and comprehensive training for staff nurses and doctors, and scaling-up of training for the paramedical workers would help.  Protocols developed for antenatal and neonatal care and extensive training provided to doctors and staff nurses, but more needs to be done.  Interventions to reduce the IMR and MMR such as the establishment of CEmONCs within 30 minutes reach, CEmONC PHC at a rate of one per block, auditing of every maternal death enabling the identification of the circumstances leading to the death, and establishment of 24x7 delivery centers in all PHCs were effective.  CEmONCs’ reach to tribal areas was ensured by extending 20 tribal mobile medical units, birth waiting room in 17 PHCs in the foothills of tribal villages, provision of feeding and dietary charges for 7 days for AN mothers and an attender in 34 tribal PHCs and training/placement of 2,650 ASHA workers in 15 tribal/ hard to reach districts. These interventions helped increase access to health care by tribal populations.  Need for expediting the process of filling up the vacant posts, identification of the mentors in maternity wings for continuation of quality of care, revision of the training 71 methodology to encompass a consolidated and comprehensive approach and replicating the team approach. HMIS  Prior to the HMIS, no real time data was available, evidence based program management was stalled, retrieval of old manual records was ineffective and time consuming.  Human resource constraint needs addressing as the entire program is handled by 5 medical officers with the help of one ELCOT Deputy Manager.  Ways needs to be found to improve receptivity by hospital staff, connectivity and server stabilization, and basic computer knowledge.  To augment and expedite the standardization, there should be mapping of existing process and rationalization of input forms.  To minimize fragmentation of vendors, efforts are needs to ideally have single vendor for IT infrastructure.  Bank contributed immensely in the implementation of HMIS by providing key inputs and support in defining the functional requirements of different modules, supporting the capacity building process, promoting collaboration with non- governmental sectors and external experts at different stages of application development, and in the adoption of the quality assurance mechanism.  Next actions should include: bringing down the number of vendors, recruiting project management from within the TNHSP team, retaining the HMIS team until the system reaches the self-sustainable level, taking stock of the Phase I inventory and starting to plan replacement of IT equipment’s during phase 1 implementation, and immediately starting the procurement process of System Integrator.  Successful implementation of the College Management Information System was a milestone in Tamil Nadu’s medical education. A strong IT team is now needed to handle the CMS application where it will help the future generation. Health care waste management (HCWM)  From a pilot program, the HCWM was up-scaled to 449 health facilities. Over 49,000 health staff were rained/retrained in Regional Training Centers.  Sustainable training and retraining of health staff at regular intervals was recommended. Accreditation  The main objective of accreditation is to improve the quality of treatment and provide the safety for patients and employees. The process helps to rectify the defects. Out of 46 hospitals taken up by the GoTN for accreditation, 15 hospitals are already in the final stage of accreditation. Quality in health care system  TNHSP immensely contributed to skills development in the health sector.  Development of an Infection Control and Waste Management System was a milestone achievement.  Various manuals were prepared and published.  Quality of care indicators were developed and are being used to monitor the quality of care prevailing in the hospitals. 72  TNHSP provided training to improve interpersonal communication.  ISMRs were introduced and are being prepared every month; data is used to also grade hospitals.  Poison treatment centers were established.  Introduction of Quality Assurance System for laboratory investigations was introduced.  Introduction of a system for the rational use of medicines was introduced.  Ongoing attention is needed to change the mind-set of the Health Care providers for new activities.  Reports prepared in the health sector need to be validated systematically.  Training on administrative procedures needs to be provided to the CMOs.  Increased recruitment of specialists and staff nurses is needed. Universal health coverage  For policy makers, the ultimate goal is that all citizens have access to health care which is the basis of Universal Health Coverage.  Focus should be on ensuring services to the bottom quintile of population as they suffer the most.  The idea the Government of Kerala adopted was to identify top 20 percent of conditions which constitute 80 percent of morbidity in the state (22 conditions were identified that caused 70-80% of morbidity). In Kerala, government emphasized training of doctors to manage the disease conditions, which causes 80 percent of morbidity, at the PHC level, and to also ensure adequate supply of essential medicines including insulin and NCD related drugs at the PHC level.  It was suggested that traditional institutions namely the health service centers (HSCs) and ICDS centers need to be focused and further strengthened to sustain the gains made in maternal and child health including immunization and family welfare, to strengthen the capability of infectious diseases case management systems in secondary and tertiary facilities, and consider making the insurance mechanism universal so that all facilities may be able to generate adequate resources.  It was suggested that the Chief Minister’s Comprehensive Health Insurance Scheme can be a tool to extend universal health care, perhaps by expanding the scheme to include middle class, package of essential services and possibility of shifting towards primary care and offering financial protection to the population. End line studies NCD interventions:  Patient exit survey showed a sharp increase from baseline in the proportion of patients who received drugs for 30 days from PHCs (range of 40% in Theni to around 90% in Villupuram).  A high proportion of patients also received dietary counselling from nurses and doctors at the PHC level.  The proportion of patients who adhered to their drug regimen also increased.  