Retail drug shops market in Uganda incentives, effect on health care system, and implications for child health

Abstract

The under-five mortality rate in Uganda has drastically declined to 55 deaths per 1,000 live births as of 2018 compared to 175 deaths per 1,000 live births in 1990, although it is still high. Febrile conditions – malaria, pneumonia, and diarrhoea – account for the largest burden of these deaths, yet these conditions are preventable. The Sustainable Development Goals (SDG 3.2) target reducing under-five mortality to 25 deaths per 1,000 live births by 2030. Achieving the SDG target requires more effort in expanding access to and utilization of essential and quality health care for children more especially in regions such as Sub-Saharan Africa (SSA) where the burden is still high. In many low-and-middle-income countries, including Uganda, private retail drug shops play a crucial role in providing access to essential care for children and hence can contribute to reducing under-five mortality. There are already discussions on how to implement hitherto public sector interventions, into the retail drug shop in order to expand the benefits of these interventions. In Uganda, for example, there are studies that have shown that drug shops can implement Integrated Community Case Management of childhood illnesses (iCCM) intervention, and also provide a range of health care services, including, family planning. Furthermore, there are studies that have demonstrated that retail drug shops can use rapid diagnostic technologies to improve quality of care, among others. Given that iCCM in Uganda is implemented through the public sector and community health workers (CHWs), it presents a missed opportunity for many children who bypass CHWS and use retail drug shops as their first (and sometimes main) point of contact with the health system. While there is now focus on working with the retail drug shop to implement such interventions as iCCM, the retail health market is not yet comprehensively understood – in terms of market characteristics and the incentive structure – factors that partly influence retail health market operations and performance. Makerere University School of Public Health (MakSPH) together with partners implemented an intervention study aimed at introducing a modified iCCM at the private retail drug shops in rural areas of Mbarara and Bushenyi districts, in South Western Uganda, between 2013 and 2015. The aim of the intervention was to determine the feasibility and effect of iCCM on the quality, appropriateness of care, and rational use of medicines in registered private retail drug shops. This Ph.D. study was nested on the main intervention study (AXEX study). The Ph.D. study specifically aimed to characterize the retail drug shop market, associated incentives, and determine the effect of an iCCM intervention on availability and affordability of paediatric febrile illness care in the private sector registered drug shops in rural South-Western Uganda, in order to contribute to improved access and utilization of appropriate care and child survival. Methods. Four sub-studies (I – IV) constituting this Ph.D. thesis were conducted using cross-sectional quantitative design (sub-study 1), cross-sectional mixed methods design (sub-study II), quasi experimental design (sub-study III), and exploratory qualitative process evaluation design (sub-study IV). Sub-study I aimed to describe the operational characteristics of the private retail drug shop market in rural South Western Uganda. In this sub-study, we surveyed 74 registered private retail drug shops where we interviewed drug shop attendants using a structured questionnaire. In addition, we vi conducted 428 exit interviews with care-seekers who had sought care for under-five children at the drug shop during the study period. Descriptive analysis was conducted with means, medians, and proportions reported. The outcomes of interest included; training and skills of drug sellers, common childhood illnesses reported, the common medicines stocked and stocking decisions, prescription and dispensing decisions, use of diagnostics, drug shop operational environment, and type of clients served. Sub-study II aimed to analyse the incentives and motivations of private retail drug shop providers in rural South Western Uganda. To achieve this, a survey of 74 drug shops in both districts was conducted at baseline, with an additional detailed investigation done on a sub-sample of 20 drug shops using structured questionnaires with close-ended and open-ended items. In addition, we included four (4) in-depth interviews with government officials in the two study districts. The qualitative interviews supplemented the quantitative data. We conducted a descriptive analysis and reported the outcomes in terms of means, medians, and proportions as appropriate as possible. The qualitative information from the open-ended section of the structured questionnaire was manually analysed following deductive manifest and content analysis. The four key informant interviews were also analysed following a process that involved reading the transcripts several times, coding and generating themes, using thematic content analysis. The quotable quotations from the qualitative analysis supplemented quantitative results where necessary. The main outcomes of focus in sub study II included; motivations for start-up, incentives for drug shop operation, sales volumes, price mark-ups, profitability, and regulatory oversight, among others. Sub-study III aimed to determine the effect of implementing an integrated Community Case Management (iCCM) intervention at private retail drug shops on availability and affordability of paediatric febrile care provided at retail drug shops in rural South Western Uganda. This sub-study utilized data from a survey of 74 drug shops (Intervention=40, Comparison=34) conducted before the intervention and 48 drug shops (Intervention=32, Comparison=16) after the intervention. Supplementary data was extracted from 428 care-seeker interviews (Intervention=212, Comparison=216) and 553 exit interviews (Intervention=285, Comparison=268) conducted before and after the intervention respectively. Supplementary qualitative data were collected from 18 focus Group Discussions with caretakers of under-five children, 8 Focus Group Discussions with Village Health Teams, 47 in-depth interviews with drug sellers, and 19 in-depth interviews with government officials. The WHO/HAI standardized methodology of assessing the availability and affordability of medicines was used. We conducted a difference Difference-in-difference (DiD) analysis to ascertain the effect of the intervention on the availability and affordability of iCCM recommended medicines. In addition, we conducted a manifest and content analysis of qualitative, and the qualitative findings reinforced the quantitative findings. The main outcome measures were the proportion of outlets reporting availability of selected medicines; affordability of treatment regimens after comparing treatment costs with day’s wage rates of lowest paid government worker; and the proportion of care-seekers reporting affordability of treatment regimens. We also reported the views and opinions of care-seekers, caregivers, and VHTs, about the effect of the intervention on