Assessing Facility Capacity, Costs of Care, and Patient Perspectives HEALTH SERVICE PROVISION IN MADHYA PRADESH CCESS, OTTLENECKS, OSTS, AND QUITY A B C E PUBLIC HEALTH FOUNDATION OF INDIA INSTITUTE FOR HEALTH METRICS AND EVALUATION UNIVERSITY OF WASHINGTON UNITED NATIONS CHILDREN’S FUND
49
Embed
Assessing Facility A B Capacity, Costs of Care, CE OSTS ......Assessing Facility Capacity, Costs of Care, and Patient Perspectives HEALTH SERVICE PROVISION IN MADHYA PRADESH CCESS,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Assessing Facility Capacity, Costs of Care, and Patient Perspectives
HEALTH SERVICE PROVISION IN MADHYA PRADESH
CCESS,OTTLENECKS,OSTS, ANDQUITY
A B CE
PUBLIC HEALTH FOUNDATIONOF INDIA
INSTITUTE FOR HEALTH METRICS AND EVALUATIONUNIVERSITY OF WASHINGTON
UNITED NATIONS CHILDREN’S FUND
CCESS,OTTLENECKS,OSTS, ANDQUITY
A B CE
5 Acronyms6 Termsanddefinitions8 Executivesummary11 Introduction13 ABCEprojectdesign18 MainfindingsHealth facility profiles Facility capacity and characteristics Patient perspectives Efficiency and costs50 Conclusionsandpolicyimplications56 Annex
HEALTH SERVICE PROVISION IN MADHYA PRADESH
Table of Contents
Assessing Facility Capacity, Costs of Care, and Patient Perspectives
PUBLIC HEALTH FOUNDATIONOF INDIA
INSTITUTE FOR HEALTH METRICS AND EVALUATIONUNIVERSITY OF WASHINGTON
UNITED NATIONS CHILDREN’S FUND
32
About IHME
About Public Health Foundation of India Collaborations
About this report
The Public Health Foundation of India (PHFI) is a public-private initiative to build institutional capacity in India for strengthening training, research, and policy development for public health in India. PHFI adopts a broad, integrative approach to public health, tailoring its endeavors to Indian conditions and bearing relevance to countries facing similar challenges and concerns. PHFI engages with various dimensions of public health that encompass promotive, preventive, and therapeutic services, many of which are often lost sight of in policy planning as well as in popular understanding.
The Institute for Health Metrics and Evaluation (IHME) is an independent global health research center at the University of Washington that provides rigorous and comparable measurement of the world’s most important health problems and evaluates the strategies used to address them. IHME makes this information freely available so that policymakers have the evidence they need to make informed decisions about how to allocate resources to best improve population health.
This project has immensely benefitted from the key inputs and support from the Madhya Pradesh (MP) state government. Approvals and valuable support for this project were received from the Madhya Pradesh state government and district officials, which are gratefully acknowledged.
Assessing Facility Capacity, Costs of Care, and Patient Perspectives: Madhya Pradesh provides a comprehensive as-sessment of health facility performance in Madhya Pradesh, including facility capacity for service delivery, efficiency of service delivery, and patient perspectives on the service they received. Findings presented in this report were produced through the ABCE project in Madhya Pradesh, which aims to collate and generate the evidence base for improving the cost-effectiveness and equity of health systems. The ABCE project in Madhya Pradesh is funded by UNICEF through a grant from AusAID. The ABCE project in India is funded through the Disease Control Priorities Network (DCPN), which is a multi-year grant from the Bill & Melinda Gates Foundation to comprehensively estimate the costs and cost-effectiveness of a range of health interventions and delivery platforms.
About UNICEF
The United Nations Children’s Fund (UNICEF) works in 190 countries and territories to protect the rights of every child. UNICEF has spent 70 years working to improve the lives of children and their families. Defending children’s rights throughout their lives requires a global presence, aiming to produce results and understand their effects. UNICEF in India is fully committed to working with the Government of India to ensure that each child born in this country gets the best start in life, thrives, and develops to his or her full potential.
54
Acronyms Acknowledgments
We especially thank all of the health facilities and their staff in Madhya Pradesh, who generously gave of their time and facilitated the sharing of facility data that made this study possible. Our special thanks to the Government of Madhya Pradesh, especially the Principal Secretary, Health Commissioner, Mission Director and Deputy Director Child Health, Rajashree Bajaj. We are also most appreciative of patients of the facilities who participated in this work, as they too were giving of their time and were willing to share their experiences with the field research team.
At PHFI, we wish to thank Rakhi Dandona and Lalit Dandona, who served as the principal investigators for the ABCE project in India. We also wish to thank G. Anil Kumar for guidance with data collection, management, and analysis. The quantity and quality of the data collected for the ABCE project in India is a direct reflection of the dedication of the field team. We thank the India field coordination team, which included Md. Akbar, G. Mushtaq Ahmed, and S.P. Ramgopal. We also recognize and thank Venkata Srinivas, Amit Kumar, Simi Chacko, and Ranjana Kesarwani for data management and coordination with field teams.
At IHME, we wish to thank Christopher Murray and Emmanuela Gakidou, who served as the principal investigators. We also recognize and thank data analysts and Post-Bachelor Fellows at IHME: Roy Burstein, Alan Chen, Emily Dansereau, Katya Shackelford, Alexander Woldeab, Alexandra Wollum, and Nick Zyznieuski for managing survey programming, sur-vey updates, data transfer, and ongoing verification at IHME during fieldwork. We are grateful to others who contributed to the project: Michael Hanlon, Santosh Kumar, Herbie Duber, Kelsey Bannon, Aubrey Levine, and Nancy Fullman. Finally, we thank those at IHME who supported publication management, editorial support, writing, and design: Joan Williams, Adrienne Chew, and Michaela Loeffler.
At Bhopal field office of UNICEF, we wish to thank health specialists Gagan Gupta and Vandana Bhatia, who played a critical role in coordinating with state and district government through the entire duration of the project. We are also thankful to support from Syed Hubbe Ali and Rahul Bhadoria who supported coordination with district teams.
This report was drafted by Marielle Gagnier, Lauren Hashiguchi, and Nikhila Kalra of IHME and Rakhi Dandona and G. Anil Kumar of PHFI.
Funding for this research comes from the Bill & Melinda Gates Foundation under the Disease Control Priorities Network (DCPN), and from UNICEF, provided by AusAID.
ABCE Access, Bottlenecks, Costs, and EquityANC Antenatal careANM Auxiliary nurse midwifeAusAID Australian Agency for International DevelopmentCH Civil hospitalCHC Community health centreDCPN Disease Control Priorities NetworkDEA Data envelopment analysisDH District hospitalDOTS Directly observed therapy, short-courseIHME Institute for Health Metrics and EvaluationIPHS Indian Public Health StandardsMP Madhya PradeshNCD Non-communicable diseasesPHC Primary health centrePHFI Public Health Foundation of IndiaSFA Stochastic frontier analysisSHC Sub health centreSTI Sexually transmitted infectionUNICEF United Nations Children’s FundWHO World Health Organization
76
A B C E I N M A D H YA P R A D E S H
Table 1 defines the types of health facilities in Madhya Pradesh; this report will refer to facilities according to these definitions.
Table1 Health facility types in Madhya Pradesh1
1 Directorate General of Health Services, Ministry of Health & Family Welfare, and Government of India. Indian Public Health Standards (IPHS) Guidelines. New Delhi, India: Government of India, 2012.
T E R M S A N D D E F I N I T I O N S
Districthospital(DH)This type of facility is the secondary referral level for a given district. Its objective is to provide comprehensive secondary health care services to the district’s population. DHs are sized according to the size of the district population, so the number of beds varies from 75 to 500.
Civilhospital(CH)These facilities are sub-district/sub-divisional hospitals below the district and above the block level hospitals (CHC). As First Referral Units, they provide emergency obstetrics care and neonatal care. These facilities serve populations of 500,000 to 600,000 people and have a bed count varying between 31 and 100.
Communityhealthcentre(CHC)These facilities constitute the secondary level of health care and were designed to provide referral as well as specialist health care to the rural population. They act as the block-level health administrative unit and as the gate-keeper for referrals to higher-level facilities. Bed strength ranges up to 30 beds.
Primaryhealthcentre(PHC)These facilities provide rural health services. PHCs serve as referral units for primary health care from Sub-Centres and refers cases to CHCs and higher-order public hospitals. Depending on the needs of the region, PHCs may be upgraded to provide 24-hour emergency hospital care for a number of conditions. A typical PHC covers a population of 20,000 to 30,000 people and hosts about six beds.
Subhealthcentre(SHC)Along with PHCs, these facilities provide rural health care. SHCs typically provide outpatient care, which includes immunizations, and refer inpatient care and deliveries to higher-level facilities.
Health facility types in Madhya Pradesh
Terms and definitions
Definitions presented for key technical terms used in the report.
Constrainta factor that facilitates or hinders the provision of or access to health services. Constraints exist as both “supply-side,” or the capacity of a health facility to provide services, and “demand-side,” or patient-based factors that affect health-seeking behaviors (e.g., distance to the nearest health facility, perceived quality of care received from providers).
DataEnvelopmentAnalysis(DEA)an econometric analytic approach used to estimate the efficiency levels of health facilities.
Efficiencya measure that reflects the degree to which health facilities are maximizing the use of the resources available in producing services.
Facilitysamplingframethe list of health facilities from which the ABCE sample was drawn. This list was based on a 2012–2013 facility inventory published by the Madhya Pradesh state government.
Inpatientvisita visit in which a patient has been admitted to a facility. An inpatient visit generally involves at least one night spent at the facility, but the metric of a visit does not reflect the duration of stay.
Inputstangible items that are needed to provide health services, including facility infrastructure and utilities, medical supplies and equipment, and personnel.
Outpatientvisita visit at which a patient receives care at a facility without being admitted.
Outputsvolumes of services provided, patients seen, and procedures conducted, including outpatient and inpatient care, laboratory and diagnostic tests, and medications.
Platforma channel or mechanism by which health services are delivered.
StochasticFrontierAnalysis(SFA)an econometric analytic approach used to estimate the efficiency levels of health facilities.
9
E X E C U T I V E S U M M A R Y
W
Executive summary
ith the aim of establishing universal health coverage, India’s national and state gov-ernments have invested significantly in expanding and strengthening the public
health care sector. This has included a particular com-mitment to extending its reach to rural populations and reducing disparities in access to care for marginalized groups. However, in order to realize this goal, it is neces-sary for the country to critically consider the full range of factors that contribute to or hinder progress towards it.
Since its inception in 2011, the Access, Bottlenecks, Costs and Equity (ABCE) project has sought to compre-hensively identify what and how components of health service provision- access to services, bottlenecks in de-livery, costs of care, and equity in care received- affect health system performance in several countries. Through the ABCE project, multiple sources of data, including facility surveys and patient exit interviews, are linked together to provide a nuanced picture of how facili-ty-based factors (supply-side) and patient perspectives (demand-side) influence optimal service delivery.
Led by the Public Health Foundation of India (PHFI) and the Institute for Health Metrics and Evaluation (IHME), the ABCE project in Madhya Pradesh is uniquely posi-tioned to inform the evidence-base for understanding the country’s drivers of health care access and costs of care. Derived from a state representative sample of 203 facilities, the findings presented in this report provide governments, international agencies, and development partners alike with actionable information that can help identify areas of success and targets for improving health service provision.
The main topical areas covered in this report move from an assessment of facility-reported capacity for care, to quantifying the services actually provided by facilities and the efficiency with which they operate; tracking facil-ity expenditures and the costs associated with different types of service provision; and comparing patient per-spectives of the care they received across different types of facility. Further, we provide an in-depth examination
and comparison of facility-level outputs, efficiency, capac-ity and patient experiences. It is with this information that we strive to provide the most relevant and actionable in-formation for health system programming and resource allocation in Madhya Pradesh.
• Health facilities generally reported a high availability of a subset of key services. Services such as antenatal care, routine deliveries, general medicine, and lab ser-vices were widely available across facilities.
• Few facilities reported available services for non-com-municable diseases (NCDs). Just over half of district hospitals reported providing cardiology (53%), but few provided psychiatry (6%) or chemotherapy (6%), and availability decreased at lower levels of the health system.
• Basic medical equipment such as scales, stethoscopes, and blood pressure apparatus were widely available at all health facility levels, but laboratory equipment such as glucometers and blood chemistry analyzers were less readily available at lower facility levels. For example, over a quarter of community health centres and nearly half of primary health centres did not have glucometers. This shows limited capacity for testing in the health system, with particular implications for diag-nosing and treating NCDs.
• Gaps also emerged with regard to imaging equipment across health facilities. 40% of civil hospitals had a functional ultrasound, and CT scans were available in just 29% of district hospitals.