As a result of NCD awareness interventions, a high percentage of individuals are aware of the harmful effects of tobacco and salt. 73  Recommendations: (i) Sustain diabetes and hypertension screening in the public sector and support with adequate infrastructure and human resources. (ii) Improve cervical cancer and breast cancer screenings through better awareness and improvements to the health system. (iii) Strengthen follow-up mechanisms post-screening. (4) Ensure that NCD nurses are posted in order to continue service delivery, and ensure availability of adequate drugs for 30 days for patients. (5) Sustain induction and refresher trainings; with doctors, focus on case management to improve prescription practices around achieving blood pressure control and glycemic control, and on targeting organ complications; with nurses, focus on screening/ counselling skills. (6) Develop patient-focused education programs for diabetes and hypertension to improve treatment and adherence rates. (7) Utilize TV as key information source during campaigns, and focus on obesity and physical activity, importance of long-term treatment and adherence for patients, cancer screening, and other important changes in behavior. (8) Actively involve health workers in awareness programs and explore other innovative ways of engaging communities in behavior change campaigns. End line assessment of quality of care and patient satisfaction in the hospitals under the project:  Significant improvements from baseline to end line: (1) in infrastructure at the hospitals (accessibility, power, water, and equipment), (2) in the availability of services, such as laboratory services, pharmacy services, and emergency services, a reduction in time taken to register, better conditions of wards and toilets, (3) a sharp increase in patient’s engagement with IEC materials, (4) around half of health facility workers believed that they would benefit from further training, and health workers felt that there could be an improvement in their residential quarters, (5) health workers believed that there were improvements in supervision and the frequency of staff meetings, (6) in the satisfaction with infrastructure, staff behavior and treatment outcomes (however, in-patients appeared more satisfied with services than out-patients), and (7) patients chose government services due to their perceptions of good quality, affordability and accessibility, availability. However, some study respondents expressed concerns with hospitals in some districts regarding treatment outcomes, communication skills and dual practice of doctors.  Recommendations included (i) improvements in the conditions of diagnostic services, imaging services and facility vehicles, (ii) further sensitization of public regarding health services offered by the health facility, such as NCD services, (iii) ensure good behaviour of staff to public such as communication skills, particularly at registration, (iv) ensure availability of water and soap in toilets and overall cleanliness of toilets, (v) reduce waiting times by re-visiting registration process flowchart. Evaluation of CEmONC and Tribal Health activities  Achievements of CEmONCs: (1) Overall reduction in the MMR from 109 to 68 in the past 10 years, increases in LSCS and night-time LSCS, increases in institutional deliveries and increases in treatment for pregnancy complications. (2) Increases in maternal admissions and complicated maternal admissions. (3) Efficiency (utilization of equipment) also improved - a sharp increase in the utilization of scans for Obstetrics/Gynecology cases from 2011-2012 onwards, attributed to the supply of equipment and training. (4) In terms of quality of care, high proportions of women 74 reported receiving antenatal care, having birth companions when eligible, having their babies weighed at birth, and having access to NICU facilities when required. (5) In terms of patient satisfaction, patients reported that the availability of free treatment was a primary reason for selecting government facilities, followed by good quality of care. Patients however reported that the provision of bed linens, the regular changing of sheets, and the cleanliness of toilets were inadequate.  Achievements of tribal health initiatives: (1) Implementation of Mobile Outreach Services, Counselling Services, the Bed Grant Program and the Sickle Cell Anemia to improve access to health services. Patients reported an increase in lab visits, and improvements in the quality of services. There has also been a sharp increase in patients counselled through the Counselling Program. Responsibilities of these Counsellors have also increased considerably. However, language was considered as major barrier for availing services from the counsellors of other communities. The Sickle Cell Anemia intervention seemed to have a strong impact, through increased awareness of disease status and reported effectiveness of treatment. Patients also appeared satisfied with the Bed Grant Programme, and a high proportion of patients reported using the facilities for deliveries and ailments such as fever, headaches, etc.  Recommendations for CEmONCs: (1) Clean linen and regular changing of bed sheets for new mothers and babies, and cleanliness of toilets for new mothers. (2) Availability of blood transfusion services for all patients. (3) Consider connecting health workers at the field level for effective monitoring of complicated cases. (4) Explore the use of effective induction and acceleration of labor in order to bring down caesarian section rates, and encourage vaginal delivery wherever possible. (5) Consider separate ICU for CEmONC to handle the critical and high risk cases. (6) Examine whether inputs, such as human resources and number of beds, are in line with the increase in patient demand for CEmONC services. (7) Consider the use of staff exclusively focused on recordkeeping, which would allow for nurses to spend more time on patient care. (8) Consider increasing the posting of CEmONC trained MBBS doctors at CEmONC centers for more basic care, so that specialists can focus on advanced cases. (9) Consider the use of a non-medical team to follow up on referred cases and newborns (for at least one month post discharge). (10) Medical doctors posted at PHCs should be adequately trained for early referral in case of complicated deliveries.  Recommendations for tribal health care: (i) Mobile Outreach Services (MOS) should include antenatal care, the full immunization schedule and reducing anemia among adolescent girls. (ii) Increase awareness of program amongst doctors and paramedics, and ensure the selection of NGOs based on performance and willingness to reach remote areas. (iii) Explore the use of separate areas for counseling, provide periodic refreshers to counselors, and provide rewards for good performance. (iv) Integrate the work of the Counseling program with the MOS, and with other health workers such as Village Health Nurses. (v) Continue Bed Grant Program given comfort of tribal communities in accessing facilities run by organizations known to them/in their area. (vi) Expand awareness programs for the Bed Grant Scheme to increase utilization. (vii) Consider taking advantage of strong internet connectivity by developing online programs, and uploading daily case reports. Assessing the Training and Practices on Infection Control and Waste Management (ICWM) 75  Observed 100% coverage of the ICWM initiative in all sampled hospitals, 95% training coverage in sampled hospitals, increased practice of labelling of bins, increased availability of storage room for waste, and improved knowledge levels on waste segregation and use of personal protective equipment.  Infection control officers and infection control committees have been installed and are operational. Recommendations:  Examine continued use of needle destroyers.  