There is no printed copy available to order.
Utilisation of key reproductive, maternal and child health services has improved in Mali. However, the pace of improvement varies across services, and there is also a large disparity in service utilisation across regions. Service utilisation at the population level is determined by various factors such as demand for a service and, among those with demand for the service, ability to access quality care. To understand the determinants of service utilisation and trends in utilisation, it is crucial to study service utilisation in the context of access to quality care by using data from both the population and health facilities.
In 2018, Mali conducted a Demographic Health Survey (DHS) and a Service Availability and Readiness Assessment (SARA). This provided a relatively rare opportunity to study both the service environment and service utilization in the country. The 2018 Mali SARA collected data from a representative sample of facilities in the country, which provided information on the availability of facilities, availability of services among the sampled facilities, and service readiness among the facilities that provide specific services. The two elements of access to care that can be studied with data from SARA include geographic accessibility (by using a crude proxy indicator) and service quality.
This study aims to increase our understanding of the service environment and its association with service utilisation at the regional level for eight services, using data from the 2018 SARA and the 2018 DHS. The services include family planning, antenatal care (ANC), delivery care, childhood vaccination, child health services, malaria diagnosis and treatment, intermittent preventive treatment in pregnancy for malaria (IPTp), and counselling and testing for human immunodeficiency virus (HIV). For each of the select services, the specific aims are to (1) assess service readiness, by domain and average across domains; (2) examine service availability, both unadjusted and adjusted; and (3) explore associations between service utilization and various measures of service availability. In addition, we provide a region-specific summary of the service environment and utilization that will facilitate use of the study findings at the regional and national levels.
For each of the eight services, we constructed a service readiness score for the four domains of service readiness (staff and guidelines, equipment, diagnostics, and medicine and commodities). In addition, for each service, we created a summary service readiness score that averaged the four domain scores. Operational capacity among facilities that offer specific services varies greatly across services, although this generally ranges between 60 and 90 of the maximum score of 100. Across the eight service areas, readiness is lower in the staffing and guidelines and the diagnostics domains than in the equipment and medicines and commodities domains. Overall, childhood vaccination readiness is relatively high and consistent across the four domains. Among the regions, Bamako has relatively low service readiness across all service areas. In Kayes, Sikasso, and Mopti, the readiness score is higher than or close to the national average across the eight service areas.
We also calculated service availability (percent of facilities that offer specific services), which is relatively high for all services except HIV counselling and testing. We then calculated the service availability adjusted for service readiness (percent of facilities that offer a specific service with operational capacity to provide the service). The average adjusted availability is lower than the unadjusted availability by 22 percentage points across all regions and services areas. To address the large variation in facility density across regions, we further adjusted the availability by using the relative facility density score against the World Health Organization (WHO) benchmark (ranging from 0.45 in Mopti and Segou and 0.8 in Bamako). After the final adjustment, the ranking of regions changed substantially because regions with low availability and/or readiness scores have relatively higher facility density (such as Bamako) and vice versa. For each region, we produced detailed information on utilization, service readiness, and service availability, along with relative comparisons against the national average and regional ranges. Finally, we assessed the relationships between service utilization and the three service availability measures.
Although it is difficult to draw statistical inferences based on a small number of regions, we found no statistically significant linear correlation between utilisation and service availability in any service area. When we examine associations between utilization and service availability adjusted for readiness and facility density, statistically significant linear correlation was observed in two areas: a negative association with malaria IPTp (correlation coefficient: -0.74, p-value: 0.03) and a positive association with HIV counselling and testing (correlation coefficient: 0.72, p-value: 0.04).
It is important to note that any significant correlation based on observational data does not indicate causality. A negative association can reflect targeted interventions that can improve the service environment in regions with low utilization. The overall lack of associations with most services suggests the importance of other elements of access in service utilization, which are not addressed in our study. These include information, sociocultural acceptability, and affordability. We attempted to control for geographic accessibility, another important element for access, by adjusting service readiness with the relative density score across regions. However, since the geographic distribution of facilities differs from the geographic distribution of population in many regions, the aggregate regional-level facility density may not have been an effective measure for adjusting geographic access.
In summary, we assessed the service environment at the regional level with various quality and availability measures for eight select services in Mali. Although most services are commonly offered at facilities, readiness to provide specific services varies greatly across regions and services. Further, relatively low facility density indicates lower accessibility at the population level. When assessing the relationship between utilization and various service availability measures, we found significant associations in only a few services. We speculate that other elements of access to care contribute to differences in service utilization at the regional level. However, it is important to note that lack of association between the service environment and utilization at the population level does not imply that service quality is unimportant. Receiving high-quality service is a patient right, regardless of its impact at the population level.