• A service capacity gap emerged for the majority of health facilities across several types of services. Many
facilities reported providing a given service but lacked full capacity to properly deliver it, for instance lacking functional equipment or medications. For example, while nearly all hospitals reported providing routine delivery care, only 29% of district hospitals and 7% of civil hospitals were fully equipped to do so. Discor-dances like these have substantial programmatic and policy implications for the health system in Madhya Pradesh, highlighting continued challenges in ensur-ing facilities have all the supplies they need to provide a full range of services.
• Functional electricity was available at all hospitals and community health centres, and the large major-ity (94%) of primary health centres. 43% of sub health centres had electricity, showing substantial improve-ment on figures from past government surveys.
• Access to piped water was generally high at hospi-tals (100%), though much lower at primary and sub health centres (33% and 9% respectively). Similarly, while there was universal availability of flush toilets at hospitals, they were available at 69% of sub health centres. These figures reflect investments into improv-ing physical infrastructure at health facilities, though some discrepancies remain between high and low-level facilities.
• 28% of primary health centres had access to a com-puter, and only 9% had access to a phone. Access to emergency vehicles was relatively high at district hospitals (94%), civil hospitals (73%) and community health centres (59%), but much lower at primary health centres (4%). Given that these types of facilities often play key referral functions, these findings have serious implications for coordinating the care and transporta-tion of patients.
• In general, hospitals reported that they staffed more nurses than doctors, and district hospitals in particular employed high proportions of non-medical personnel.
Lower-level facilities tended to employ more para- and non-medical staff than nurses or doctors.
• Staff numbers were concentrated at district hospi-tals. Civil hospitals had the second highest number of personnel, but this was a quarter of that at district hospitals. Health centres averaged between 2 and 28 staff. While some of this variation is a result of service provision and population size, this also demonstrates relative shortages in human resources for health.
• Between 2008 and 2012, the average number of out-patient visits remained stable. Civil hospitals had nearly triple the outpatient visits of community health centres, while primary health centres had ten times the visits of sub health centres.
• Inpatient visits increased for all facility types between 2008 and 2012.
• The number of immunization doses administered between 2008 and 2012 remained stable for all facility types.
• In generating estimates of facility-based efficiency, or the alignment of facility resources with the number of patients seen or services produced, we found a wide range of efficiency levels within and across facility types. The average efficiency score of district hospitals ranged from 73% to 88%, with a platform average of 78%. Civil hospitals were between 57% and 72% effi-cient. Community health centres were between 39% and 79% efficient. The range of efficiency scores for primary health centres was from 22% to 49%.
• If they operated at optimal efficiency, district hospitals could provide 42,684 additional outpatient visits with the same inputs (including physical capital and per-sonnel), while primary health centres could produce 10,576 additional outpatient visits.
• These efficiency scores indicate that there is consid-
8
1110
erable room for health facilities to expand service production given their resources existing resources. Future work on pin-pointing specific factors that heighten or hinder facility efficiency, and how effi-ciency is related to the quality of service provision, should be considered.
• Spending on personnel accounted for the majority of annual spending across facility types. Community health centres spent a slightly lower proportion of their total expenditures on personnel than other platforms, while the proportion of expenditure on medical sup-plies was highest at primary health centres.
• Most patients had travel times of less than 30 min-utes to a facility for care, and travel times were shorter for patients seeking care at lower-level facilities than higher-level ones; 42% of patients who went to district hospitals traveled more than 30 minutes, while none at sub-health centres traveled for that long.
• The large majority of patients waited less than 30 min-utes to receive care at all platforms, and nearly all patients seeking care at community health centres (94%) and primary health centres (97%) received care within 30 minutes. At district hospitals, 21% of patients waited more than 30 minutes to receive care.
• At hospitals, patients receiving care from nurses or auxiliary nurse midwives reported relatively higher lev-
els of satisfaction than those treated by doctors. The opposite tended to be true of community and primary health centres. Satisfaction with both doctors and nurses was highest at district hospitals, and lower at civil hospitals and health centres.
• Similar proportions of patients (52%-57%) were sat-isfied with the cleanliness at the facility they visited at all facility types-- except sub health centres, which received particularly low ratings (13% satisfied with cleanliness). Similar trends were observed for privacy.
• The vast majority of patients received all drugs that they were prescribed during their visits. Proportions of patients receiving all prescribed drugs ranged from 98-100% across all platforms.
With its multidimensional assessment of health ser-vice provision, findings from the ABCE project in Madhya Pradesh provide an in-depth examination of health facility capacity, costs of care, and how patients view their inter-actions with the health system. Madhya Pradesh’s health provision landscape was markedly heterogeneous, and will likely continue to evolve over time. This highlights the need for continuous and timely assessment of health service delivery, which is critical for identifying areas of successful implementation and quickly responding to service disparities or faltering performance. Expanded analyses would also allow for an even clearer picture of the trends and drivers of facility capacity, efficiencies, and costs of care. With regularly collected and analyzed data, capturing information from health facilities, recipi-ents of care, policymakers, and program managers can yield the evidence base to make informed decisions for achieving optimal health system performance and the eq-uitable provision of cost-effective interventions throughout Madhya Pradesh.
Introduction
The performance of a country’s health sys-tem ultimately shapes the health outcomes experienced by its population, influencing the ease or difficulty with which individuals
can seek care and facilities can address their needs. At a time when international aid is plateauing1 and the gov-ernment of India has prioritized expanding many health programs,2,3 identifying health system efficiencies and promoting the delivery of cost-effective interventions has become increasingly important.
Assessing health system performance is crucial to opti-mal policymaking and resource allocation; however, due to the multidimensionality of health system functions,4
comprehensive and detailed assessment seldom occurs. Rigorously measuring what factors are contributing to or hindering health system performance – access to services, bottlenecks in service delivery, costs of care, and equity in service provision throughout a country – provides crucial information for improving service delivery and popula-tion health outcomes.
The Access, Bottlenecks, Costs, and Equity (ABCE) project was launched globally in 2011 to address these gaps in information. In addition to India, the multi-pronged, multi-partner ABCE project has taken place in seven other countries (Bangladesh, Colombia, Ghana, Kenya, Lebanon, Uganda, and Zambia). In India, the ABCE project was undertaken in six states – Andhra Pradesh and Telangana, Gujarat, Madhya Pradesh, Odisha, and Tamil Nadu.
The ABCE project, with the goal of rigorously assessing the drivers of health service delivery across a range of set-tings and health systems, strives to answer these critical
1 Institute for Health Metrics and Evaluation (IHME). Financing Global Health 2015: Development assistance steady on the path to new Global Goals. Seattle, WA: IHME, 2016. 2 Planning Commission Government of India. Eleventh Five Year Plan (2007-12). New Delhi, India: Government of India, 2007. 3 Planning Commission Government of India. Twelfth Five Year Plan (2012-17). New Delhi, India: Government of India, 2012.4 Murray CJL, Frenk J. A Framework for Assessing the Performance of Health Systems. Bulletin of the World Health Organization. 2000; 78 (6): 717-731.
questions facing policymakers and health stakeholders in each country or state for public sector health care service delivery:• What health services are provided, and where
are they available?
• What are the bottlenecks in provision of these services?
• How much does it cost to produce health services?
• How efficient is provision of these health services?
Findings from each country’s ABCE work will pro-vide actionable data to inform their own policymaking processes and needs. Further, ongoing cross-country analyses will likely yield more global insights into health service delivery and costs of health care. These eight countries have been purposively selected for the overar-ching ABCE project as they capture the diversity of health system structures, composition of providers (public and private), and disease burden profiles. The ABCE project contributes to the global evidence base on the costs of and capacity for health service provision, aiming to de-velop data-driven and flexible policy tools that can be adapted to the particular demands of governments, de-velopment partners, and international agencies.
The Public Health Foundation of India (PHFI) and the Institute for Health Metrics and Evaluation (IHME) com-pose the core team for the ABCE project in India, and they received vital support and inputs from the state Ministry of Health and Family Welfare for data collection, analysis, and interpretation. The core team harnessed information from distinct but linkable sources of data, drawing from a state-representative sample of health facilities to cre-ate a large and fine-grained database of facility attributes, expenditure, and capacity, patient characteristics, and outcomes. By capturing the interactions between facility characteristics and patient perceptions of care, we have been able to piece together what factors drive or hinder optimal and equitable service provision in rigorous, da-ta-driven ways.
1312
A B C E I N M A D H YA P R A D E S H
We focus on the facility because health facilities are the main points through which most individuals interact with the health system or receive care. Understanding the ca-pacities and efficiencies within and across different types of public sector health facilities unveils the differences in health system performance at the level most critical to patients – the facility level. We believe this information is immensely valuable to governments and development partners, particularly for decisions on budget alloca-tions. By having data on what factors are related to high facility performance and improved health outcomes, pol-icymakers and development partners can then support evidence-driven proposals and fund the replication of these strategies at facilities throughout India.
The ABCE project in India has sought to generate the evidence base for improving the cost-effectiveness and equity of health service provision. In this report, we ex-amine facility capacity across platforms, as well as the efficiencies and costs associated with service provision for each type of facility. Based on patient exit interviews, we
consider the factors that affect patient perceptions of and experiences with state’s health system. By considering a range of factors that influence health service delivery, we have constructed a nuanced understanding of what helps and hinders the receipt of health services through facili-ties in the state of Madhya Pradesh.
The results discussed in this report are far from ex-haustive; rather, they align with identified priorities for health service provision and aim to answer questions about the costs of health care delivery in the respective state in India. This report provides an in-depth examina-tion of health facility capacity across different platforms, specifically covering topics on human resource capacity, facility-based infrastructure and equipment, health ser-vice availability, patient volume, facility-based efficiencies, costs associated with service provision, and demand-side factors of health service delivery as captured by patient exit interviews.
Table 2 defines the cornerstone concepts of the ABCE project: Access, Bottlenecks, Costs, and Equity.
AccessHealth services cannot benefit populations if they cannot be accessed; thus, measuring which elements are driving improved access to – or hindering contact with – health facilities is critical. Travel time to facilities, user fees, and cultural preferences are examples of factors that can affect access to health systems.
BottlenecksMere access to health facilities and the services they provide is not sufficient for the delivery of care to popula-tions. People who seek health services may experience supply-side limitations, such as medicine stockouts, that prevent the receipt of proper care upon arriving at a facility.
CostsHealth services cost can translate into very different financial burdens for consumers and providers of such care. Thus, the ABCE project measures these costs at several levels, quantifying what facilities spend to provide services.
EquityVarious factors influence how populations interact with a health system. The nature of these interactions either facilitates or obstructs access to health services. In addition to knowing the cost of scaling up a given set of services, it is necessary to understand costs of scale-up for specific populations and across population-related factors (e.g., distance to health facilities). The ABCE project aims to pinpoint which factors affect the access to and use of health services and to quantify how these factors manifest.
Access, Bottlenecks, Costs, and Equity
Table2 Access, Bottlenecks, Costs, and Equity
F
ABCE project design
or the ABCE project in India, we conducted primary data collection through a two-pronged approach:
1. A comprehensive facility survey administered to a representative sample of health facilities in select states in India (the ABCE Facility Survey).
2. Interviews with patients as they exited the sampled facilities.
Here, we provide an overview of the ABCE sur-vey design and primary data collection mechanisms. All ABCE survey instruments are available online at http://www.healthdata.org/dcpn/india.
ABCEFacilitySurveyThrough the ABCE Facility Survey, direct data collec-
tion was conducted from a state-representative sample of health service platforms and captured information on the following indicators for the five fiscal years (running from April to March of the following year) prior to the survey:
• Inputs: the availability of tangible items that are needed to provide health services, including in-frastructure and utilities, medical supplies and equipment, pharmaceuticals, personnel, and non-medical services.
• Finances: expenses incurred, including spending on infrastructure and administration, medical supplies and equipment, pharmaceuticals including vaccines, and personnel. Facility funding from different sources (e.g., central and state governments) and revenue from service provision were also captured.
• Outputs: volume of services and procedures pro-duced, including outpatient and inpatient care, emergency care, and laboratory and diagnostic tests.
• Supply-sideconstraintsandbottlenecks: factors that affected the ease or difficulty with which patients received services they sought, including bed availability, pharmaceutical availability and stockouts, cold-chain capacity, personnel availability, and service availability.
Table 3 provides more information on the specific indicators included in the ABCE Facility Survey.
