Assessment Of Primary Health Care System Performance In .

Transcription

Health Systems & Reform, 2(4):302–318, 2016Ó 2016 Bill and Melinda Gates FoundationISSN: 2328-8604 print / 2328-8620 onlineDOI: 10.1080/23288604.2016.1234861Research ArticleAssessment of Primary Health Care SystemPerformance in Nigeria: Using the Primary HealthCare Performance Indicator Conceptual FrameworkDaniel H. Kress1,*, Yanfang Su122and Hong Wang1Bill & Melinda Gates Foundation, Seattle, WA, USAVeritas Health Systems, Seattle, WA, USACONTENTSIntroductionConceptual FrameworkData SourcesPHC Performance: Outcomes, Outputs, Service Delivery, andInputsSystemBenchmarking Nigeria’s PHC PerformancePolicy LandscapeConcluding RemarksReferencesAbstract—Health gains oftentimes associated with income growthhave been stubbornly slow in Nigeria in the past 25 years. Oneplausible reason for this stagnation is underperformance in thecountry’s primary health care (PHC) system. The Primary HealthCare Performance Indicators conceptual framework is used toexamine Nigeria’s PHC system and possible causes ofunderperformance. Analysis was conducted using a variety ofsources including recent facility level information from the WorldBank Service Delivery Indicators Survey. Results show that Nigeriahas a relative abundance of PHC centers, reasonable geographicaccess to PHC, and relatively high health worker density. However,the performance of the PHC system is hindered by (1) segmentedsupply chains; (2) a lack of financial access to PHC; (3) a lack ofinfrastructure, drugs, equipment, and vaccines at the facility level;and (4) poor health worker performance. Altogether, these factorsreflect two overarching system-level challenges—financing andgovernance—that are key root causes of the dysfunctions observedin the PHC system in Nigeria. Compared with peer Africancountries, Nigeria ranks low on nearly all PHC performanceindicators. The government has taken important steps to addressthese root causes of underperformance, but policy gaps remain inachieving sustainable and equitable provision of PHC for the peopleof Nigeria.INTRODUCTIONKeywords: inputs, Nigeria, outcomes, outputs, primary health care, system,service deliveryReceived 17 June 2016; revised 31 August 2016; accepted 3 September2016.*Correspondence to: Daniel H. Kress; Email: daniel.kress@gatesfoundation.org302Primary health care (PHC) is the backbone of a health system. Furthermore, quality PHC initiatives have been recognized as fundamental to improving health outcomes.1 Thestrength of a country’s primary care system was negativelyassociated with mortality in Organization for EconomicCooperation and Development countries,2 and PHC also hasimproved population health in low- and middle-incomecountries.3,4 The Declaration of Alma-Ata in 1978,5 the 1987

Kress et al.: Assessment of Primary Health Care System Performance in NigeriaBamako Initiative,6 and the 2006 Abuja Call7 all emphasizedthe importance of investing in PHC for health. Following theWorld Health Report “Health Systems: Improving Performance” in 2000,8 the World Health Organization’s WorldHealth Report “Primary Health Care (Now More ThanEver)” in 2008 asserted that PHC reforms can deliver equitable health services and secure the health of communities.9Given that PHC is essential to strengthening health systems10and achieving the Sustainable Development Goals,11 it isimportant to understand fundamental causes of underperformance of PHC systems.In this article, we examine the performance of Nigeria’sPHC system. The country emerged as Africa’s largest economyin 2014[a] with a per capita gross domestic product (GDP) of5,991 USD. However, compared with other countries, Nigeriahas underperformed on important health outcomes such aschild mortality (Figure 1). One of the potential reasons for thisis the poor performance of the country’s PHC system.Researchers have sought root causes of poor PHC coverage,focusing largely on two factors. First, many argue that poor performance is due to lack of sufficient health care facilities.12-15Some scholars have argued that low PHC coverage is a resultof an insufficient health workforce.14,16,17 The literature largelypoints to bottlenecks in primary health care inputs, including303health facilities and health workers, to explain Nigeria’s poorperformance in PHC coverage. However, these factors do notconvey the whole story. This article draws upon a holistic conceptual framework to examine Nigeria’s primary health caresystem. We explore the extent to which service delivery, inputs,health financing, and governance limits the performance ofNigeria’s primary health care system. The aim of this article isto identify root causes of PHC underperformance, highlightareas of future research, and provide a framework by which thefuture policy agenda can be shaped.In this article, we first introduce the conceptual framework, entitled the Primary Health Care Performance Initiative (PHCPI), followed by the data sources for analysis.Then, we examine each component of the PHCPI frameworkin the following order: outcomes, outputs, service delivery,inputs, and system. We also conduct country comparisons tobenchmark PHC performance in Nigeria. Lastly, we examinethe policy landscape, followed by concluding remarks.CONCEPTUAL FRAMEWORKWe have used the PHCPI conceptual framework to identifykey factors that contribute to low coverage of PHC in Nigeria(Figure 2). This framework is particularly useful because itFIGURE 1. Child Mortality Rate and GDP Per Capita for 173 Countries, 2015. Source: Adapted from Ref. 23