Ensure that all hospitals obtain authorization and renew authorization with the State Pollution Control Board.  In hilly areas, consider the use of deep burial facilities, in consultation with the State Pollution Control Board.  Consider the use of online refresher courses on the management of bio-medical waste and incorporate videos in the training.  The Regional Training Centres should conduct quarterly consolidation of pre- and post- tests conducted, and TNHSP should provide regular feedback, through reviewing reports, surprise visits and regular checks.  Staff should be repeatedly motivated by their managers to use protective gear.  Examine the considerable under-reporting of needle stick injuries at hospitals.  Ensure the availability of bio-medical waste storage rooms with clearly demarcated spaces that are accessible by vehicles, and with separate exits for the waste collection.  Ensure that responsibilities and roles for Infection Control Officer are clearly defined, and that these individuals are supported by the hospitals. Public private partnerships Tamil Nadu Chief Minister’s Comprehensive Health Insurance Scheme  A four year old scheme, and provides financial protection to families earning less than Rs 72,000 per annum. Its execution entrusted to three TPAs: Vital Healthcare, MD India, and Medi Assist India.  The program was recently extended to cover marginalized populations such as differently abled persons, refugees from Sri Lanka, widow pensioners and old age pensioners.  It also provides treatment for highly technical procedures. Follow up procedures are also covered by the scheme in case of certain major procedures, surgeries and treatments.  Health camps are conducted once a month to identify and register eligible patients. Quality assurance measures include medical audits, standardized procedures, periodic review of medical/technical guidelines by experts and constant vigilance to prevent money collection from the beneficiaries. Outsourcing of housekeeping services:  Because of poor sanitation services in government hospitals, housekeeping services were outsourced to a competitively selected firm for four district government hospitals for two years.  The exercise resulted in an improvement of physical cleanliness, better safety and crowd regulations, and a rise in the satisfaction levels among patients and providers, as well resulted in cost effectiveness for TNHSP. 76  Key gaps in the pilot included variations in compliance among the facilities, unplanned allocation of human resources, weak procedures used by the agency and improper placement of security staff.  Based on overall positive results obtained from the pilot, TNHSP outsourced housekeeping services for 48 hospitals. The scaled-up program included carrying out of quality measures such as electronic reporting, biometric attendances for housekeeping personnel, appointment of nodal officer and regular training of housekeeping personnel and supervisors.  Weekly reviews were held of the vendor where compliance issues were sorted out. TNHSP has derived following results from this outsourcing experience: cleaner hospitals, cleaner toilets, proper biomedical waste management, proper parking of vehicles within hospital premises, zero theft incidences involving hospital goods, and improvement in the aesthetic appearance of the hospitals.  Following issues have emerged: frequent attrition of housekeeping personnel, carrying out personal work of providers during hospital duty hours, insufficient use of chemicals, lesser use of modern equipment, and variation of wages among the districts. These issues are being sorted out on an ongoing basis with the contractor. There were three additional presentations under the PPP:  Leveraging PPP for technology & innovations (emergency ambulance services)  Free hearse service  Impact of STEMI and need for upscaling Tribal health  Secretary, Nilgiris Adivasi Welfare Association (NAWA) described the range of programs available to tribal populations, including mobile outreach and the Sickle Cell Anemia Interventions.  Discussed the innovative use of retired health workers given difficulties in recruiting health personnel to hill areas.  Concluded that PPPs in partnership through NGOs in Tribal Health is cost effective and result oriented. Non Communicable Diseases Scaling up of Non-communicable diseases intervention program:  A large scale program covering four NCDs (hypertension, diabetes, cervical cancer and breast cancer) throughout the state was first of its kind in the entire country.  Key lessons learnt from the NCD pilots that informed the scaled up program were: (a) ensure dedicated human resources, (b) ensure uninterrupted supply of reagents and drugs with additional funding, (c) ensure maintenance of equipment, (d) carry out periodic reorientation of staffs, (e) ensure follow up of patients with suspected cancers on screening, and (f) improve the data quality and analysis and corrections of reports at the district level.  NCD program was carried out in coordination with various departments including the educational department, rural development department, labor department, ESI and municipal administration and corporation. 77  Self-help group women were sensitized on NCDs and were encouraged to go for screening in collaboration with the rural development department.  Major challenges during implementation were: (1) Human resources and capacity building issues, (2) Structural issues - Identifying space for conducting procedures and privacy for women, (3) Data issues - HMIS issues in PHCs, (4) Procedural/protocol issues, (5) Social issues, (6) Budget issues, (7) Procurement cum logistic issues, (8) Administrative issues, (9) Integration of levels of health care, and (10) Follow up issues.  Recommendations:  Outsourcing of Human resources to overcome the attrition of NCD staffs.  Periodic training for addressing knowledge gaps and skills.  Use of the Chief Minister’s Insurance Scheme for diagnostic and treatment services and conducting outreach programs.  Strengthening IEC activities and sensitizing self-help groups.  Frequent meetings between NCD team and TCS, inspections and video conferencing.  Display boards for beneficiaries to inform the services available.  Comprehensive exit plan to sustain the program through (NHM- NPCDCS). Implementation challenges and resolution in the NCD Program (TNHSP): Experience from concurrent evaluation  Major challenges (input indicators): staff nurses were posted in other departments, inadequate lab technicians, delay in procurement of equipment, not prescribing the available drugs for 30 days due to fear of stock-outs, challenges in implementing HMIS due to lack of computers in NCD clinics etc.  Major challenges (process indicators): statistics were shown based on the total number of people screened for hypertension and diabetes in phase 1 districts of Tamil Nadu from October 2012 to September 2013 and the average follow up visits of those screened positive for hypertension and diabetes from October 2013 to August 2014  Major challenges (cancer screening): inadequate trained nurses, high false negatives, lack of involvement of health workers in follow ups.  