1514
A B C E I N M A D H YA P R A D E S H A B C E P R OJ E C T D E S I G N
Figure2 Sampling strategy for health facilities in a district in the ABCE survey in India
Selected facilities are in blue; unselected facilities from the sampling frame are in grey.DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre; SHC: Sub health centre
SURVEY MODULE SURVEY CATEGORY KEY INDICATORS AND VARIABLES
Module 1:Facilityfinances andinputs
Inputs Input funding sources, managing authority and maintenance information
Availability and functionality of medical and non-medical equipment
Finances Salary/wages, benefits, and allowances
Total expenses for infrastructure and utilities; medical supplies and equipment; pharmaceuticals; administration and training; non-medical services, personnel (salaries and wages, benefits, allowances)
Performance and performance-based financing questions
Revenue User fees; total revenue and source
Personnel characteristics Total personnel by cadre
Funding sources of personnel
Health services provided and their staffing; administrative and support services and their staffing
Module 2:Facilitymanagementanddirectobservation
Facility management and infrastructure characteristics
Characteristics of patient rooms; electricity, water, and sanitation
Facility meeting characteristics
Guideline observation
Direct observation Latitude, longitude, and elevation of facility. Facility hours, characteristics, and location; waiting and examination room characteristics
Availability and functionality of medical furniture, equipment, and supplies
Inventory of procedures for sterilization, sharp items, and infectious waste
Inventory of personnel
Module 6:Facilityoutputs
Facility capacity Fund and vehicle availability for referral and emergency referral
General service provision Inpatient care and visits; outpatient care and visits; emergency visits; home or outreach visits
Laboratory and diagnostic tests
Module 7:Vaccines
Facility procedures for vaccine supply, delivery and disposal
Source from vaccine obtained
Personnel administering vaccine
Procedures to review adverse events
Disposal of vaccines
Vaccine availability, storage, and output
Stock availability and stockouts of vaccines and syringes
Types and functionality of storage equipment for vaccines
Temperature chart history; vaccine inventory and vaccine outputs; vaccine outreach and home visits
Vaccine sessions planned and held
Table3Modules included in the ABCE Facility Survey in India Figure1 Sampled districts in Madhya PradeshSampledesignA total of 17 districts in Madhya Pradesh were selected
for the ABCE survey (Figure 1). The districts were selected using three strata to maximize heterogeneity: proportion of full immunization in children aged 12–23 months as an indicator of preventive health services; proportion of safe delivery (institutional delivery or home delivery assisted by skilled person) as an indicator of acute health services; and proportion of urban population as an indicator of overall development. The districts were grouped as high and low for urbanization based on median value, and into three equal groups as high, medium, and low for the safe delivery and full immunization indicators. Sixteen districts were selected randomly from each of the various combi-nations of indicators, and in addition the capital district was selected purposively.
Within each sampled district, we then sampled pub-lic sector health facilities at all levels of services based on the structure of the state health system (Figure 2).
1716
A B C E I N M A D H YA P R A D E S H
Table5 Facility sample, by platform, for the ABCE project in Madhya Pradesh
FACILITY TYPE FINAL SAMPLE
Districthospital 17
Civilhospital 15
Communityhealthcentre 34
Primaryhealthcentre 69
Subhealthcentre 68
Totalhealthfacilities 203
Table4Types of questions included in the Patient Exit Interview Survey in India
SURVEY CATEGORY TYPES OF KEY QUESTIONS AND RESPONSE OPTIONS
Directobservationofpatient Sex of patient (and of patient’s attendant if surveyed)
Directinterviewwithpatient Demographic questions (e.g., age, level of education attained, caste)
Scaled-response satisfaction scores (e.g., satisfaction with medical doctor)
Open-ended questions for circumstances and reasons for facility visit, as well as visit characteristics (e.g., travel time to facility)
Reporting costs associated with facility visit (user fees, medications, transportation, tests, other), with an answer of “yes” prompting follow-up questions pertaining to amount
In each sampled district, one district hospital (DH); one civil hospital (CH, from a total of two or three) for each sampled DH; two community health centres (CHC, from a total of two to five) for each sampled CH; two primary health centres (PHC, from a total of two to four) for each sampled CHC; and one sub centre (SHC, from a total of one to four) for each sampled PHC were randomly se-lected for the study. Some of the sampled districts did not have a civil hospital and some districts had more than one; all available CHs in sampled districts were sampled.
PatientexitinterviewsurveyA fixed number patients or attendants of patients were
interviewed at each facility, based on the expected out-patient density for the platform. A target of 24 patients were interviewed at district hospitals, 16 at CH, 12 at CHC, 10 at PHC and five at SHC. Patient selection was based on a convenience sample. The main purpose of the Pa-tient Exit Interview Survey was to collect information on patient perceptions of the health services they received and other aspects of their facility visit (e.g., travel time to facility, costs incurred during the facility visit, and sat-isfaction with the health care provider). Table 4 provides more information on the specific indicators included in the exit survey. This information fed into quantifying the “demand-side” constraints to receiving care (as opposed to the facility-based, “supply-side” constraints and bottle-necks measured by the ABCE Facility Survey).
DatacollectionfortheABCE surveyinMPData collection took place from January to June 2014.
Prior to survey implementation, PHFI and the data-collec-tion agency hosted a two-week training workshop for 50 interviewers, where they received extensive training on the electronic data collection software (DatStat and Sur-veybe), the survey instruments, the MP health system’s organization, and interviewing techniques. Following this workshop, a one-week pilot of all survey instruments took place at health facilities. Ongoing training occurred on an as-needed basis throughout the course of data collection.
All collected data went through a thorough verification process between PHFI and IHME and the ABCE field team. Following data collection, the data were methodically cleaned and re-verified, and securely stored in databases hosted at PHFI and IHME.
A total of 203 health facilities participated in the ABCE project. Ten facilities were replaced (one CHC, four PHCs, and five SHCs) due to data being unavailable for the years considered; the reporting chain of the sampled facility being incorrect; or the facility being functional for less duration.
A B C E P R OJ E C T D E S I G N
1918
M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
Main findingsHealth facility profiles
The delivery of facility-based health ser-vices requires a complex combination of resources, ranging from personnel to phys-ical infrastructure, that vary in their relative
importance and cost to facilities. Determining what fac-tors support the provision of services at lower costs and higher levels of efficiency at health facilities is critical in-formation for policymakers to expand health system coverage and functions within constrained budgets.
Using the ABCE MP facility sample (Table 5), we analyzed five key drivers of health service provision at facilities:• Facility-based resources (e.g., human resources,
infrastructure and equipment, and pharmaceuticals), which are often referred to as facility inputs.
• Patient volumes and services provided at facilities (e.g., outpatient visits, inpatient bed-days), which are also known as facility outputs.
• Patient-reported experiences, capturing “demand-side” factors of health service delivery.
• Facility alignment of resources and service production, which reflects efficiency.
• Facility expenditures and production costs for service delivery.
These components build upon each other to create a comprehensive understanding of health facilities in Mad-hya Pradesh, highlighting areas of high performance and areas for improvement.
Facilitycapacityandcharacteristics
ServiceavailabilityAcross and within district hospitals, civil hospitals and
community health centres in MP (Table 6), several nota-ble findings emerged for facility-based health service provision. While fundamental services such as antenatal care, routine deliveries, general medicine, and pharmacy were nearly universally available, only a few district hos-pitals reported available services for non-communicable diseases, such as psychiatry and chemotherapy; only half reported cardiology. District hospitals reported a wide range of services such as blood banks, surgical services, and emergency obstetrics. Civil hospitals generally of-fered fewer services than district hospitals but reported high coverages of services like OBGYN services, ante-natal care, and immunizations. Less than one-quarter of community health centres reported providing STI/HIV treatment and less than a third pediatric medicine.
HumanresourcesforhealthA facility’s staff size and composition directly affect
the types of services it provides. In general, a greater availability of health workers is related to higher service utilization and better health outcomes.1 India has a severe shortage of qualified health workers, and the workforce is concentrated in urban areas.2 The public health system has a shortage of both medical and paramedical per-sonnel. The number of primary and community health centres without adequate staff is substantially higher if high health-worker absenteeism is taken into consider-ation.3 The Indian Government is aware of the additional requirements and shortages in the availability of health workers for the future. The National Rural Health Mission,
1 Rao KD, Bhatnagar A, Berman P. So many, yet few: Human resources for health in India. Human Resources for Health. 2012; 10(19). 2 Rao M, Rao KD, Kumar AK, Chatterjee M, Sundararaman T. Human resources for health in India. The Lancet. 2011; 377(9765): 587-98.3 Hammer J, Aiyar Y, Samji S. Understanding government failure in public health services. Economic and Political Weekly. 2007; 42: 4049–58.
Table6Availability of services in health facilities, by platform
DISTRICT HOSPITAL (DH)
CIVIL HOSPITAL (CH)
COMMUNITY HEALTH CENTRE (CHC)
Total OBGYN services 100% 100% 100%
Routine births 100% 93% 100%
Emergency obstetrics 100% 93% 100%
Antenatal care 100% 100% 100%
Surgical services 100% 80% 76%
Cardiology 53% 0% 3%
Psychiatric 6% 7% 0%
Accident, trauma, and emergency 100% 60% 59%
Ophthalmology 100% 73% 74%
Pediatric 100% 73% 29%
General anesthesiology 94% 47% 6%
Blood bank 100% 27% 9%
Dentistry 94% 67% 6%
DOTS treatment 94% 93% 82%
STI/HIV 88% 80% 24%
Immunization 100% 100% 88%
Internal/general med 100% 100% 94%
Mortuary 100% 47% 59%
Burns 100% 53% 29%
Orthopedic 100% 40% 3%
Pharmacy 100% 100% 100%
Chemotherapy 6% 0% NA
Dermatology 53% 13% NA
Alternative medicine 53% 47% 32%
Diagnostic medical 94% 87% 65%
Laboratory services 100% 100% 100%
Outreach services 53% 67% 59%
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: All values represent the percentage of facilities, by platform, that reported offering a given service at least one day during a typical week.
for instance, recommends a vastly strengthened infra-structure, with substantial increases in personnel at every tier of the public health system.4
Based on the ABCE sample, we found substantial heterogeneity across facility types in MP by considering the total number of staff in the context of bed strength (i.e., number of beds in the facility) and patient load (Figure 3). Overall, the most common medical staff at district hospitals were nurses, while at lower levels, para-medical staff outnumbered doctors and nurses. This is a reflection of the differential service offerings between higher- and lower-level facilities. Additionally, higher-level facilities tended to have a greater number of health per-sonnel overall; while a degree of this variation is due to differences in service provision and population size, some of this indicates relative shortages in human resources for health.
The volume of human resources across platforms was on the expected lines, with the greatest number of doc-tors, nurses, paramedical staff, and non-medical staff concentrated at the district hospitals, and the least at the sub-health centres. Civil hospitals reported the sec-ond highest number of personnel; however, the total personnel at these facilities was one-quarter of what was reported by district hospitals. Community health centres
4 National Rural Health Mission. Ministry of Health and Family Welfare, Govern-ment of India. Mission Document (2005-2012). New Delhi, India: Government of India, 2005.
Figure3 Composition of facility personnel, by platform
0 50 100 150 200 250Number of Staff
Sub Health Centre
Primary Health Centre
Community Health Centre
Civil Hospital
District Hospital
MP
Doctors Nurses
Para-medical staff Non-medical staff
2120
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
maintained a smaller body of health workers, an average total of 28, with most workers reported to be paramedical staff. Primary health centres reported, on average, seven health workers in total, most of which were paramedical staff. Finally, sub-health centres reported the lowest num-ber of staff, with only two paramedical and non-medical personnel who perform immunizations, simple outpatient care, and community outreach.
NursestodoctorratioThe ratio of number of nurses to number of doctors
is presented in Figure 4. A ratio greater than 1 indicates that nurses outnumber doctors; for instance, a ratio of 2 indicates that there are two nurses staffed for every one doctor. Alternatively, a ratio lower than 1 indicates that doctors outnumber nurses; for instance, a ratio of 0.5 in-dicates there is one nurse staffed for every two doctors.
In general, district hospitals reported a high ratio, indi-cating that they staff more nurses than doctors. However, the ratio reported by various district hospitals ranged from 0.7 to 9.9. About half of civil hospitals reported more nurses than doctors, and half had more doctors than nurses. There was heterogeneity among community health centres, with ratios ranging from 0.1 to 6. Finally, all primary health centres reported fewer or the same num-ber of nurses staffed as doctors.
Nursesanddoctorstoparamedical andnon-medicalstaffThe ratio of number of nurses and/or doctors to num-
ber of paramedical and/or non-medical staff in 2012 is presented in Figure 5. A ratio greater than 1 indicates that nurses and doctors outnumber paramedical and non-medical personnel; for instance, a ratio of 2 indicates that there are two nurses and/or doctors staffed for every one paramedical/non-medical staff. Alternatively, a ratio lower than 1 indicates that paramedical and/or non-medi-cal personnel outnumber nurses and/or doctors.
Most district hospitals reported ratios greater than 1, ranging from 0.5 to 1.7. All levels below district hospi-tal reported ratios less than or equal to 1. Civil hospitals had the next highest ratio of nurses and doctors to para-medical and non-medical staff, with an average of 0.4. Community health centres were quite homogenous, re-porting an average ratio of 0.3, with facilities reporting ratios that ranged from 0.1 to 0.5. Primary health centres employ more paramedical and non-medical staff than doctors and nurses, with all but two facilities reporting a ratio less than 0.7.