availability and affordability, among others. Sub-study IV aimed to describe and analyse the AXEX iCCM intervention, how it was implemented at the drug shops, and the dynamics, and processes through which the observed effects were produced. This sub-study was qualitative in nature and adopted a process evaluation design approach. Qualitative data were generated through key informants with district health officials and health workers, drug vii shop attendants, and FGDs with caregivers and VHTs. Deductive manifest and content analysis were used and a health markets framework proposed by Bloom and colleagues was adopted to guide the analysis. The main outcomes of focus were Perceptions and behaviours of stakeholders, effect on intervention on health system components, and factors that influenced intervention implementation. Summary Results: From Sub-study 1, it was found that most drug shops were operated by individuals with health related training mainly midwifery, nursing, clinical medicine, and to some extent pharmacy assistants. Relatedly, the physical environment within which most drug shops operate is consistent with recommended standards and guidelines. The most commonly stocked medicines by drug shops were paracetamol, quinine, cough syrup, ORS/Zinc, Septrin (co-trimoxazole), and ACTs. Our study also found that prescription and dispensing decisions and practices were largely influenced by market factors, including client demand, client preferences, client finances, and profitability. In sub-study II, it was found that retail drug shop start-ups were mainly motivated by the need to make a profit, although there were additional social benefits that accrued to drug shop operators, including the need to contribute to the community. Indeed, drug sellers enjoy social recognition and expanded social networks, and these social considerations partly influenced drug seller practices. It was also found that drug prices varied significantly across outlets with mark-up prices ranging between 43% and over 100%. It was estimated that drug shops make more than 90% per month after offsetting operational expenses. Finally, the retail health market regulatory system within the study area was found to be largely poorly streamlined and weak in capacity and under-resourced, and certain regulatory actions often resulted in perverse incentives to the operations of the retail market. In sub-study III, we found that overall; implementing an iCCM intervention increased the availability of iCCM recommended medicines for paediatric febrile conditions at the retail drug shops. A higher proportion of surveyed retail drug shop outlets reported a higher availability of medicines for paediatric febrile conditions in the intervention district compared to the comparison district, after implementing the intervention. Besides, a higher proportion of care-seekers in the intervention district reported that drug prices and costs of treatment were lower and within their ability to pay, compared to care-seekers in the comparison district, after the implementation of the intervention. Further analysis done for the intervention district shows that the intervention increased the affordability of medicines. The cost of treatment courses (doses) was lower and required less of the daily wage incomes after the intervention compared to before the intervention. Our sub-study IV found thatretail health markets constitute various actors and stakeholders, including community health workers, government health workers in health facilities (centers), district health officials, Ministry of Health officials, NDA, drug sellers and owners, care seekers, and the community leadership. These actors operate and interact in a dynamic system with rules, beliefs, and incentives that influence the nature and quality of interactions and ultimately goods and services provided. Each of the actors appreciated and benefited from implementing iCCM at the retail drug shop. For example, the district health team, the NDA, and MOH appreciated the partnership and collaboration with the project team in monitoring and supervision, which resulted in an improved relationship with drug sellers and improved quality of services. The drug sellers appreciated the training and capacity building, the free RDTs and subsidized supplies, the improved trust that resulted in providing better quality services, as well as better sales and client retention. The care-seekers and the community viii appreciated the quality of care received and the efficacy of the medicines. Finally, the intervention implementation demonstrated that private market-based interventions generate spillover effects to the broader health system beyond the targeted actors and these effects can be both intended and or unintended. For example, if services were improved at drug shops, there would be a shift from the formal health service providers to the retail drug shops, which could also have positive and negative results. Conclusions 1. Most registered shops surveyed operated in environments that generally meet the minimum National Drug Authority (NDA) guidelines, and most drug shops are operated by drug sellers with health-related training. The decisions about what medicines to stock or dispense (prescribe) were largely influenced by among other factors; demand and preferences of clients or care seekers; amount of finances that clients present with at the drug shop; and profitability objective. 2. While the need to earn profit was a significant motivator for retail drug shops, there are additional social benefits that accrue to drug shop operators including social recognition, expanded social networks, and a positive feeling of contribution to the community. In some circumstances, the social motivations generate intrinsic pressure that stimulates positive provider behaviour and action. 3. The regulatory system and framework for the retail drug shop market exist, and all key players appreciate the role and benefit of regulation. However, the system is weak, under-resourced, and not well structured to carry out its regulatory mandate. There were reported constraints to licensing, monitoring, and supervision, and in some cases, regulatory actions tended to stimulate perverse incentives to compliance. 4. Implementing iCCM at retail drug shops improved availability and affordability of treatment for uncomplicated malaria, pneumonia symptoms, and non-bloody diarrhoea for under-five children in a relatively low malaria transmission setting of South Western Uganda. 5. Identifying and implementing an appropriate package of market-based incentives and consumer empowerment actions can improve retail market practices and assure better health care provision as well as protect the public against negative practices that are detrimental to public and population health in a pluralistic health system. 6. Retail drug shops are part of a complex health system with multiple components and actors whose actions are influenced by multiple factors. The success of the AXEX intervention owes itself to the multi-component nature of the intervention which enabled the inclusion of different intervention packages each targeted at different actors and having different effects but all contributing to the objective of improving health care service delivery within the retail health market