Other challenges (data usage): poor quality data generated from the facilities, time delay in receiving the data from all facilities, poor adherence to registers/ reporting formats.  Challenges (patients’ perspective): long waiting time in the facility, frequent visits for drugs and low awareness regarding the need for long term treatment.  Despite these challenges, the overall program was satisfactory because of strong political and administrative will, dedicated program managers at state/ district level, rolling out NCD program in all institutions across the state, uninterrupted and good quality drugs, NCD awareness messages reached the remote areas with the help of dedicated NCD nurses and highly motivated doctors.  Recommendations:  Train staff nurses to collect/analyze the lab samples.  Ensure availability of adequate stock of NCD drugs.  Access to computers to NCD nurses in the PHCs.  Train doctors to stick on to the protocols to improve adherence.  Purchase of consumables at the local level.  Incentives for the VHN who take women to hospitals for diagnostic work up. 78  Weekly report, monthly data analysis and shared state level summary to ensure good quality data. Ensuring patient adherence in hypertension and diabetes  Many people have been successfully screened for Hypertension, Diabetes, Cancer cervix and Cancer breast and the present approach has reached a point of saturation.  There is a need to reach the public through the private sector and missing follow ups at present is a major challenge (missed monthly follow ups, non-adherence of daily dose drugs, unawareness of the monthly, half-yearly and annual checkups, difficulty in identifying the beneficiary and in data/ tracking). Impact of non-adherence (extent to which a person’s behavior- taking medication, following a diet or making healthy lifestyle changes does not correspond with agreed upon recommendations from a health care provider) leads to significant treatment failures, costly second line management, increase in cardio vascular hospitalization and increased CVD mortality. Adherence can be improved by a SIMPLE strategy -- Simplify the regimen, Impart knowledge, Modify patient’s beliefs and behavior, Provide communication and Trust, Leave the bias and evaluate the adherence.  Under the current NCD program, patients are tracked by the NCD staff nurses, the positive individuals’ list is shared with the Village Health Nurse for tracking them in the field, and online tracking is also done with the help of HMIS. Incentives are provided to the NCD staff and VHNs for tracking patients. Training is given for the staff/ Medical Officers on how to use the HMIS platform and on updating the entries  Recommendations:  Need for daily mobile based alerts/SMS.  Train the patients as the “front line workers” and create a patient support group.  Electronic tracking through mobile apps.  Involvement of private sector and collaboration with health related sectors like nutrition, education, food safety, local bodies etc. 79 Implementation Completion Review Workshop with Stakeholders, August 28-29, 2015, Chennai - List of Invitees Name Title, Agency Mr. Bhaskar Dasgupta Director (MI Division), Department of Economic Affairs, Ministry of Finance, GOI Dr. Manivannan Deputy Drug Controller of India, Central Drug Standards Control Organization, South zone, GoI Dr. C. Vijayabaskar Hon’ble Minister for Health, GoTN Dr. Girija Vaidyanathan, Commissioner for Land Acquisition, Former Health Secretary I.A.S. & Mission Director, NHM, GoTN Mr. K. Shanmugam, I.A.S. Principal Secretary, Finance, GoTN Dr. J. Radhakrishnan, I.A.S. Health Secretary, GoTN Mr. P.W.C. Davidar, I.A.S. P & AR Secretary, Former Project Director, TNHSP. Dr. S. Vijayakumar, I.A.S. Secretary, Animal Husbandry, Dairy and Fisheries, Former Project Director, TNHSP. Mr. Pankaj Kumar Bansal, Managing Director, Chennai Metro Rail Limited, Chennai, I.A.S. Former Project Director, TNHSP. Mr. M. S. Shanmugam, I.A.S. Joint Secretary and Additional Secretary to Government, Industries Department, Former Project Director, TNHSP. Dr. K. Elangovan, I.A.S. Secretary to Government, Health and Family Welfare Department, Government of Kerala Dr. Himanshu Bhushan Director & Head, PHA Division, NHSRC, Delhi. Mr. Prasanth Subrahamanian Sr. Consultant, PHA Division, NHSRC, Delhi. Dr. Bontha V Babu Senior Scientist, ICMR Dr. Harsh Sharma Additional Project Director, UP Health Systems Project Dr. B. K. Verma Assistant Director, UP Health Systems Project Dr. Thiru. S. Ramakrishanan Advisor, NISG , Former Director General, C DAC Dr. V. R.Muraleedharan Professor, Department of Humanities and Social Sciences. IIT Madras Dr. N. Devadasan President, IPHI, Bengaluru Director and officials of Medical and Rural Health Services Director of Medical Education and key Officials Director and officials Public Health and Preventive Medicine National Institute of Epidemiology (ICMR), Chennai Vice-Chancellor, Tamil Nadu Dr. M.G.R. Medical University Deans of Medical colleges and Hospitals (20) Joint Directors of Health Services of all Districts (31) Deputy Directors, Health Services -all 42 Health Unit Districts District Project Management Coordinators of all Districts (31) Director of Drug Control and key Officials Commissioner of Indian Medicine and his key Officials Officials from Director of Family Welfare Officials from National Rural Health Mission, Tamil Nadu Officials from Director of Medical and Rural Health Services Officials from Tamil Nadu Medical Services Corporation Ltd. Officials from Tamil Nadu AIDS Control Society Officials from Anna Institute of Management Professor of Medicine, Poison Treatment Centre, Madras Medical College Professor of Cardiology, Stanley Medical College, Chennai 80 Professors from Institute of Obstetrics & Gynaecology, KG Hospital, RSRM Hospital and Institute of Child Health. National Institute of Epidemiology (ICMR), Chennai Dr.Ajith Mullasari, Director, Dept. of Cardiology, Madras Medical Mission Officials from Public Works Department Officials from Electronics Corporation of Tamil Nadu Principals of Public Heath Training Centers (2) Nodal Officers of Regional Training Centers for ICWM (9) M/s. Krystal Integrated Services Private Ltd, Mumbai Officials from Tata Consultancy Services Officials from Accenture Services Private Ltd. Tribal Development NGOs (4) 108 Emergency Management Services – EMRI Officials Indian Red Cross Society (IRCS), Tamil Nadu State Branch Officials from United India Insurance Company and Third Party Administrators, M/s Medi India, Medi Assist and Vidal Officials from Tamil Nadu Corporation for Women Development Officials from Sarva Shiksha Abhiyan Officials from Bharat Sanchar Nigam Limited (BSNL) Officials from Cancer Institute, Adyar Officials from Tamil Nadu Health Systems Project (TNHSP) Former Officials of TNHSP including Finance Officers who worked during the various phases of the Project spanning planning stage, initial implementation period, launching of pilot schemes, scaling up & so on. Consultants hired for conducting evaluation studies of various project activities during the entire Project period Certain Vendors who have supplied equipment and electronic items Hospital