In isolation, facility staffing numbers are less meaning-ful without considering a facility’s overall patient volume and production of specific services. For instance, if a fa-
cility mostly offers services that do not require a doctor’s administration, failing to achieve the doctor staffing tar-get may be less important than having too few nurses. Further, some facilities may have much smaller patient volumes than others, and thus “achieving” staffing tar-gets could leave them with an excess of personnel given patient loads. While an overstaffed facility has a different set of challenges than an understaffed one, each reflects a poor alignment of facility resources and patient needs. To better understand bottlenecks in service delivery and areas to improve costs, it is important to assess a facili-ty’s capacity (inputs) in the context of its patient volume and services (outputs). We further explore these find-ings in the “Efficiency and costs” section. As part of the ABCE project in India, we compare levels of facility-based staffing with the production of different types of health services. In this report, we primarily focus on the delivery of health services by skilled medical personnel, which includes doctors, nurses, and other paramedical staff. It is possible that non-medical staff also contribute to ser-vice provision, especially at lower levels of care, but the ABCE project in India is not currently positioned to ana-lyze these scenarios.
InfrastructureandequipmentHealth service provision depends on the availability of
adequate facility infrastructure, equipment, and supplies (physical capital). In this report, we focus on four essen-tial components of physical capital: power supply, water and sanitation, transportation, and medical equipment, with the latter composed of laboratory, imaging, and other medical equipment. Table 7 illustrates the range of physical capital, excluding medical equipment, available across platforms.
Power supplyAll hospitals and community health centres reported
access to a functional electrical supply. Among smaller facilities, 6% of primary health centres and 57% of sub health centres lacked functional electricity. No facilities re-ported solely relying on a generator for power.
These results demonstrate some improvement in the availability of electricity at the lowest platform level compared to 2007–2008, when only 7% of sub health centres had a regular electric supply.5 However, inade-quate access to consistent electric power has substantial
5 International Institute for Population Sciences (IIPS). District Level Household and Facility Survey (DLHS-3), 2007-08: India, Madhya Pradesh. Mumbai, India: IIPS, 2010.
Figure4 Nurse to doctor ratio, by platform
Vertical bars represent the platform average ratio.
0 2 4 6 8 10
District Hospital Civil Hospital
Community Health Centre Primary Health Centre
Figure5Ratio of nurses and doctors to paramedical and non-medical staff, by platform
Vertical bars represent the platform average ratio.
0 2 4 6 8 10
District Hospital Civil Hospital
Community Health Centre Primary Health Centre
Table7 Availability of physical capital, by platform
DISTRICT HOSPITAL (DH)
CIVIL HOSPITAL (CH)
COMMUNITY HEALTH CENTRE (CHC)
PRIMARY HEALTH CENTRE (PHC)
SUB HEALTH CENTRE (SHC)
Functionalelectricity 100% 100% 100% 94% 43%
Pipedwater 100% 100% 76% 33% 9%
Flushtoilet 100% 100% 85% 91% 69%
Handdisinfectant 94% 100% 94% 91% 71%
Any4-wheeledvehicle 100% 80% 79% 6% NA
Emergency4-wheeledvehicle 94% 73% 59% 4% NA
Landlinephone 0% 40% 56% 9% NA
Computer 100% 87% 94% 28% NA
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: Values represent the percentage of facilities, by platform, that had a given type of physical capital
2322
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
implications for health service provision, particularly for the effective storage of medications, vaccines, and blood samples.
WaterandsanitationThere was a high availability of improved sanitation.
Generally, more facilities had sewer infrastructure than functional piped water. All hospitals, 85% of community health centres, and 91% of primary health centres had sewer infrastructure, though this was found in only 69% of sub health centres. Hand disinfectant was broadly available as a supplementary sanitation method at most platform levels, though again was less available in sub health centres. Access to piped water declined at lower levels of the health system as all hospitals but only 76% of community health centres, 33% of primary health centres, and 9% of sub health centres had piped water. Among all facilities, 30% reported a severe shortage of water at some point during the year. These findings show a mixture of both notable gains and ongoing needs for facility-based water sources and sanitation practices among primary care facilities.
TransportationandcomputersFacility-based transportation and modes of communi-
cation varied across platforms. In general, the availability of a vehicle, irrespective of its emergency capabilities, substantially decreased down the levels of health plat-forms. Only 4% of primary health centres had emergency transportation and 6% had any four-wheeled vehicles at all, which means transferring patients under emergency circumstances from these facilities could be fraught with delays and possible complications. This transportation gap and the coordination of transport might be further exacerbated by the relatively low availability of landline phones at lower-level facilities. The availability of a func-tional computer in facilities exceeded that of phones across all platforms but sub centres.
EquipmentFor three main types of facility equipment – medical,
lab, and imaging – clear differences emerge across levels of health service provision, with Table 8 summarizing the availability of functional equipment by platform.
We used WHO’s Service Availability and Readiness As-sessment (SARA) survey as our guideline for what types of equipment should be available in hospitals and primary
care facilities.6 Table 8 illustrates the distribution of SARA scores across platforms. In general, hospitals had greater availability of medical equipment, and notable deficits in essential equipment availability were found in the lower levels of care. Lacking basic equipment such as scales and blood pressure cuffs can severely limit the collection of important patient clinical data, and the large majority of facilities across all platforms did carry these. Micro-scopes and corresponding components were largely prevalent among all facilities, except at primary health centres, where all had slides but almost half did not have a microscope to use them with. Additional testing capac-ity was relatively high at district hospitals but declined across lower platforms. For instance, 100% of district hospitals and only 67% of civil hospitals and community health centres had a blood chemistry analyzer. All district hospitals and 80% of civil hospitals had both a functional glucometer and test strips for the glucometer. However, in both community health centres and primary health centres, more facilities had glucometer test strips than had a glucometer itself. This indicates limited capacity for addressing non-communicable diseases (NCDs) such as diabetes, for which this equipment is necessary. District hospitals had good availability of imaging equipment, with the notable exception of CT scans, which were avail-able in only 29% of facilities. Civil hospitals show patchier availability of imaging equipment, as nearly half had no ECG and 60% had no ultrasound. Community health cen-tres had poor availability of essential imaging equipment.
Overall, these findings demonstrate gradual improve-ments in equipping health facilities with basic medical equipment in MP, as well as the continued challenge of ensuring that these facilities carry the supplies they need to provide a full range of services. Measuring the avail-ability of individual pieces of equipment sheds light on specific deficits, but assessing a health facility’s full stock of necessary or recommended equipment provides a more precise understanding of a facility’s service capacity.
6 World Health Organization (WHO). Service Availability and Readiness Assessment (SARA) Survey: Core Questionnaire. Geneva, Switzerland: WHO, 2013.
Table8Availability of functional equipment, by platform
DISTRICT HOSPITAL
CIVIL HOSPITAL
COMMUNITY HEALTH CENTRE
PRIMARY HEALTH CENTRE
SUB-HEALTHCENTRE
Medicalequipment
Wheelchair 100% 93% 100% 88% NA
Adult scale 100% 100% 100% 96% 94%
Child scale 100% 100% 100% 96% 91%
Blood pressure apparatus 100% 100% 100% 100% 94%
Stethoscope 100% 100% 100% 99% 85%
Light source 100% 93% 82% 65% 38%
Labequipment
Glucometer 100% 80% 74% 52% NA
Test strips for glucometer 100% 80% 82% 58% NA
Hematologic counter 100% 73% 47% 22% NA
Blood chemistry analyzer 100% 67% 56% 10% NA
Incubator 94% 80% 56% 13% NA
Centrifuge 100% 100% 91% 36% NA
Microscope 100% 100% 97% 52% NA
Slides 100% 100% 100% 100% 99%
Slide covers 100% 100% 94% 87% 93%
Imagingequipment
X-ray 100% 80% 74% NA NA
ECG 100% 53% 29% NA NA
Ultrasound 88% 40% 3% NA NA
CT scan 29% NA NA NA NA
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: Availability of a particular piece of equipment was determined based on facility ownership on the day of visit. Data on the number of items present in a facility were not collected. All values represent the percentage of facilities, by platform, that had a given piece of equipment.
2524
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
Table9Availability of tests and functional equipment to perform routine antenatal care, by platform
Facilities fully equipped for ANC provision based on above tests and equipment availability
88% 33% 65% 10% 12%
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: Availability of a given ANC item was determined by its availability at a facility on the day of visit. All values represent the percentage of facilities, by platform, that had the given ANC item. The service summary section compares the total percentage of facilities reporting that they provided ANC services with the total percentage of facilities that carried all of the functional equipment to provide ANC services.
FocusonserviceprovisionFor the production of any given health service, a
health facility requires a complex combination of the ba-sic infrastructure, equipment, and pharmaceuticals, with personnel who are adequately trained to administer nec-essary clinical assessments, tests, and medications. Thus, it is important to consider this intersection of facility re-sources to best understand facility capacity for care. In this report, we further examined facility capacity for a subset of specific services – antenatal care, delivery, gen-eral surgery, and laboratory testing. For these analyses of service provision, we only included facilities that re-ported providing the specific service, excluding facilities that were potentially supposed to provide a given service but did not report providing it in the ABCE Facility Sur-vey. Thus, our findings reflect more of a service capacity “ceiling” across platforms, as we are not reporting on the facilities that likely should provide a given service but have indicated otherwise on the ABCE Facility Survey.
AntenatalcareservicesIn MP, according to the National Family Health Sur-
vey-4, 36% of women had at least four antenatal care (ANC) visits during their last pregnancy.7 This is a low level of coverage; moreover, it neither reflects what services were actually provided nor the quality of care received. Through the ABCE Facility Survey, we estimated what pro-portion of facilities stocked the range of tests and medical equipment to conduct a routine ANC visit. It is important to note that this list was not exhaustive but represented a number of relevant supplies necessary for the provision of ANC.
The availability of tests and functional equipment for ANC is presented in Table 9. While all hospitals and com-
7 International Institute for Population Sciences (IIPS). National Family Health Survey (NFHS-4), 2015-2016: Madhya Pradesh Factsheet. Mumbai, India: IIPS, 2016.
Table10 Availability of blood tests and functional equipment to perform routine delivery care, by platform
DISTRICT HOSPITAL
CIVIL HOSPITAL
COMMUNITY HEALTH CENTRE
PRIMARY HEALTH CENTRE
Testingavailability
Hemoglobin 100% 100% 94% 83%
Glucometer and test strips 100% 80% 68% 48%
Cross-match blood 100% 40% NA NA
Medicalequipment
Blood pressure apparatus 100% 100% 100% 100%
IV catheters 100% 100% 100% 86%
Gowns 100% 100% 100% 74%
Measuring tape 88% 93% 100% 70%
Masks 100% 100% 100% 86%
Sterilization equipment 94% 80% 56% 42%
Adult bag valve mask 100% 87% 59% 42%
Ultrasound 88% 40% NA NA
Deliveryequipment
Infant scale 100% 87% 88% 67%
Scissors or blade to cut umbilical cord 100% 93% 100% 86%
Facilities fully equipped for delivery services based on the above tests and equipment availability 29% 7% 15% 1%
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: Availability of a given delivery item was determined by its availability at a facility on the day of visit. All values represent the percentage of facilities, by platform, that had the given delivery item. The service summary section compares the total percentage of facilities reporting that they provided routine delivery services with the total percentage of facilities that carried all of the recommended pharmaceuticals and functional equipment to provide routine delivery services.
2726
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
munity health centers in this survey reported providing ANC services, many were not adequately supplied for care. District hospitals were, on the whole, well equipped for ANC services, with 88% of facilities having all the nec-essary tests and equipment. However, only one-third of civil hospitals were equipped: for example, 60% had no available ultrasound. Urinalysis and hemoglobin testing were actually available in more sub health centres than primary care centres, but facilities across all health centre platforms lacked many essential tests.
DeliverycareservicesEighty-one percent of deliveries in MP are in a health
facility, and 70% are in public facilities.8 Availability of essential equipment is necessary for providing high-qual-ity delivery care; these results are presented in Table 10. Availability was generally highest in district hospitals, declining at lower levels. While most civil hospitals, com-munity health centres, and primary health centres offered routine delivery services, none had all essential tests and equipment available. Only 29% of district hospitals were fully equipped. An ultrasound machine was absent from 60% of civil hospitals.
8 International Institute for Population Sciences (IIPS). National Family Health Survey (NFHS-4), 2015-2016: Madhya Pradesh Factsheet. Mumbai, India: IIPS, 2016.
Table11Availability of blood tests and functional equipment to perform general surgery, by platform
Table12 Availability of laboratory tests, by platform
DISTRICT HOSPITAL (DH)
CIVIL HOSPITAL (CH)
COMMUNITY HEALTH CENTRE(CHC)
PRIMARY HEALTH CENTRE(PHC)
Blood typing 100% 93% 79% 16%
Cross-match blood 100% 40% NA NA
Complete blood count 100% 80% 59% 12%
Hemoglobin 100% 100% 94% 83%
HIV 100% 87% 79% 17%
Liver function 100% 67% 65% NA
Malaria 100% 100% 94% 83%
Renal function 94% 53% 41% 4%
Serum electrolytes 82% 7% 12% NA
Spinal fluid test 82% 13% 9% NA
Syphilis 100% 80% 76% NA
Tuberculosis skin 100% 100% 85% 29%
Urinalysis 100% 100% 91% 77%
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: Availability of a given test was determined by its availability at a facility on the day of visit. All values represent the percentage of facilities, by platform, that had the given test.