304Health Systems & Reform, Vol. 2 (2016), No. 4FIGURE 2. Primary Health Care Performance Initiative Conceptual Frameworkhighlights a critical area—service delivery—that has beenlargely neglected in PHC performance measurement.The PHCPI framework is based on several important priorsystems frameworks, such as the control knobs framework,18health system performance assessment,19 economic modelsof supply and demand, and Starfield’s key characteristics ofhigh-performing primary health care systems.20 The PHCPIconceptual framework reflects a structure similar to the commonly used input–process–output–outcome logic model,indicating logical relationships between constructs. Weincluded a “system” domain prior to the inputs domain toindicate the importance of the modifiable PHC system structure as emphasized in the control knobs framework. Furthermore, the framework exhibits an overall directionality ofinfluence, where the system domain influences the inputsdomain, which affects the complex interplay within the service delivery domain. Thereafter, successful service deliverycontributes to effective outputs, which subsequently affectoutcomes. In this article, we examine the system, input, andservice delivery constraints that are leading to underperformance in outputs and outcomes. There is directionality to theconceptual model, and in this article we chose to first highlight the end point of the model—outcomes—followed byeach previous component of the health system. We chose todo this because it is important to first understand the outcomes that need changing and then closely examine key rootcauses of the outcome, from most proximal to the most distal.We used a simplified version of PHCPI, focusing on keyidentified indicators. For example, due to data unavailability,we do not cover Starfield’s person-centered PHC servicedelivery, which is an important component in the originalPHCPI framework.DATA SOURCESWe use a variety of data sources in this article to understandPHC performance in Nigeria. These sources include theDemographic and Health Surveys for outcome indicators,21the Nigeria General Household Survey regarding PHCaccess,22 the World Development Indicators regarding poverty headcount,23 the World Health Organization (WHO)National Health Account for financing data,24 the WHOGlobal Health Workforce statistics for health worker densitydata,25 and the Advancing Child Health via Essential Medicine Vendors survey for Patent and Proprietary MedicineVendors (PPMVs) data.26 In addition to these data sets, wemainly rely on a relatively new data source (the Nigeria Service Delivery Indicator survey)27 for insights into what ishappening in health facilities.The Service Delivery Indicator (SDI) data were collectedthrough multicountry health facility surveys, allowing for acomparison between Nigeria and other countries when examining primary health care performance. SDI surveys havebeen carried out in Tanzania (2012), Senegal (2012), Kenya(2013), Nigeria (2013), and Uganda (2014). Table 1 showsthe sample size for each country.Though sampling strategies were adapted to eachcountry’s situation, the same general method (i.e., multistageclustered sampling) was used. The sampling strategy allowedfor disaggregation by geographic location (rural and urban)in all five countries and by provider type (public and private)in Uganda, Kenya, and Nigeria (only public health facilitieswere surveyed in Tanzania and Senegal). According to published World Bank SDI country reports, data are representative at the national level for Uganda and Kenya. Noinformation is provided on the issue of representativeness of