Superintendents, Chief Medical Officers, Nodal Officers & Assistant Nodal Officers for Accreditation from Government Hospitals Tambaram, Cuddalore, Hosur, Erode, Manaparai, & Aruppukottai Ms. Sai Subashri Raghavan Solidarity and Action Against The HIV Infection in India Dr. Varun Goyal Solidarity and Action Against The HIV Infection in India Dr. Bushra Binte Alam Task Team Leader, TNHSP, World Bank Dr. Preeti Kudesia Former Task Team Leader, TNHSP, World Bank Dr. Ramesh Govindaraj Lead Health Specialist, World Bank Ms. Sangeeta Carol Pinto Operations Officer, World Bank Mr. Atin Rastogi Procurement Specialist, World Bank Mr. Rohit Gawri I.T. Specialist, World Bank Mr. Owen Smith Senior Economist, World Bank Ms. Shreelata Rao Sheshadri Social Development Consultant, World Bank 81 Annex 7. Summary of Borrower's ICR Project Implementation The project met its development objectives as is evidenced in its achieving all agreed results from the original and additional financing period. Component I Increasing Access to and Utilization of Health Services 1. Subcomponent 1 (Reduction of Maternal and neonatal mortality) increased the number of complicated maternal admissions, number of ultra-sonograms, number of blood transfusions, number of night caesareans and reduced referral outs from hospitals. The MMR declined from 111 to 79 during the period 2004 to 2013 and the IMR from 37 to 21 during the period 2005 to 2013. 2. Subcomponent 2 (Improving Tribal Health) increased access to health care for the Tribal population through the provision of mobile outreach health services, sickle cell anemia intervention program, bed grant scheme for NGO-run hospitals in Tribal areas and Tribal patients’ counselors in health facilities. 3. Subcomponent 3 (facilitating use of hospitals by the poor and the disadvantaged) posted a total of 492 Patient Counselors in CEmONCs (Comprehensive emergency Obstetric and Newborn Care) and hospitals located in the Tsunami affected districts. The counselors guided patients and counseled them on preventive and promotive health behaviors. However, their services were discontinued from October 2011. Ambulance Services with 700+ ambulances were established to provide emergency transportation in the state. A fleet of 63 mortuary vans provided free hearse services in state. Housekeeping services in 48 large government hospitals and regional diagnostic laboratories in five district headquarters hospitals were outsourced as a PPP model. Component II Developing Effective models to Combat Non Communicable Diseases and Accidents 4. Subcomponent 1 (Health Promotion). Worksite, school based and community based interventions were rolled out to promote healthy lifestyles in support of the NCD program. More than 50% of the eligible state population has been sensitized on risk factors of NCDs and counseled on healthy life styles. The behavior change communication activities resulted in an increased number of people accessing the health facilities as evidenced from the end line survey reports for CEmONCs, Tribal Health and NCD Programs. 5. Subcomponent 2 Pilot testing and state-wide scale up of Non Communicable Diseases (NCD) control directly benefitted an increased number of persons who were screened for hypertension, diabetes, and cancers of the breast and cervix. All screened positive patients were offered treatment preventing complications like stroke, myocardial infarction, kidney failures, full blown cancers etc. 6. Subcomponent 3 was dropped to prevent duplication of interventions carried out by the World Bank funded road sector project in Tamil Nadu. A Helmet usage survey in ten major cities was conducted for ten months. A workshop on Road Traffic Accidents Prevention and trauma care was conducted in 2006. Dedicated ‘Poison Treatment Centers’ were set up in 66 secondary care hospitals, in addition to such centers in all medical College Hospitals. 82 Component III Building Capacity for Oversight and Management of the Health System 7. Subcomponent 1 Monitoring and Evaluation developed and successfully deployed a comprehensive Health Management Information System across the entire public health system in the state to facilitate management of hospital functions and public health management. Staff were trained in the use of the system, and a helpdesk was set up to provide both backend and user support. 8. Subcomponent 2 (Improving quality of care) A set of quality care indicators were introduced and health care providers trained on its use. Monthly reports were collected and feedback to the hospitals was provided after analysis, this resulted in a marked improvement in the performance and quality of activities in the hospitals. Quality was further improved by institutionalizing quality improvement circles in project facilities. Several protocols, guidelines and capacity strengthening programs were initiated under the project. Twelve large government hospitals secured accreditation from the National Board of Accreditation for Hospitals. 9. Subcomponent 3 (Health care Waste Management) Government of Tamil Nadu developed and implemented an integrated infection control and waste management plan with operational procedures, standardized protocols and training modules to institutionalize a comprehensive Infection Control and Waste Management system in all Government Health Institutions in Tamil Nadu. 10. Subcomponent 4. Strategy Development and Implementation consists of (i) Establishing a Strategic Planning Unit. The Strategic Planning Cell was established within the Project Management Unit as a think tank for the project and to undertake studies and policy research for improving the efficiency and effectiveness of the Health Systems. (ii) Establishing Public-Private Partnership (PPP) wing for fruitful partnerships with all non-governmental stakeholders in Health. (iii) Conducting a Health Insurance Pilot. The Project Implementation Plan had proposed a pilot community based Health insurance scheme. However the Government of Tamil Nadu implemented a Health insurance scheme with their own budget from 2009. Hence, while financing for this activity was no longer supported by project, administration of the scheme was done by TNHSP. (iv) Strengthening Project Management. A four tier management structure was created with (a) State Empowered Committee; (b) Project Steering Committee; (c) Project Management Unit and (d) District Project Management Unit. Component IV Maximizing the Efficiency of the Public Sector to deliver essential services 11. Subcomponent 1 Rationalization of Secondary care facilities was achieved by supporting works at 225 project facilities, and CEmONCs at eight Government Medical Colleges. Necessary infrastructure and equipment was also provided to all project hospitals based on detailed facility surveys and agreed criteria. 12. Subcomponent 2 Rationalizing of equipment was achieved with the provision of equipment and establishment of systems to maintain and manage these through a cadre of bio-medical engineers. This enhanced the provision of services, reduced referral outs and resulted in improved patient satisfaction. 