DISTRICT HOSPITAL (DH)
CIVIL HOSPITAL (CH)
COMMUNITY HEALTH CENTRE(CHC)
PRIMARY HEALTH CENTRE(PHC)
Testingavailability
Hemoglobin 100% 100% 94% 83%
Cross-match blood 100% 40% NA NA
Medicalequipment
Blood pressure apparatus 100% 100% 100% 100%
IV catheters 100% 100% 100% 86%
Sterilization equipment 94% 80% 59% 42%
Gowns 100% 100% 100% 74%
Masks 100% 100% 100% 86%
Adult bag valve mask 100% 87% 59% 42%
Surgicalequipment
Surgical scissors/blade 100% 93% 100% 86%
Thermometer 100% 87% 65% 51%
General anesthesia equipment 100% 73% 29% 1%
Scalpel 88% 93% 91% 58%
Suction apparatus 100% 87% 65% 36%
Retractor 100% 87% 65% 39%
Nasogastric tube 94% 80% 62% 36%
Blood storage unit/refrigerator 100% 40% 15% NA
Intubation equipment 100% 67% 41% 17%
Servicesummary
Facilities reporting general surgery services 100% 80% 76% 55%
Facilities fully equipped for general surgery services based on the above tests and equipment availability 82% 25% 12% 0%
NA: Not applicable to this platform according to standards.
LOWEST AVAILABILITY HIGHEST AVAILABILITY
Note: Availability of a given surgery item was determined by its availability at a facility on the day of visit. All values represent the percentage of facilities, by platform, that had the given surgery item. The service summary section compares the total percentage of facilities reporting that they provided general surgery services with the total percentage of facilities that carried all of the recommended functional equipment to provide general surgery services.
2928
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
GeneralsurgeryservicesAvailability of essential tests and equipment for gen-
eral surgery services are presented in Table 11. Eighty-two percent of district hospitals had all of the essential items; availability was substantially lower in civil hospitals and community health centres, while no primary health cen-tres were fully equipped. Essential medical equipment was mostly available in both types of hospitals, and avail-ability of surgical equipment was also relatively high, with the exception of blood storage units in civil hospi-tals. Most civil hospitals also notably lacked cross-match blood tests. There were large gaps, particularly in medical and surgical equipment, in community health centres and primary health centres. No primary health centre was fully equipped to carry out surgery. It is also crucial to consider the human resources available to perform surgical proce-
dures, as assembling an adequate surgical team is likely to affect patient outcomes. Given the nature of documen-tation of human resources in the records, such data could not be captured, but future work on assessing surgical ca-pacity at health facilities should collect this information.
LaboratorytestingThe availability of laboratory tests is presented in Table
12. While all district hospitals, civil hospitals, and commu-nity health centres offer the range of laboratory services, there were gaps in test availability. Availability was gen-erally high in district hospitals, and decreased at lower facility levels. Serum electrolyte tests, useful as part of a metabolic panel and to measure symptoms of heart dis-ease and high blood pressure, had very low availability in civil hospitals (7%) and community health centres (12%).
Figure6Number of outpatient visits, by platform
Each line represents outpatient visits for an individual facility, with the bold line depicting the average for the platform. Scales are different for each platform.
020
000
4000
060
000
8000
0V
isits
2008 2009 2010 2011 2012
OP visits by facility OP visits average
CHC
020
0000
4000
0060
0000
Vis
its
2008 2009 2010 2011 2012
OP visits by facility OP visits average
DH
050
0010
000
1500
020
000
2500
0V
isits
2008 2009 2010 2011 2012
OP visits by facility OP visits average
PHC0
5000
010
0000
1500
0020
0000
Vis
its
2008 2009 2010 2011 2012
OP visits by facility OP visits average
CH
010
0020
0030
00V
isits
2008 2009 2010 2011 2012
OP visits by facility OP visits average
SHC
Figure7Number of inpatient visits (excluding deliveries), by platform
Each line represents inpatient visits for an individual facility, with the bold line depicting the average for the platform. Scales are different for each platform.
020
0040
0060
0080
0010
000
Vis
its
2008 2009 2010 2011 2012
IP visits by facility IP visits average
CHC
020
000
4000
060
000
8000
0V
isits
2008 2009 2010 2011 2012
IP visits by facility IP visits average
DH
020
040
060
080
010
00V
isits
2008 2009 2010 2011 2012
IP visits by facility IP visits average
PHC
050
0010
000
1500
0V
isits
2008 2009 2010 2011 2012
IP visits by facility IP visits average
CH
3130
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
Spinal fluid tests were also rare among facilities below district hospitals. Most facilities were equipped to test for malaria, HIV, and TB, with the exception of primary health centres, which had a low availability of TB (29%) and HIV (17%) tests.
FacilityoutputsMeasuring a facility’s patient volume and the num-
ber of services delivered, which are known as outputs, is critical to understanding how facility resources align with patient demand for care. Figure 6 illustrates the trends in average outpatient volume across platforms and over time.
OutpatientvisitsThe number of outpatient visits by year, by platform,
is presented in Figure 6. In general, the average num-ber of outpatient visits remained stable over five fiscal years, with slight declines for civil hospitals, community health centres, and primary health centres. Patient vol-ume was highest in district (average of 176,970–188,262 visits per year). Civil hospitals reported an average of 60,084–71,016 visits per year, which was near triple the number reported by community health centres (average of 20,383–24,342 visits per year). Primary health centres reported more than 10 times more outpatient visits (aver-age of 4,295–5,196 visits per year) than sub-health centres (average of 402–443 visits per year).
Figure8 Number of deliveries, by platform
Note: Each line represents delivery visits for an individual facility, with the bold line depicting the average for the platform. Scales are different for each platform.
010
0020
0030
00D
eliv
erie
s
2008 2009 2010 2011 2012
Deliveries by facility Deliveries average
CHC
4000
6000
8000
1000
012
000
Del
iver
ies
2008 2009 2010 2011 2012
Deliveries by facility Deliveries average
DH
050
010
0015
0020
00D
eliv
erie
s
2008 2009 2010 2011 2012
Deliveries by facility Deliveries average
PHC0
2000
4000
6000
8000
Del
iver
ies
2008 2009 2010 2011 2012
Deliveries by facility Deliveries average
CH
Figure9 Number of immunization doses administered, by platform
Note: Each line represents immunization doses for an individual facility, with the bold line depicting the average for the platform. Scales are different for each platform.
050
0010
000
1500
0D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses by facility Immunization doses average
CHC
020
000
4000
060
000
8000
010
0000
Dos
es a
dmin
iste
red
2008 2009 2010 2011 2012
Immunization doses by facility Immunization doses average
DH
020
0040
0060
0080
00D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses by facility Immunization doses average
PHC
010
000
2000
030
000
4000
0D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses by facility Immunization doses average
CH
010
0020
0030
0040
00D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses by facility Immunization doses average
SHC
3332
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
InpatientvisitsInpatient visits generally entail more service demands
than outpatient visits, including ongoing occupancy of facility resources such as beds. The reported number of inpatient visits (other than deliveries) by year are pre-sented in Figure 7. Over time, the average number of inpatient visits have increased for all platforms. District hospitals provided care for an average of 22,708–28,782 inpatient visits per year, with one facility reporting con-sistently more than 60,000 inpatient visits per year. Civil hospitals provided care for an average of 6,176–6,656 visits per year, while community health centres provided one-third as many visits (an average between 2,502 and 2,618 inpatient visits per year). Primary health centres re-ported substantially fewer inpatient visits (on average 85–106 visits per year). It is important to note that the ABCE Facility Survey did not capture information on the length of inpatient stays, which is a key indicator to moni-tor and include in future work.
DeliveriesThe reported number of deliveries, by platform and
over time, is presented in Figure 8. District hospitals re-ported an average between 6,136 and 6,688 deliveries in each year of observation, which his triple that of civil hospitals (an average of 2,178–2,344 deliveries per year). While many hospitals experienced an increase in the number of deliveries over time, several hospitals reported decreasing numbers over the five years of observation. Community health centres reported an annual average number of deliveries between 1,117 and 1,279. Few deliv-eries were reported in primary health centres (an average of 214–245 deliveries per year). The ratio of deliveries to inpatient visits is higher among the lower platforms.
Table13 Characteristics of patients interviewed after receiving care at facilities
Note: Educational attainment refers to the patient’s level of education or the attendant’s educational attainment if the interviewed patient was younger than 18 years old.
ImmunizationThe number of immunization doses administered over
time, by platform, is presented in Figure 9. Generally, the average number of doses administered remained sta-ble over the five years. District hospitals reported many more immunization doses administered (annual aver-ages between 37,091 and 42,922) than civil hospitals (annual averages between 13,171 and 13,693) and commu-nity health centres (annual averages between 3,652 and 4,736). Facilities at the PHC and SHC level are central to immunization delivery; primary health centres reported an average of 708–1,033 doses per year while sub health centres reported slightly more, with an average of 1,228–1,366 doses per year.
PatientperspectivesA facility’s availability of and capacity to deliver ser-
vices is only half of the health care provision equation; the other half depends upon patients seeking those health services. Many factors can affect patients’ decisions to seek care, ranging from associated visit costs to how pa-tients view the care they receive. These “demand-side” constraints can be more quantifiable (e.g., distance from facility) or intangible (e.g., perceived respectfulness of
Figure10Patient travel times to facilities, by platform
0 20 40 60 80 100Percent (%)
SHC
PHC
CHC
CH
DH
< 30 min. > 30 min.
DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre; SHC: Sub health centre
Figure11Patient wait times at facilities, by platform
0 20 40 60 80 100Percent (%)
SHC
PHC
CHC
CH
DH
< 30 min. > 30 min.
DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre; SHC: Sub health centre
the health care provider), but each can have the same im-pact on whether patients seek care at particular facilities or have contact with the health system at all.
In order to measure demand-side constraints and pa-tient experiences, interviews were conducted with 2,092 patients or their attendants at public facilities (Table 13). The majority of patients were male (52%), though at all fa-cility levels except district hospitals, most patients were female. One-third of patients identified as part of a sched-uled caste/scheduled tribe, and nearly half identified as another backwards caste. Most patients had some educa-tion (70%), though primary health centres tended to see patients with less education. Over half (54%) of patients were under the age of 30.
TravelandwaittimesThe amount of time patients spend traveling to facili-
ties and then waiting for services can substantially affect their care-seeking behaviors. Most patients had travel times of less than 30 minutes to a facility for care (Figure 10). Travel time was shorter for patients seeking care at lower-level facilities than higher-level ones; 42% of pa-tients who went to district hospitals traveled more than 30 minutes, while none at sub-health centres traveled for as
3534
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
Figure12 Patient scores of facilities, by platform
0 20 40 60 80 100Percent (%)
SHC
PHC
CHC
CH
DH
<6 6-78-9 10
Table14Proportion of patients satisfied with facility visit indicators, by platform
DISTRICT HOSPITAL
CIVIL HOSPITAL
COMMUNITY HEALTH CENTRE
PRIMARY HEALTH CENTRE
SUB HEALTH CENTRE
Staffinteractions
Nurse/ANM Medical provider respectfulness 83% 82% 74% 69% 100%
Clarity of provider explanations 79% 77% 75% 68% 75%
Time to ask questions 72% 75% 72% 67% 75%
Doctor
Medical provider respectfulness 78% 64% 70% 71% NA
Clarity of provider explanations 79% 64% 76% 73% NA
Time to ask questions 74% 67% 74% 77% NA
Facilitycharacteristics
Cleanliness 57% 52% 55% 57% 13%
Privacy 66% 55% 65% 58% 25%
LOWEST PROPORTION HIGHEST PROPORTION
NA: Results not applicable.
DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre; SHC: Sub health centre
Note: Facility ratings were reported along a scale of 0 to 10, with 0 as the worst facility possible and 10 as the best facility possible.
long. This finding is not unexpected, as these are the clos-est health facilities for many patients, particularly those in rural areas. It also reflects the fact that many patients travel longer distances to receive the kind of specialized care offered at hospitals.
Wait time is also an important determinant of patient satisfaction. The large majority of patients waited less than 30 minutes to receive care at all platforms (Figure 11), and nearly all patients seeking care at community health centres (94%) and primary health centres (97%) received care within 30 minutes. At district hospitals, 21% of pa-tients waited more than 30 minutes to receive care.