Kress et al.: Assessment of Primary Health Care System Performance in NigeriaNumber of ObservationsFacilitiesHealth WorkerAbsenceClinical iaKenyaTABLE 1. Sample Size by Country. Source: Adapted from Ref. 37the data at national level in Nigeria, but we feel that it ishighly unlikely that these data could be representative at thenational level given that data were collected in only 12 out of36 states.Table 2 summarizes the data modules in the SDI survey.In particular, provider ability was measured using clinicalvignettes, which are validated clinical cases that are designedto test provider knowledge for how to treat certain commonconditions associated with primary care. Using SDI datafrom 12 surveyed states in Nigeria, we generated nationaland state-level averages for key indicators. The quality ofinterstate comparisons in Nigeria is relatively high becauseof high levels of intrastate facility sampling.305and under-five mortality) for cross-validation and found thatinfant mortality has declined by 21% from 1990 to 2013, andunder-five mortality (U5M) declined by 34% over the sameperiod. Though this represents a decline, it is a decline that isslower than expected when compared to benchmark countries (i.e., Kenya, Uganda, Tanzania, and Senegal) over time(Figure 3). Furthermore, though Millennium DevelopmentGoal four targeted a U5M rate reduction by two thirdsbetween 1990 and 2015, Nigeria did not meet this target andunderperformed compared to peer countries.28 According toWHO estimation, U5M was 105 per 1,000 live births inNigeria in 2015, equaling about 760,000 deaths given thelarge population size in this country.The first step in assessing key root causes of the slow rateof mortality decline (the outcome in the PHCPI framework)is to look at coverage trends of key interventions (outputs inthe PHCPI framework).OutputsLooking back at trends in intervention coverage over the last25 years, the overall trend in intervention coverage is quiteflat (Figure 4), with 2013 coverage levels largely below 40%for each of the indicators. We now want to examine otherfactors in the PHCPI conceptual framework that will hopefully shed light on the persistently low levels of coverage.PHC PERFORMANCE: OUTCOMES, OUTPUTS,SERVICE DELIVERY, AND INPUTSService DeliveryOutcomesAccessThe mortality decline has not been as rapid as expected inNigeria. We used two indicators (i.e., infant mortality rateAlthough there are some isolated pockets where availabilityof services is limited, overall Nigeria appears to have aModule NumberModule TitleIntervieweeModule oneFacilityinformationHealth facility superintendent/most seniorhealth worker presentModule twoStaff rosterModule 2AFirst visitModule 2BSecond visitModule threeCase simulationsHealth facility superintendent/most seniorhealth worker presentHealth facility superintendent/most seniorhealth worker presentObservation on ten randomly selected healthworkersTen randomly selected health workersModule fourHealth facilityfinancingHealth facility superintendent/most seniorhealth worker present/accountantModule fiveExit interviewPatients exiting the facilityTABLE 2. Service Delivery Indicator Survey Instruments. Source: Adapted from Ref. 37ContentGeneral information about the facility, includinginfrastructure, equipment, materials andsupplies, and availability of drugsList of all health workers and theircharacteristicsMeasures availability of workers and theircharacteristicsClinician information, introduction, and sevenconsecutive clinical vignettesFinancial cash and non-cash support,expenditures, user fees, planning, andfinancial managementExit interview of patients

306Health Systems & Reform, Vol. 2 (2016), No. 4FIGURE 3. Mortality Trends Over Time. Source: Adapted from Ref. 21FIGURE 4. Long-Term Stagnation in Coverage of Basic Health Interventions. Source: Adapted from Ref. 21

Kress et al.: Assessment of Primary Health Care System Performance in Nigeria307FIGURE 5. Reasons for Absences. Source: Adapted from Ref. 38sufficient facility density and, as a result, most Nigerians havegeographic access to primary health care. This is confirmedwhen looking at the results from the General Household Survey (2013) that indicates that 75% of rural respondents residewithin two kilometers of a public PHC facility, and 95%reside within eight kilometers.22 However, financial access isa major challenge. The average cost of a public PHC visit is2.30 USD for child patients and 3.20 USD for adult patients(Table 3). However, it can go up to as much as 8 USD, whichis extremely burdensome for the 45% of Nigerians who liveon less than 2 USD a day and 28% who live on less than1.25 USD a day, according to World Development Indicators.23 User fees f

domain, which affects the complex interplay within the ser-vice delivery domain. Thereafter, successful service delivery contributes to effective outputs, which subsequently affect outcomes. In this article, we examine the system, input, and service delivery constraints that are leading to underperform-ance in outputs and outcomes. There is directionality to the conceptual model, and in this .