83 13. Subcomponent 3 Human Resource Planning & Development involved training as the core activity. Training for specialists and paramedical staff for CEmONCs, tribal counselors, specialists and paramedical staff for deployment of NCD screening and treatment program, all clinical and administrative staff in the public health system in Health Management Information Systems, all staff in public health institutions in the state on Infection Control and Waste Management, staff of project supported facilities in Quality of Care; bio-medical engineers in equipment maintenance and management; as well as Human Resource Development was undertaken in the project. Training modules were developed and feedback obtained from the trainees on the quality and usefulness of the program. This was the first time in the history of Health Department that such a massive training program on diverse subjects to improve health care service delivery was undertaken. This improved the performance of the health care providers. 14. Subcomponent 4 Enhancing Management of Public Facilities. The project successfully improved the management of project hospitals through (a) provision of hospital management training to doctors, ministerial staff and nurses (b) enhancing financial powers of the chief medical officers for condemnation (c) reviewing monthly performance reports from HMIS for all project hospitals (d) grading of hospitals into A,B,C & D categories based on performance and providing support to C & D categories for improving performance (e) periodic inspections and reviews of the hospitals (f) medical and prescription audits; and (g) computerized inventory management of stores. 84 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders NA 85 Annex 9. Details on NSS 2004 and NSS 2014 Data on Access and utilization of health services by poorest 40% and scheduled tribe (ST) populations in Tamil Nadu The important project goal of improving access and utilization of health services by poor, disadvantaged and tribal groups was measured as changes among the (a) poorest 40% of the population (bottom two quintiles) and (b) Scheduled Tribes in: i. Proportion of population reporting ailment in last 15 days ii. Percentage of those reporting illness accessing any form of care iii. Number of hospitalization cases per 1000 in (a) the private sector and (b) public sector. Data Source and Sample Data from two nationally representative surveys conducted by the National Sample Survey Organization (NSSO) is used to understand the access and utilization of health services by poor and tribal groups:  Survey on Morbidity and Health care, NSS 60th round (January - June 2004)  Survey on Social Consumption: Health, NSS 71st round (January - June 2014) Table 1: Description of sample size for Tamil Nadu, NSS 2004 and 2014 Sample persons Sample persons Sample households Sample persons hospitalized in last 365 reporting any ailment days in last 15 days NSS 2004 Total ST Total ST Total ST Total ST Rural 2540 63 10348 247 1090 22 1100 5 Urban 2599 15 10946 56 1104 5 1255 1 All 5139 78 21294 303 2194 27 2355 6 NSS 2014 Total ST Total ST Total ST Total ST Rural 1960 45 8237 197 1604 31 1288 19 Urban 1957 10 7853 43 1588 5 1657 8 All 3917 55 16090 240 3192 36 2945 27 Results Indicator 1: Proportion of persons reporting ailment in last 15 days Table 2 reports the proportion of persons reporting an ailment24, measured as the number of living persons reporting ailment (per 1000 persons) during 15-day reference period by different background characteristics for 2004 and 2014. For the State as a whole, in 2004, 9.4 percent individuals have reported any ailment during the reference period of last 15 days and this has increased to 16.5 percent in 2014. 24 Due to the change in coverage and difference in concepts and definitions in some important parameters in the two rounds, the results of NSS 71st round are not strictly comparable with the results of NSS 60th round. This is applicable to the indicator of persons reporting ailment in last 15 days (see Annexure 1 for details). 86 Table 2: Reporting of any ailment in last 15 days by social group and wealth quintile in Tamil Nadu, NSS 2004 and 2014 NSS 2004 NSS 2014 Background characteristics % ailing % ailing Place of residence Rural 9.5 14.6 Urban 9.6 18.4 Social group Scheduled Tribe (ST) 0.8 10.3 Scheduled Caste (SC) 9.1 13.1 Other backward classes (OBC) 10.0 17.3 Others 8.6 30.7 MPCE Quintiles Lowest 8.2 12.5 Second 8.8 14.0 Third 8.0 15.5 Fourth 12.1 18.3 Highest 10.9 22.8 MPCE group Lower 40% 8.5 13.2 Upper 60% 10.2 18.7 All 9.5 16.5 Note: MPCE – monthly per capita consumption expenditure information as available in the NSS 2004 and NSS 2014 surveys. The MPCE quintiles are constructed using the MPCE distribution of all households of the State (rural + urban) as reference distribution. In 2004, around 0.8 per cent of ST persons have reported any ailment and in 2014 this has increased to 10.3 per cent. It may be noted that the sample size of ST persons (see Table 1) is smaller than other groups because of its relatively low share of 1.1 percent in the State’s total population (as per Census of India, 2011). The reporting of ailments is noted to vary across quintiles of monthly per capita expenditure (MPCE). In 2004, about 8.5 per cent of persons in the bottom two MPCE quintiles (poor 40 %) reported of any ailment during the last 15 days whereas this proportion has increased to 13.2 per cent in 2014. However, the reporting of ailments is higher among the richer 60% individuals in both periods. Indicator 2: Percentage of spells of ailment treated on medical advice in last 15 days Table 3 reports the percentage of spells of ailment treated on medical advice during 15-day reference period by different background characteristics for 2004 and 2014. For the State, in 2004, 81.3 percent spells of ailment reported during the reference period of last 15 days were treated on medical advice. This proportion has increased to 97.3 percent in 2014 with considerable narrowing of rural-urban differential. 87 Table 3: Percentage of spells of ailment treated on medical advice by social group and wealth quintile in Tamil Nadu, NSS 2004 and 2014 NSS 2004 NSS 2014 Background characteristics % treated % treated Place of residence Rural 78.4 96.5 Urban 86.6 98.0 Social group Scheduled Tribe (ST) 96.2 93.3 Scheduled Caste (SC) 71.4 96.9 Other backward classes (OBC) 83.8 97.4 Others 90.8 100.0 MPCE Quintiles Lowest 76.1 98.5 Second 75.9 96.6 Third 79.5 96.1 Fourth 81.6 99.2 Highest 90.4 97.0 MPCE group Lower 40% 76.0 97.5 Upper 60% 84.2 97.3 All 81.3 97.3 Note: MPCE – monthly per capita consumption expenditure information as available in the NSS 2004 and NSS 2014 surveys. The MPCE quintiles are constructed using the MPCE distribution of all households of the State (rural + urban) as reference distribution. In 2004, 96 per cent of spells of ailment among ST persons are reportedly treated, in 2014 it was 93.3 per cent. The proportion of ailments treated during 2014 is lower among STs than the estimates for other social groups. It may be noted that the proportion of ailments treated among STs in 2004 is estimated to be much higher than other social groups but this may be affected due to small sample (see Table 1). In 2004, there