PatientsatisfactionratingsWe report primarily on factors associated with patient
satisfaction with provider care and perceived quality of services by patients with regard to medicine availability and hospital infrastructure, as these have been previously identified to be of significance in the patient’s perception
Figure13Availability of prescribed drugs at facility, by platform
0 20 40 60 80 100Percent (%)
SHC
PHC
CHC
CH
DH
Got none/some of the drugs Got all perscribed drugs
DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre; SHC: Sub health centre
of quality of health services in India.9
Ratings of patient satisfaction, based on a scale from one to 10, with 10 being the highest score, are presented in Figure 12. Overall, patients were satisfied with the care they received and, in general, ratings were higher for higher-level platforms. 12.5% of patients receiving care at a sub health centre gave a rating below six, while no pa-tients gave a rating of 10. Among all patients who gave a rating, only 1.9% rated their facility a 10. While commu-nity health centres had the highest proportion of patients give a 10-rating (2.9%), it also had the highest proportion of patients who rated lower than six (18.1%).
Patients were also asked more detailed questions about satisfaction with providers and facility character-istics (Table 14). Less than two-thirds of patients were satisfied with facility cleanliness or privacy of facilities.
9 Rao KD, Peters DH, Bandeen-Roche K. Towards patient-centered health services in India—a scale to measure patient perceptions of quality. International Journal for Quality in Health Care. 2006; 18(6):414-421.
FemaleMale
>=40 years16-39 yearsOther caste
Backwards casteSchooling
No schoolingNot prescribed all drugs
Prescribed all drugsWait time <30 min
Wait time >=30 minDH
SHCCHC
CH
0 1 2 3Odds Ratio
Figure14Determinants of satisfaction with doctors
Dotted vertical line represents an odds ratio of one. Black points represent the reference groups, which all carry an odds ratio of one. Compared to the referent category, significant odds ratios and 95% confidence intervals are represented with blue points and horizontal lines, respectively. Odds ratios that are not significant are represented by green points, and their 95% confidence intervals with a green horizontal line. Any confidence intervals with an upper bound above 3 were truncated for ease of interpretation.
DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre
3736
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
These ratings were lowest at sub health centres, where only 13% were satisfied with cleanliness and 25% were satisfied with privacy. Three parameters were assessed to document satisfaction with health providers – being treated respectfully by the provider, clarity of explana-tion provided by the provider, and that provider gave enough time to ask questions about a health problem or treatment – using a five-point Likert scale, with the high-est ratings of good and very good responses combined as satisfied, and rest as not satisfied. Using the three pa-rameters of satisfaction, a composite satisfaction variable was created separately for doctors and nurses – if a pa-tient reported good/very good for all three parameters, it was categorized as satisfied. At district hospitals and civil hospitals, patients receiving care from nurses and auxiliary nurse midwives (ANMs) reported higher levels of satisfaction with respectfulness, clarity, and time than those receiving care from doctors. This trend was re-
FemaleMale
>=40 years16-39 yearsOther caste
Backwards caste
SchoolingNo schooling
Not prescribed all drugsPrescribed all drugs
Wait time <30 minWait time >=30 min
CHSHCPHCCHC
CH
0 1 2 3Odds Ratio
Figure15Determinants of satisfaction with nurses/ANMs
Dotted vertical line represents an odds ratio of one. Black points represent the reference groups, which all carry an odds ratio of one. Compared to the referent category, significant odds ratios and 95% confidence intervals are represented with blue points and horizontal lines, respectively. Odds ratios that are not significant are represented by green points, and their 95% confidence intervals with a green horizontal line. Any confidence intervals with an upper bound above 3 were truncated for ease of interpretation.
DH: District hospital; CH: Civil hospital; CHC: Community health centre; PHC: Primary health centre; SHC: Sub health centre
Table15Input-output model specifications
CATEGORY VARIABLES
Model 1
Inputs Expenditure on personnelExpenditure on pharmaceuticalsAll other expenditure
versed in community health centres and primary health centres, where patients were more satisfied with doctors. Satisfaction with both nurse and doctor interactions were lower for patients seeking care at community and primary health centres than district hospitals.
Access to to affordable drugs has been interpreted to be part of the right to health. Among 1,968 patients who were prescribed drugs and attempted to obtain those drugs during the visit, 1,927 received all prescribed drugs (Figure 13). This ranged from 98% of patients at primary health centres to 100% of patients at sub-health centres.
Reasons for patient satisfaction of medical care are complex, so a multivariable logistic regression was con-ducted to measure the association of select patient and facility characteristics that could determine patient satis-faction with medical doctors (Figure 14) and nurses/ANMs (Figure 15). For each characteristic, the odds ratio (OR) is presented. An odds ratio greater than 1.0 indicates that there are greater odds of being satisfied with care as compared to the reference group. An odds ratio below 1.0 indicates that there are lower odds of being satisfied with care than the reference group.
Longer wait time to receive attention was associated with lower patient satisfaction with doctors (OR: 0.44, 95% confidence interval [CI]: 0.31–0.64). Compared to patients of another caste, there was slightly higher satis-faction with doctors for patients of backwards caste (OR: 1.28, 95% CI: 1.02–1.61). Patients younger than 40 years were more satisfied with doctors than were older patients (OR: 1.35, 95% CI: 1.04–1.73). There was no difference in satisfaction by platform.
Considering all selected patient and facility charac-teristics, no factors significantly increased the odds of a patient being satisfied with their care.
EfficiencyandcostsThe costs of health service provision and the efficiency
with which care is delivered by health facilities go hand-in-hand. An efficient health facility uses resources well, producing a high volume of patient visits and services without straining its resources. Conversely, an ineffi-cient health facility is one where the use of resources is not maximized, leaving usable beds empty or medical staff seeing very few patients per day. We present techni-cal efficiency analysis for district hospitals, civil hospitals,
community health centres and primary health centres. Community health centres are stratified by levels (L2 and L3), due to the types of services provided.
AnalyticalapproachAn ensemble model approach was used to quantify
technical efficiency in health facilities, combining results from two approaches – the restricted versions of Data Envelopment Analysis (rDEA) and Stochastic Distance Function (rSDF).10 Based on this analysis, an efficiency score was estimated for each facility, capturing a facility’s use of its resources. Relating the outputs to inputs, the rDEA and rSDF approaches compute efficiency scores ranging from 0% to 100%, with a score of 100% indicat-ing that a facility achieved the highest level of production relative to all facilities in that platform.
This approach assesses the relationship between in-puts and outputs to estimate an efficiency score for each facility. Recognizing that each type of input requires a different amount of facility resources (e.g., on average, an inpatient visit uses more resources and more com-plex types of equipment and services than an outpatient visit), we applied weight restrictions to rescale each fa-cility’s mixture of inputs and outputs. The incorporation of additional weight restrictions is widely used in order to improve the discrimination of the models. Weight re-strictions are most commonly based upon the judgment about the importance of individual inputs and outputs, or reflect cost or price considerations. The resulting ensem-ble efficiency scores were averaged over five years and between the two input models.
For these models, service provision was categorized into outpatient visits, inpatient visits, delivery and immu-nization. Two input-output specifications were used, with the inputs being different in the two models. The inputs and outputs are listed in Table 15. The detailed data uti-lized for this analysis is documented in the annex. The average and range of inputs and outputs for the variables is presented in Table 16.
10 Di Giorgio L, Flaxman AD, Moses MW, Fullman N, Hanlon M, Conner RO, et al. Efficiency of Health Care Production in Low-Resource Settings: A Monte-Carlo Simulation to Compare the Performance of Data Envelopment Analysis, Stochastic Distance Functions, and an Ensemble Model. PLOS ONE. 2016; 11(2): e0150570.
3938
A B C E I N M A D H YA P R A D E S H M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
CostsofcareTotal expenditure, by district and platform, is pre-
sented in Table 17. In terms of annual total expenditures, trends in average facility spending varied by platform be-tween 2008 and 2012 (Figure 16). All platforms recorded higher spending in 2012 than 2008. Spending on person-nel accounted for the majority of annual spending across facility types. Community health centres spent a slightly lower proportion of their total expenditures on personnel than other platforms, while the proportion of expenditure
Table16Average and range of inputs and outputs, by platform. INR denotes Indian Rupees.
on medical supplies was highest at primary health centres (Figure 17).
It is important to note that data availability on the in-puts and output indicators varied across the facilities and platforms, with more non-availability for PHCs. Facilities with five years of missing data for any input or output vari-able were dropped from analysis. In addition, the data were smoothed where necessary based on the trends seen in inputs or outputs for that facility.
To further illustrate the production of outputs per in-
puts – in this case, staff – a simple ratio of outpatient visits (Figure 18), inpatient visits (Figure 19), deliveries (Figure 20), and immunization doses (Figure 21) per staff are pre-sented. District hospitals produced an average of 1,099 outpatient visits per staff, though the ratio ranged greatly. The average ratio for civil hospitals was 1,334 visits per staff, for community health centres (L2) was 862, for com-munity health centres (L3) was 1,007, and for primary health centres was 844. This gradient was similar for inpa-tient visits, with district hospitals providing 157 inpatient
visits per staff, civil hospitals providing 125, commu-nity health centres (L2) providing 80, community health centres (L3) providing 128, and primary health centres providing 24. The range of inpatient visits per staff was low for primary health centres, where inpatient visits are rare. Overall, as expected, outpatient visits accounted for the overwhelmingly large majority of the patients seen per staff per day across the platforms.
Fewer deliveries were performed per staff than other services, with an average of 43 deliveries per staff in
Table17 Average annual cost in INR, by district and platform, last fiscal year. INR denotes Indian Rupees.
DISTRICTDISTRICT
HOSPITALCIVIL
HOSPITAL
COMMUNITY HEALTH CENTRE
(LEVEL2)
COMMUNITY HEALTH CENTRE
(LEVEL3)PRIMARY
HEALTH CENTRE
District 1 61,159,191 14,831,786 7,005,634 1,130,722
District 2 69,952,031 9,606,974 12,509,894 2,436,466
District 3 41,250,883 10,624,417 18,054,811 3,096,714
District 4 102,884,044 7,580,404 2,366,454
District 5 85,012,043 13,608,939 7,735,776 2,958,584 1,875,066
District 6 13,831,620 19,032,197 2,052,144
District 7 18,899,505 6,378,878 13,260,717 1,332,077
District 8 75,703,007 26,144,943 1,843,673
District 9 69,735,120 13,145,077 3,206,382 3,450,501
District 10 106,374,009 7,235,921 11,221,508 1,648,587
District 11 96,412,260 23,133,987 5,915,959 1,865,989
District 12 79,646,985 11,711,570 3,599,145
District 13 67,113,598 25,463,896 8,086,259 1,218,292
District 14 30,910,343 12,225,018
District 15 70,999,502 11,114,390 3,501,974
District 16 56,773,327 8,237,254 2,173,071
District 17 37,089,941 5,939,195 15,349,913 4,038,913
4140
A B C E I N M A D H YA P R A D E S H
Figure16 Average total and type of expenditure, by platform, 2008–20120
2040
6080
100
Exp
endi
ture
in 1
00,0
00 R
upee
s
2008 2009 2010 2011 2012
Personnel
Pharmaceuticals and consumables Other
COMMUNITY HEALTH CENTRES, LEVEL 20
5010
015
0E
xpen
ditu
re in
100
,000
Rup
ees
2008 2009 2010 2011 2012
Personnel
Pharmaceuticals and consumables Other
COMMUNITY HEALTH CENTRES, LEVEL 30
5010
015
020
025
0E
xpen
ditu
re in
100
,000
Rup
ees
2008 2009 2010 2011 2012
Personnel
Pharmaceuticals and consumables Other
CIVIL HOSPITALS
020
040
060
080
01,
000
Exp
endi
ture
in 1
00,0
00 R
upee
s
2008 2009 2010 2011 2012
Personnel
Pharmaceuticals and consumables Other
DISTRICT HOSPITALS
PRIMARY HEALTH CENTRES
Figure17 Average percentage of expenditure type, by platform, in 2012
district hospitals, 44 per staff in civil hospitals, 47 per staff in community health centres (L2), 47 per staff in commu-nity health centres (L3), and 38 per staff in primary health centres. For immunization doses, 234 doses were admin-istered per staff in district hospitals, 274 per staff in civil hospitals, 183 per staff in community health centres (L2), 189 per staff in community health centres (L3), and 156 per staff in primary health centres.
EfficiencyresultsUsing the five fiscal years of data to estimate the effi-
ciency scores for all facilities, two main findings emerged. First, efficiency scores were relatively higher for district hospitals and community health centres (level 3), with the primary health centres having the lowest efficiency across all the platforms. Second, the range between the facilities with highest and lowest efficiency scores was large within platforms suggesting that a substantial performance gap may exist between the average facility and facilities with the highest efficiency scores. Figure 22 depicts this range of facility efficiency scores across platforms for MP.
The five-year average efficiency of district hospitals ranged from 72.5% to 87.6%, with a platform average of 78.2%. Civil hospitals were between 57.2% and 72.2% effi-cient. Community health centres (L2) were between 39.1% and 68.4% efficient. Community health centres (L3) were
slightly more homogenous, ranging from 69.3% to 78.9% efficient. The range of efficiency scores was wide for pri-mary health centres, from 21.5% to 49%.
Efficiency by district is presented in Table 18. There is variation in facility efficiency both between and within districts, however, the primary health centres were signifi-cantly inefficient in all the districts.
If all facilities were perfectly efficient, many more pa-tient services could be provided with the same inputs (Figure 23). On average, district hospitals could provide 42,684 additional outpatient visits with the same inputs, while primary health centres could see an average of 10,576 additional outpatient visits. Community health cen-tres (L2) could administer an average of 3,319 additional immunization doses with the same inputs if all facilities were efficient.
Given observed levels of facility-based resources (beds and personnel), it would appear that many facilities had the capacity to handle much larger patient volumes than they reported. Figure 23 displays this gap in poten-tial efficiency performance across platforms, depicting the possible gains in total service provision that could be achieved if every facility in the ABCE sample operated at optimal efficiency.
We found that all types of facilities could expand their outputs substantially given their observed resources. Based on our analyses, the highest level of care, district hospitals, had the greatest potential for increasing service provision without expanding current resources. Over-all, based on our estimation of efficiency, a large portion of health facilities could increase the volume of patients seen and services provided with the resources available to them.
At the same time, many reports and policy documents emphasize that pronounced deficiencies in human re-sources for health exist across India in the public sector health system, such that “significant [human resources for health] will be required to meet the demand” for health services.11 Our results suggest otherwise, as most facilities in the ABCE sample had the potential to bolster service production given their reported staffing of skilled person-nel and physical capital.
These findings provide a data-driven understanding of facility capacity and how health facilities have used their
11 Rao M, Rao KD, Kumar AK, Chatterjee M, Sundararaman T. Human resources for health in India. The Lancet. 2011; 377(9765): 587-98.
05
1015
2025
Exp
endi
ture
in 1
00,0
00 R
upee
s
2008 2009 2010 2011 2012
Personnel
Pharmaceuticals and consumables Other
0 20 40 60 80 100Percent of Total Expenditure
Primary Health Centre
Community Health Centre L3
Community Health Centre L2
Area Hospital
District Hospital
Personnel
Pharmaceuticals and consumables
Other
L2: Level 2; L3: Level 3
M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
4342
A B C E I N M A D H YA P R A D E S H
Figure18Outpatient load per staff by platform 0
1000
2000
3000
4000
5000
Vis
its
2008 2009 2010 2011 2012
OP visits per staff by facilityOP visits per staff average
DH
050
010
0015
0020
0025
00V
isits
2008 2009 2010 2011 2012
OP visits per staff by facilityOP visits per staff average
CHC2
DH: District hospital; CH: Civil hospital; CHC2: Community health centre level 2; CHC3: Community health centre level 3; PHC: Primary health centreNote: each line represents an individual facility, with the bolded line depicting the average for the platform. Scales are different for each platform type.
020
0040
0060
00V
isits
2008 2009 2010 2011 2012
OP visits per staff by facilityOP visits per staff average
CH
600
800
1000
1200
1400
1600
Vis
its
2008 2009 2010 2011 2012
OP visits per staff by facilityOP visits per staff average
CHC3
010
0020
0030
00V
isits
2008 2009 2010 2011 2012
OP visits per staff by facilityOP visits per staff average
PHC
Figure19 Inpatient load per staff by platform
010
020
030
0V
isits
2008 2009 2010 2011 2012
IP visits per staff by facilityIP visits per staff average
DH
050
100
150
200
250
Vis
its
2008 2009 2010 2011 2012
IP visits per staff by facilityIP visits per staff average
CHC2
010
020
030
040
0V
isits
2008 2009 2010 2011 2012
IP visits per staff by facilityIP visits per staff average
CH
050
100
150
200
250
Vis
its
2008 2009 2010 2011 2012
IP visits per staff by facilityIP visits per staff average
CHC3
050
100
150
200
Vis
its
2008 2009 2010 2011 2012
IP visits per staff by facilityIP visits per staff average
PHC
DH: District hospital; CH: Civil hospital; CHC2: Community health centre level 2; CHC3: Community health centre level 3; PHC: Primary health centreNote: each line represents an individual facility, with the bolded line depicting the average for the platform. Scales are different for each platform type.
M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
4544
A B C E I N M A D H YA P R A D E S H
Figure20Deliveries per staff by platform0
5010
015
0D
eliv
erie
s
2008 2009 2010 2011 2012
Deliveries per staff by facilityDeliveries per staff average
DH
050
100
150
Del
iver
ies
2008 2009 2010 2011 2012
Deliveries per staff by facilityDeliveries per staff average
CHC2
DH: District hospital; CH: Civil hospital; CHC2: Community health centre level 2; CHC3: Community health centre level 3; PHC: Primary health centreNote: each line represents an individual facility, with the bolded line depicting the average for the platform. Scales are different for each platform type.
050
100
150
Del
iver
ies
2008 2009 2010 2011 2012
Deliveries per staff by facilityDeliveries per staff average
CH
050
100
Del
iver
ies
2008 2009 2010 2011 2012
Deliveries per staff by facilityDeliveries per staff average
CHC3
010
020
030
0D
eliv
erie
s
2008 2009 2010 2011 2012
Deliveries per staff by facilityDeliveries per staff average
PHC
Figure21Immunizations per staff per day by platform
020
040
060
0D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses per staff by facilityImmunization doses per staff average
DH
020
040
060
0D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses per staff by facilityImmunization doses per staff average
CHC2
DH: District hospital; CH: Civil hospital; CHC2: Community health centre level 2; CHC3: Community health centre level 3; PHC: Primary health centreNote: each line represents an individual facility, with the bolded line depicting the average for the platform. Scales are different for each platform type.
020
040
060
080
010
00D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses per staff by facilityImmunization doses per staff average
CH
020
040
060
0D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses per staff by facilityImmunization doses per staff average
CHC3
050
010
0015
00D
oses
adm
inis
tere
d
2008 2009 2010 2011 2012
Immunization doses per staff by facilityImmunization doses per staff average
PHC
M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
4746
A B C E I N M A D H YA P R A D E S H
Table18District-wise efficiency scores (%), by platform
DISTRICT/PLATFORM
DISTRICT HOSPITAL CIVIL HOSPITAL
COMMUNITY HEALTH CENTRE
LEVEL 2
COMMUNITY HEALTH CENTRE LEVEL 3 PRIMARY HEALTH CENTRE
1 1 2 3 1 2 1 1 2 3 4
District 1 66.9 68.3 42.5 33.5 82.5 42.1
District 2 72.5 39.1 66.7 38.6 8.1
District 3 58.6 40.3 85.1 50.1
District 4 86 46 51.8 43.3
District 5 91.3 45.4 71.5 57.5 68.4 69.3 50.5 27.4
District 6 70.9 78 31.1 6.2
District 7 71.8 90.6 50.6 51.1 77.2 51.1 29.9
District 8 63.5 50.4 43.2 11.5
District 9 72.7 58.9 57.2 20.5 80.7 22.5
District 10 80.1 78.2 78.9 27.8 49 19.5 19.2
District 11 75.8 73.5 67.9 64.2 58.9 48.8
District 12 74.5 34 38.1 25.8 57.5
District 13 77.8 77.6 72.2 54.9 40
District 14 87.9 83.5
District 15 89 71.3 53
District 16 77 70 71.9 25.8 21.9 24.4
District 17 87.6 64.5 58.7 20.9 13.4
White cells were either dropped from analysis due to data availability, or there were no more facilities to sample from that platform. There were no civil hospitals in Districts 2, 4, 6, 12, 14, 15, 16, and 17.
020
4060
8010
0
District Hospital Civil Hospital
Community Health Centre L2 Community Health Centre L3
Primary Health Centre
Figure22Range of efficiency scores across platforms
L2: Level 2; L3: Level 3Note: One data point per five-year facility average.
Note: Each circle represents the five-year facility average efficiency score; IQR refers to intra-quartile range.
resources in MP; at the same time, they are not without limitations. Efficiency scores quantify the relationship be-tween what a facility has and what it produces, but these measures do not fully explain where inefficiencies orig-inate, why a given facility scores higher than another, or what levels of efficiency are truly ideal. It is conceivable that always operating at full capacity could actually have negative effects on service provision, such as longer wait times, high rates of staff burnout and turnover, and com-promised quality of care. These factors, as well as less tangible characteristics such as facility management, are all important drivers of health service provision, and fu-ture work should also assess these factors alongside measures of efficiency.
M A I N F I N D I N G S : H E A LT H FA C I L I T Y P R O F I L E S
4948
A B C E I N M A D H YA P R A D E S H
Figure23 Observed and estimated additional visits that could be produced given observed facility resources
0 2,000 4,000 6,000 8,000Deliveries
Primary Health Centre
Community Health Centre L3
Community Health Centre L2
Area Hospital
District Hospital
Deliveries
Observed Estimate additional deliveries
0 50,000 100000 150000 200000Outpatient visits
Primary Health Centre
Community Health Centre L3
Community Health Centre L2
Area Hospital
District Hospital
Outpatient visits
Observed Estimate additional visits
0 20,000 40,000 60,000 80,000Immunization doses
Primary Health Centre
Community Health Centre L3
Community Health Centre L2
Area Hospital
District Hospital
Immunization Doses
Observed Estimate additional doses
0 10,000 20,000 30,000 40,000Inpatient visits
Primary Health Centre
Community Health Centre L3
Community Health Centre L2
Area Hospital
District Hospital
Inpatient visits
Observed Estimate additional visits
L2: Level 2; L3: Level 3
DELIVERIES IMMUNIZATION DOSES
INPATIENT VISITSOUTPATIENT VISITS
5150
Conclusions and policy implications
To achieve its mission to “expand the reach of health care and establishing universal health coverage,”1 India has strived over the past 10 years to expand and strengthen the public
sector of health care, with a focus on reaching rural areas. The country recognizes disparities and has sought to en-act policies and implement programs to expand access to essential and special services for marginalized groups. Our findings show that these goals are ambitious but at-tainable, if the country focuses on rigorously measuring health facility performance and costs of services across and within levels of care, and if it can align the different dimensions of health service provision to support optimal health system performance.
FacilitycapacityforserviceprovisionOptimal health service delivery, one of the key build-
ing blocks of the health system,2 is linked to facility capacity to provide individuals with the services they need and want. With the appropriate balance of skilled staff and supplies needed to offer both essential and spe-cial health services, a health system has the necessary foundation to deliver quality, equitable health services.
The availability of a subset of services including immu-nization, DOTS treatment, OBGYN services, laboratory services, and general medicine was generally high across facility types in Madhya Pradesh, reflecting the expansion of these services throughout the state. However, differ-ences remain between high- and lower-level platforms. For example, while STI/HIV services were available in most district hospitals, they were available in only 24% of community health centers. Moreover, substantial gaps were identified between facilities reporting availability of these services and having the full capacity to actually deliver them. While almost all facilities, across platforms,
1 Planning Commission Government of India. Twelfth Five Year Plan (2012-17). New Delhi, India: Government of India, 2012. 2 World Health Organization (WHO). Everybody’s Business: Strengthening health systems to improve health outcomes: WHO’s Framework for Action. Geneva, Swit-zerland: WHO, 2007.
indicated that they provided routine delivery care, only 29% of district hospitals and no lower-level facilities had the full stock of medical supplies and equipment to opti-mally provide these services as relevant to that platform. These gaps were also evident for ANC and general sur-gical services, and in all facility types, though they were more pronounced at lower levels. In general, district hos-pitals were well equipped with medical and laboratory equipment, though with limited capacity for imaging ser-vices. The availability of equipment declined through the levels of the system, particularly with regard to laboratory equipment and imaging equipment. Closing these gaps and making sure that all facilities are fully equipped to optimally provide essential services warrants further policy consideration. Chronic diseases (e.g., cardiovascular diseases, mental health disorders, diabetes, and cancer) and injuries are the leading causes of death and disability in India, and are projected to increase in their contribution to the burden of disease during the next 25 years.3,4,5 Much of the care for chronic diseases and injuries is provided in the private sector and can be very expensive.45 These study findings also document notably lacking NCD-related services at all levels of care, including cardiology, psychiatry, and chemotherapy. Only 53% of the district hospitals provide cardiology services, and only 6% report providing psychi-atric care. Such gaps in the health system will exacerbate disparities by not dealing appropriately with NCDs while continuing to strive to eliminate major infectious diseases like tuberculosis, HIV, and malaria, or to reduce neonatal and infant mortality. Furthermore, there also is a paucity
3 GBD 2015 Mortality and Causes of Death Collaborators. Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980–2015: a systematic analysis for the Global Burden of Disease Study 2015. The Lancet. 2016; 388:1459–1544. 4 Patel V, Chatterji S, Chisholm D, Ebrahim S, Gopalakrishna, G, Mathers C et al. Chronic diseases and injuries in India. The Lancet. 2011; 377: 413-28.5 GBD 2015 DALYs and HALE Collaborators. Global, regional, and national disability-adjusted life-years (DALYs) for 315 diseases and injuries and healthy life expectancy (HALE), 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. The Lancet. 2016 Oct 7; 388:1603–1658.