was a clear income-gradient in treatment of ailments with considerable disadvantages for poor individuals. However, the estimates for 2014 reveal significant reduction in rich-poor gap in treatment seeking for reports of ailment. In 2004 about 76 per cent cases of ailments among the poorer 40% individuals were treated on medical advice and this proportion has significantly increased to 97.5 per cent in 2014. The bridging of gap in treatment seeking for ailments both across social groups and across income class emerges as a noteworthy feature of the Tamil Nadu health system. Indicator 3a: Cases of hospitalization per 1000 persons during the last 365 days Table 4 reports the number of hospitalization cases per 1000 persons during the 365-day reference period by different background characteristics for 2004 and 2014. For the State, in 88 2004, 41.9 cases of hospitalization per 1000 persons is reported during the reference period of last 365 days. This proportion has increased to 56.2 per 1000 persons in 2014 but there is an increasing rural-urban differential in hospitalization cases. Table 4: Cases of hospitalized per 1000 persons during the last 365 days by social group and wealth quintile in Tamil Nadu, NSS 2004 and 2014 NSS 2004 NSS 2014 Background characteristics Cases per 1000 Cases per 1000 Place of residence Rural 42.0 53.7 Urban 41.8 58.8 Social group Scheduled Tribe (ST) 13.1 15.5 Scheduled Caste (SC) 39.8 55.7 Other backward classes (OBC) 43.6 57.4 Others 39.0 69.5 MPCE Quintiles Lowest 29.1 43.2 Second 37.8 40.6 Third 46.4 57.1 Fourth 48.5 65.6 Highest 48.1 76.9 MPCE group Lower 40% 33.4 41.9 Upper 60% 47.6 66.0 All 41.9 56.2 Note: MPCE – monthly per capita consumption expenditure information as available in the NSS 2004 and NSS 2014 surveys. The MPCE quintiles are constructed using the MPCE distribution of all households of the State (rural + urban) as reference distribution. In 2004, hospitalization among ST population was estimated to be 13.1 per 1000 persons. This is significantly lower than other social groups including the SC population. Importantly, the 2014 survey finds only a small increase in the hospitalization cases among the ST population and it is estimated to be 15.5 cases of hospitalization per 1000 persons. However, there is significant increase in utilization of inpatient care among other social groups. Similar to all treatment-seeking for ailments, it is noted that there is a significant income gradient in utilization of hospital-based care and the rich-poor gap has increased between the two survey periods. In 2004 about 33.4 hospitalization cases per 1000 persons were reported among the poorest 40% population and 47.6 among the richer 60%. In 2014 hospitalization cases among poor 40% have increased to 41.9 per 1000 whereas the same has increased to 66.0 hospitalization cases per 1000 among the richer 60%. Clearly, the absolute differential among the rich and poor in hospitalization cases has widened between 2004 and 2014. 89 Indicator 3b: distribution of hospitalization cases by type of hospital (public or private) Table 5 presents the distribution of hospitalization cases by type of hospital (public and private sector) and by different background characteristics for 2004 and 2014. For the State, in 2004, 39.5 per cent of the hospitalization cases were in public sector and 60.5 per cent cases were in private hospitals. The proportion of hospitalization in private sector has increased in recent years. In 2014, the share of public hospitals in total hospitalization is estimated to be 34.6 per cent whereas the share of private hospitals has increased to 65.4 per cent. Table 5: Per 1000 distribution of hospitalization cases during the last 365 days by type of hospital and by social group and wealth quintile in Tamil Nadu, NSS 2004 and 2014 NSS 2004 NSS 2014 Background characteristics Public Private Total Public Private Total Place of residence Rural 40.8 59.2 100.0 40.4 59.6 100.0 Urban 37.2 62.8 100.0 29.3 70.7 100.0 Social group Scheduled Tribe (ST) 54.7 45.3 100.0 54.5 45.5 100.0 Scheduled Caste (SC) 60.3 39.7 100.0 54.8 45.2 100.0 Other backward classes (OBC) 33.8 66.2 100.0 29.2 70.8 100.0 Others 21.5 78.5 100.0 10.0 90.0 100.0 MPCE Quintiles Lowest 48.9 51.1 100.0 53.7 46.3 100.0 Second 51.9 48.1 100.0 54.2 45.8 100.0 Third 43.5 56.5 100.0 34.8 65.2 100.0 Fourth 37.9 62.1 100.0 31.9 68.1 100.0 Highest 21.8 78.2 100.0 15.0 85.0 100.0 MPCE group Lower 40% 50.6 49.4 100.0 53.9 46.1 100.0 Upper 60% 34.4 65.6 100.0 26.3 73.7 100.0 All 39.5 60.5 100.0 34.6 65.4 100.0 Note: MPCE – monthly per capita consumption expenditure information as available in the NSS 2004 and NSS 2014 surveys. The MPCE quintiles are constructed using the MPCE distribution of all households of the State (rural + urban) as reference distribution. There is a significant difference in type of hospital use by social groups. During both the survey years 2004 and 2014 the ST population have reported relatively higher use of public hospitals (55 percent) than private hospitals (45 percent). Moreover, the share of public and private sector in hospitalization has not changed for the ST population whereas there is an increase in utilization of private hospital care among OBCs and other non-SC/ST groups. Use of type of hospital for inpatient care is associated with economic status of the individuals. In 2004, among the poor 40% population, about 51 per cent of the hospitalization cases was in public sector whereas in 2014 this proportion has increased to 54 per cent. However, in case of 90 richer 60 per cent population, the use of private sector has increased from 66 per cent in 2004 to about 74 per cent in 2014. This has also led to a reduced share of public sector in total hospitalization among the richer sections of the population. Additional Details In 2004, disabilities were included as ailments, but in 2014, pre-existing disabilities were not included: “[D]ue to the change in coverage and difference in concepts and definitions in respect of some important parameters followed in the two rounds, the results of NSS 71st round are not strictly comparable with the results of NSS 60th round. While making any comparison, these differences may be taken into consideration. In the 60th round and earlier surveys on health, persons with disabilities were regarded as ailing persons. In this round, pre-existing disabilities were considered as chronic ailments provided they were under treatment for a month or more during the reference period, but otherwise were not recorded as ailments. Disabilities acquired during the reference period (that is, whose onset was within the reference period) were, however, recorded as ailments” (NSSO 2015: pp.2). The specific instructions for collection of information regarding ailment during last 15 days: NSS 2004 Schedule 25.0 (Instructions to Field Staff, Chapter 5 page 122) 5.4.11 Column 11: whether ailing anytime during last 15 days: For each member of the household, it will be enquired whether he/she suffered from any ailment anytime during last 15 days. Those who suffered from any ailment, code 1 will be recorded for them. Otherwise, code 2 will be recorded. It may be noted that some