C O N C L U S I O N S A N D P O L I C Y I M P L I C AT I O N S
of essential equipment for NCD services, including glu-cometer/test strips and blood chemistry analyzer. While functional ultrasound machines were present in 88% of district hospitals, they were notably lacking in civil hos-pitals (40%) where they are still considered essential. Furthermore, though functional CT scans are considered essential, they were available only in 29% of the district hospitals. These findings support the need for immedi-ate action to scale up interventions for chronic diseases through improved public health and primary health care systems that are essential for the implementation of cost-effective interventions.45
According to recent studies, India has a severe short-age of human resources for health: it has a shortage of qualified health workers, and the workforce is concen-trated in urban areas.6 In the context of a shortage of qualified health personnel at all levels of the health sys-tem, but especially rural areas,7,8,9 results reveal disparate staffing patterns between facilities. Hospitals employ a large number of staff. At the lower, community levels, paramedical staff including nurses and ANMs provide the majority of care to patients (based on reported staff-ing). These staffing patterns are not unexpected in the hierarchy of care. However, nurses do not have much authority or say within the health system, and the re-sources to train them are still inadequate. A call has been made to the government to urgently address the issues of human resources through a comprehensive national policy for human resources to achieve univer-sal health care in India.48 However, it should be noted that despite the shortfall in human resources, the study findings suggest suboptimal efficiency in production of services with the given level of human resources.
6 Rao M, Rao KD, Kumar AK, Chatterjee M, Sundararaman T. Human resources for health in India. The Lancet. 2011; 377(9765): 587-98.7 Planning Commission Government of India. Twelfth Five Year Plan (2012-17). New Delhi, India: Government of India, 2012.8 Hazarika I. Health Workforce in India: Assessment of Availability, Production and Distribution. WHO South East Asia Journal of Public Health. 2013; 2(2): 106-112.9 Rao M, Rao KD, Kumar AK, Chatterjee M, Sundararaman T. Human resources for health in India. The Lancet. 2011; 377(9765): 587-98.
InfrastructureandequipmentAdequate operational infrastructure is essential for the
functioning of a facility, which in turn affects the efficiency of service provision. In Madhya Pradesh, all hospitals and community health centres and almost all primary health centres had access to functioning electricity, and no fa-cilities reported being solely dependent on a generator. This means a higher quality of service provision, as it al-lows for reliable storage of medications, vaccines, and laboratory samples. Access to piped water was universal at hospitals but more variable at health centres. This be-ing said, the majority of facilities across these platforms reported having flush toilets. That so many facilities re-ported access to essential resources like water, sanitation, and electricity likely reflects India’s commitment10,11 to upgrade all facilities so they meet Indian Public Health Standards. However, less than half of sub health centres had functional electricity and only 9% had piped water. This suggests that there should be a sustained focus on making sure that these resources reach the lowest levels of the health system. Communication is also an important facet of health ser-vice delivery. Limited facilities reported access to a landline phone; however, it is important to point out that mobile phones are widely available and in use by the staff. Access to emergency vehicles was also generally low: only 4% of primary health centres had an emergency vehicle available and 94% had no vehicle available at all. To address this, Madhya Pradesh now has a reasonable network of 108 and 102 ambulances which are readily available for use in emergencies.12,13
10 Planning Commission Government of India. Eleventh Five Year Plan (2007-12). New Delhi, India: Government of India, 2007.11 Planning Commission Government of India. Twelfth Five Year Plan (2012-17). New Delhi, India: Government of India, 2012. 12 India Infoline News Service. GVK EMRI 108 Neonatal Ambulances in Madhya Pradesh. IIFL [cited 2017 Nov 1]. Available from: https://www.indiainfoline.com/article/news-corporate/gvk-emri-108-neonatal-ambulances-in-mad-hya-pradesh-113103107152_1.html 13 Afridi S. Soon, dial 102 to get ambulance service in MP, medical advice 24x7. Hindustan Times. 2015 Feb 03 [cited 2017 Nov 1]. Available from: http://www.hindustantimes.com/bhopal/soon-dial-102-to-get-ambulance-service-in-mp-medi-cal-advice-24x7/story-uCLEt4khR9og7mm7tdGrVJ.html
5352
A B C E I N M A D H YA P R A D E S H
FacilityproductionofhealthservicesOverall, the number of outpatient visits by year and
platform was stable over the five years of observation. Outpatient visits were considerably lower at the lower health facilities. The volume of inpatient visits and deliver-ies increased slightly over the five years of observation for most platforms. The highest volumes of visits were held by district hospitals, followed by civil hospitals. Facility ex-penditure is dominated by personnel costs – accounting for, on average, at least half of total costs.
Efficiency scores reflect the relationship between facility-based resources and the facility’s total patient volume each year. Average efficiency scores by platform ranged from 34.8% to 78.2%, indicating patient volume could substantially increase with the observed levels of resources and expenditure. Within each platform, there is great variation in the efficiency of health facilities be-tween and within districts. With this information, we estimated that facilities could substantially increase the number of patients seen and services provided each, based on their observed levels of medical personnel and resources. As India seeks to strengthen public sector care to reduce the heavy burden of out-of-pocket expendi-tures,14,15 stakeholders may seek to increase efficiency by providing more services while maintaining personnel, ca-pacity (beds), and expenditure.
Further use of these results requires considering ef-ficiency in the context of several other factors, including quality of care provided, demand for care, and expedi-ency with which patients are seen.
The policy implications of these efficiency results are both numerous and diverse, and they should be viewed with a few caveats. A given facility’s efficiency score cap-tures the relationship between observed patient volume and facility-based resources, but it does not reflect the expediency with which patients are seen; the optimal provision of services; demand for the care received, and equity in provision of services to serve those who are dis-advantaged.16 These are all critical components of health service delivery, and they should be thoroughly consid-ered alongside measures of efficiency. On the other hand,
14 Planning Commission Government of India. Eleventh Five Year Plan (2007-12). New Delhi, India: Government of India, 2007.15 Kumar AKS, Chen LC, Choudhury M, Ganju S, Mahajan V, Sinha A et al. Financ-ing health care for all: challenges and opportunities. The Lancet. 2011; 377: 668-79. 16 UNICEF. Narrowing the gaps: The power of investing in the poorest children. New York, NY: UNICEF, 2017.
quantifying facility-based levels of efficiency provides a data-driven, rather than strictly anecdotal, understand-ing of how much Madhya Pradesh health facilities could potentially expand service provision without necessarily increasing personnel or bed capacity in parallel.
CostsofcareAverage facility expenditure per year differed sub-
stantially across platforms. We were unable to estimate the costs of care by type of services (such as outpatients, inpatients, deliveries, immunization, etc.) or by type of disease/condition (such as TB, diabetes, etc.), as such data are not readily available at the facilities. Estimating such costs of care and identifying differences in patient costs across the type of platforms is critical for isolating areas to improve cost-effectiveness and expand less costly services, especially for hard-to-reach populations.
Nevertheless, these results on expenditures offer in-sights into each state’s health financing landscape, a key component to health system performance, in terms of cost to facilities and service production. While these costs do not reflect the quality of care received or the specific services provided for each visit, they can enable a compelling comparison of overall health care expenses across states within India. Future studies should aim to capture information on the quality of services provided, as it is a critical indicator of the likely impact of care on patient outcomes.
PatientperspectivesPatient satisfaction is an important indicator of pa-
tient perception of the quality of services provided by the health care sector.17,18 Evaluation of services by pa-tients is important for purposes of monitoring, increasing accountability, recognizing good performance, and adapting patient-centric services, and for utilization of services and compliance with treatment. This report examined patient perspectives at public facilities; a ma-jor strength of this study is that patient satisfaction was assessed across the various levels of public sector health care.
17 Mpinga EK, Chastonay P. Satisfaction of patients: a right to health indicator? Health Policy. 2011; 100(2-3):144-150.18 Baltussen RM, Yé Y, Haddad S, Sauerborn RS. Perceived quality of care of prima-ry health care services in Burkina Faso. Health Policy Plan. 2002; 17: 42-48.
C O N C L U S I O N S A N D P O L I C Y I M P L I C AT I O N S
The public health system in India is designed as a refer-ral hierarchical system to provide a continuum of health care, and as a consequence of this, failure at one level can impact the chain of care at another level.19 Although var-ious government initiatives have led to improved basic service delivery at primary care health facilities over the last few years, still a large number of patients directly visit higher-level facilities, leading to overcrowding of those facilities,20 which impacts quality of care as it stretches facility resources in terms of both infrastructure and staff. In addition, the persistent shortage of medical staff in public facilities only aggravates the crowded condition at these facilities.21
The findings of this study indicate that patients were generally satisfied with the care they received, and rat-ings and satisfaction were highest at the highest levels of care. However, many were not satisfied with the cleanli-ness or privacy at the facility they visited. Holding other factors constant, patients with wait times longer than 30 minutes were less satisfied with care from doctors. Most patients experienced short travel and wait times. Most patients traveled for less than 30 minutes to receive care, with patients at lower-level facilities reporting the shortest travel times. District hospitals had the highest proportion of patients who had to wait more than 30 minutes to re-ceive care; the lowest proportion of patients waiting more than 30 minutes were at primary health centres. Finally, fewer than 3% of patients at all levels reported being un-able to acquire prescribed drugs. Though these levels are encouraging, ensuring that all patients may obtain pre-scribed medications at the time of their visit should be a priority, as it facilitates adherence and continuity of care.
With the developmental priorities for the government of India clearly highlighting the need to increase user participation in health care service delivery for better ac-countability,22 understanding how patients perceive the quality of the existing public health services encompass-ing various dimensions of care, such as time to receive medical attention, staff behavior, etc., could contribute to
19 National Health Mission, Ministry of Health and Family Welfare, Government of India. Framework for Implementation National Health Mission (2012-2017). New Delhi, India: Government of India, 2012. 20 Bajpai V. The Challenges Confronting Public Hospitals in India, Their Origins, and Possible Solutions. Advances in Public Health 2014; 2014: 27. 21 Rao M, Rao KD, Kumar AK, Chatterjee M, Sundararaman T. Human resources for health in India. The Lancet. 2011; 377(9765): 587-98.22 Planning Commission, Government of India. Faster, sustainable and more inclusive growth: An approach to the Twelfth Five Year Plan. New Delhi, India: Government of India, 2012.
developing strategies to improve performance and utili-zation of the public health system.23
HealthinformationsystemThis study was dependent on data availability at the
facilities for the various inputs and outputs. Because of the vast extent of data that were collected for five finan-cial years across the facilities, there were several lessons regarding the common bottlenecks within the health in-formation system, both at the facility level and at the state level. In general, there is less availability of staff to capture data and also weak staff capacity for data capture, man-agement, and use (interpretation or planning) at all levels. No system of regular review of data at the facility level that could guide planning or improvement of service pro-vision was observed.
It is not possible to assess the outputs by disease/condition other than those for deliveries, as data are not captured or collated by disease groups at the facilities. At the higher-level facilities, collation of patients seen at the facilities was not readily available, and it was not possible to assess the level of duplication of patients across de-partments. Furthermore, documentation of patients as a new patient or a follow-up patient was neither standard-ized nor practiced across most health facilities. Therefore, data interpretation is possible only in terms of number of visits and not in terms of number of patients.
Data were either incomplete or inaccurate at some fa-cilities for expenditure, patient-related outputs, and staff numbers. In a significant number of facilities, in case of staff turnover the previous staff did not hand over all the documentation of previous years to the new staff, which added to non-availability of data. Expenditure documen-tation had the most bottlenecks, with these data available across various sources for a given facility. It is not possible to document the expenditures at a given facility without procuring relevant data from the facility, a higher level of facility (block level), district health society, and from the state. The most limited capacity was to capture the expen-diture on drugs, medical consumables, and supplies.
23 World Health Organization (WHO). Global Health Observatory Data Repository. Geneva, Switzerland: WHO, 2016.
5554
A B C E I N M A D H YA P R A D E S H
SummaryThe ABCE project was designed to provide policymak-
ers and funders with new insights into health systems and to drive improvements. We hope these findings will not only prove useful to policymaking in the state, but will also inform broader efforts to mitigate factors that im-pede the equitable access or delivery of health services in India. It is with this type of information that the individual building blocks of health system performance, and their critical interaction with each other, can be strengthened. More efforts like the ABCE project in India are needed to continue many of the position trends highlighted in this report and overcome the identified gaps. Analyses that take into account a broader set of the state’s facilities, including private facilities, may offer an even clearer pic-ture of levels and trends in capacity, efficiency, and cost. Continued monitoring of the strength and efficiency of service provision is critical for optimal health system per-formance and the equitable provision of cost-effective interventions throughout the states and in India.
5756
FACILITY INFORMATION INPUTS(BEDS&STAFF) OUTPUTS EXPENDITURE
District Platform Facility Year Beds Doctors Nurses Paramed Nonmed Outpatient Inpatient Vaccinations Births PersonnelInfrastructure+