ailments may be treated (either as an inpatient of a hospital or otherwise) and some untreated - both the cases should be considered here. For detailed definition of ailments please see para 1.9.46 of Chapter One. It may be further noted that  a person under medication for an ailment during the reference period, whether he/she felt sick or not, will be treated as ailing;  cases of complications arising during pregnancy or after childbirth will be considered as ailment;  untreated injuries like cuts, burns, scald, bruise etc. of minor nature will not be covered, if the informant does not consider them to be severe enough. NSS 2014 Schedule 25.0 (Instructions to Field Staff, Vol.I: NSS 71st Round C-14) 3.4.12 Column 11: whether suffering from any chronic ailment (yes-1, no-2): To make entries in column 11, the following questions should be asked for each household member:  Has the member been experiencing symptoms – persisting for more than one month on the date of survey – indicating any problem caused by an ailment affecting any organ of the body? [Exclusions: (i) Minor skin ailments (ii) Cases of headache, body ache, and minor gastric discomfort after meals, even if of a long-standing nature, unless the patient insists that they cause restriction of his/her activity. (iii) Disabilities such as congenital blindness.] IF YES, then the member is suffering from a chronic ailment on the date of survey  enter 1 in col.11  Proceed to the next household member. 91 IF NOT,  Has the member been taking a course of treatment on medical advice for a period of one month or more and continuing as on the date of survey, aimed at alleviation of the symptoms of any ailment? (Such treatment may have resulted in non-appearance of symptoms that would otherwise have appeared, during a part of the last one month, or the entire month.) [No exclusions. Treatment of pre-existing disabilities included.] IF YES, then the member is suffering from a chronic ailment on the date of survey  enter 1 in col.11  Proceed to the next household member. OTHERWISE, enter code 2 in col.11  Proceed to the next household member. 3.4.12.1 A chronic ailment may affect the stomach, lungs, nervous system, circulation system, bones and joints, eye, ear, mouth or any other organ of the body. A list of symptoms associated with various types of diseases and their codes is given in Table 3.1 (page C-16) for better understanding and reference. This list is not, however, meant to be exhaustive. 3.4.13 Column 12: whether suffering from any other ailment any time during last 15 days (yes-1, no-2): For each member (irrespective of entry in col.11) it will be asked: During the last 15 days, did the member feel any problem relating to skin, head, eyes, ears, nose, throat, arms, hands, chest, heart, stomach, liver, kidney, legs, feet or any other organ of the body? If so, code 1 will be put in col.12, irrespective of how many such ailments the member has suffered from. Note that  For the purpose of col.12, chronic ailments will be excluded.  A disability (e.g. vision loss) whose onset was during the last 15 days will be covered.  Ailments include injuries as well as illness, and may be treated or untreated.  A person who took medical advice or was under medication on medical advice for an illness or injury at any time during the reference period, whether he/she felt sick or not, must be considered as ailing (an exception is medicines given as part of routine pre-natal or post-natal care in cases of normal pregnancy without complications).  Cases of complications arising during pregnancy or after childbirth will be considered as ailment.  Each case of childbirth will be considered as a special case of ‘ailment’ (of the mother) in this survey to facilitate collection of some important data on childbirth.  Untreated injuries like cuts, burns, scald, bruise etc. of minor nature (that is, not considered severe by the informant) will not be covered. Reference: NSSO (2015) Key indicators of social consumption in India: Health, NSS 71st round, National Sample Survey Office, Ministry of Statistics and Programme Implementation, Government of India, New Delhi. 92 Annex 10. Tamil Nadu Key Indicators – National Family Health Surveys 2015 and 2005 93 94 95 Annex 11. List of Supporting Documents  Government of Tamil Nadu, 2015. Policy Note 2015-16 (Demand no.19) Health and Family Welfare Department (comprehensive update on health and health services)  Government of Tamil Nadu, Health Policy, 2003  Project Appraisal document (PAD), Tamil Nadu Health Systems Project., November 17, 2004.  Development Credit Agreement (Tamil Nadu Health Systems Project) between India and International Development Association, January 5, 2005  Project Agreement (Tamil Nadu Health Systems Project) between International Development Association and State of Tamil Nadu. January 5, 2005.  Project Paper (PP) on Additional Financing, Tamil Nadu Health Systems Project, April 5, 2010  Financing Agreement (Additional Financing for Tamil Nadu Health Systems Project) between India and International Development Association, July 6, 2010  Project Agreement (Additional Financing for Tamil Nadu Health Systems Project) between International Development Association and State of Tamil Nadu. July 6, 2010.  Aide Memoires of all World Bank missions conducted under the project.  Implementation Status and Results (ISR) documents from the project.  Status Report on Project Activities from TNHSP for World Bank Missions under the project, May 2015 and August 2015  Revised Results Monitoring Framework (updated as of September 15, 2015) from TNHSP.  End Line Evaluation of Quality of Care, IPSOS for TNHSP, May 23,2015  End Line Evaluation of Patient Satisfaction Survey, IPSOS for TNHSP, July 31, 2015  End Line Study Report for Package A - Quality of Care, Synovate 2010 for TNHSP  End Line Study Report for Package A - Patient Satisfaction, Synovate 2010 for TNHSP  End Line Study Report for Package A –Health Care Waste Management, Synovate 2010 for TNHSP  Tamil Nadu Health Systems Project – A Milestone in Healthcare, DoHFW, May 2015  TNHSP, Program Implementation Plan for Additional financing, 2010-2013, May 2010  TNHSP, Summary of the Proceedings of the Implementation Completion Review Workshop with Stakeholders held on August 28-29, 2015, Chennai  TNHSP, Implementation Completion Results Report, January 31, 2016, PMU, DoHFW  World Bank’s Country Partnership Strategy for India, 2013-2017  World Bank’s Country Strategy for India, 2009-2012  World Bank’s Country Strategy for India, 2005-2008. September 15, 2004. Report no. 29374-IN.  World Bank’s Strategy for Health, Nutrition and Population Results, 2007  Government of India, Ministry of Health and Family Welfare, National Family Health Survey 4 2015-2016. State Fact Sheet, Tamil Nadu. International Institute for Population Sciences, Mumbai, 2016. 96 MAP INSERT MAP HERE AFTER APPROVAL BY SENIOR GLOBAL PRACTICE DIRECTOR AN ORIGINAL MAP OBTAINED FROM GSD MAP DESIGN UNIT SHOULD BE INSERTED MANUALLY IN HARD COPY BEFORE SENDING A FINAL ICR TO THE PRINT SHOP. NOTE: To obtain a map, please contact the GSD Map Design Unit (Ext. 31482) A minimum of a one week turnaround is required 97