Primary health care delivery in a Sidama community undergoing demographic transition

Abstract

Lav bruk av helsetjenester på den etiopiske landsbygd.

Rundt halvparten av verdens befolkning har liten tilgang på livsviktige helsetjenester. Og, spesielt ilavinntektsland,er registreringen av sykelighets, fødsels og dødelighetsdata ikke pålitelig på grunn av mangelfulle målinger. I Etiopia er gjennomsnittlig besøksrate for helseinstitusjoner på landsbygda lav, og spesielt i tradisjonelle samfunn er det et stort underforbruk av helsetjenester. I tillegg er det usikkerhet omkring både struktur og størrelse på befolkningen. Det hersker dermed usikkerhet om for eksempel antatall barn en kvinne føder er redusert og om befolknings reduserende tiltak som prevensjon har hatt en virking. Denne studien ble utført i to områder i Sidama regionen iSør-Etiopia. Formålet med avhandlingen har vært å måle hvorledes befolkingen utnytter det eksisterende helsesystemet, samt å vurdere den befolkningsmessige utviklingen i same område.

Doktorgradsarbeidet viser en meget lav utnyttelse av helsetjenester, dårlig diagnosekapasitet og lav forekomst av ikke-smittsom sykdomsforekomst. Bruken av helsetjenester varierte mellom ulike aldergrupper, hjemstedsadresse og helseinstitusjonstyper. Studien har bruket forskjellige metoder tilberegnetfødsels- og dødsratene i området, og den forventede levealder er lavere enn tidligere vist.

Avhandlingen anbefaler å bedre tilgjengelighet for alle deler av befolkningen. Helseinstitusjoner bør plasseres nærmere der folk bor. Videre bør en styrke den diagnostisk kapasitet og helsepersonells kompetanse, spesielt innen kroniske sykdommer. Det tilråså bedre kvaliteten på tilgangen på valide befolkningsdata.

Background: Ethiopia has been working towards achieving universal health coverage (UHC) through the expansion of primary health care units to the rural community. The primary health care delivery system is one of the three levels of the Ethiopian health care delivery system at the community level. Improving access to and utilisation of primary health care units helps the realisation of the ‘health for all’ agenda of the sustainable development goals (SDGs). Health service utilisation is considered both an influencer and consequence of population change. Yet, there are variations in health service utilisation across regions in Ethiopia. The population characteristics, such as births, deaths, age, and sex distribution, of the Sidama community have not been investigated since the 2007 census. Understanding the transitions in demographic variables and their impact in health service use can give insight to health planners at the local level. As the health-seeking behaviour of the community improves, morbidity, mortality and birth rates reduce, and life expectancy increases. However, measuring health care utilisation demands good quality population and utilisation data. Ethiopia has low coverage of registration of vital events, such as births and deaths. The quality of self-reported data also lacks accuracy. In addition, the health facility-based recoding system, Health Management Information System (HMIS), also lacks completeness and quality. Therefore, understanding health service utilisation from health facility data at a local level could give insight into how to improve the data quality and the health outcomes of the community they serve. Moreover, identification of measurement and reporting errors in health demographic sites could identify knowledge gaps that could be improved.

Objectives: The overall aim of this thesis was to measure health service utilisation from health facilities and demographic data in Dale and Wonsho districts of the Sidama region in southern Ethiopia. The specific objectives were to assess the health service utilisation and disease distributions (Paper I); to assess the health service utilisation and identify factors that account for the variation in health service utilisation at kebele level (Paper II); and to identify the demographic structure and identify some limitations to how the population dynamics are studied in rural districts in the Sidama region in southern Ethiopia (Paper III).

Methods: We conducted an institution-based cross-sectional design study from 1 July 2017 to 30 June 2018 for all patients visiting health facilities in 65 primary health care units in Dale and Wonsho districts of the Sidama region. We trained and employed fifteen local data collectors and two supervisors to collect secondary data from standard registers supplied by the Federal Ministry of Health. The number of patients who visited the health facilities was 81,129. The utilisation rate was calculated as visits per person per year for new cases. We calculated the odds ratio for health service use and proportions of diseases’ diagnoses. We also presented the findings to the health managers and incorporated their suggestions for health improvement in the study (Paper I). For Paper II, we conducted an exploratory ecological study design. Fifty-four kebeles under the catchment of the 65 health facilities were included. We used ArcGIS and SaTScan software to identify and explore the spatial distribution of health service utilisation. Linear regression was used in the analysis. For Paper III, a mixed-method cross-sectional study was conducted using baseline data of the newly established Dale-Wonsho Health and Demographic Surveillance System site in 2018. A total of 5,179 randomly selected households, having 25,144 individuals, were analysed. Death-related information was also collected from the traditional burial associations called iddir. Focus group discussions and in-depth interviews were used for qualitative data collection. Logistic regression, life tables, and age reliability indices were used in the analysis.

Results: There was a low annual health service utilisation in the study area. The outpatient health service utilisation was 0.18 visits per person per year (95% CI: 0.18–0.19), with a mean of 0.17 (range: 0.01–1.19) visits per person per year. There was an uneven distribution of the low health service utilisation among the 54 rural kebeles in the two districts (Papers I–II). The rate of utilisation varied between different demographic attributes. The rural population had lower odds of health service utilisation per year by 91%, compared to the urban population (OR = 0.09; 95% CI: 0.08–0.09). The health service utilisation by children in the age group of 5–14 years was lower than by children under five years of age by 78% (OR = 0.22; 95% CI: 0.21–0.23). Females used health services four times as often as males (OR = 4.17; 95% CI: 4.09–4.25). Health professionals’ opinions on improving the health service utilisation were categorised as budget allocations and alternative financial sources, sustainable drug and material supply, health workers capacity-building, quality of services, access to health facilities, public and stakeholders’ involvement and collaboration, and supervision and record-keeping (Paper I). Kebeles with health centres had a higher service usage. More than half of the kebeles were within a 10 km distance from health centres. Thus, as the distance from the health centre increased, the utilisation rate decreased (Paper II). Around three out of seventeen (17.9%; 14,847 of 83,148) diagnoses in all age groups were for febrile illnesses. Close to half the febrile cases were among children under five years of age (46.5%; 3,827 of 8,233). The most common registered diagnosis among children was pneumonia (36.7%; 2,635 of 7,184 new visitors). Non-communicable diseases were diagnosed rarely. Family planning services constituted the most frequently utilised service (Paper I). The population in which these studies were conducted showed a lower fertility rate. There was a total fertility rate of 2.9 children/woman and a crude birth rate was 22.8/1,000 population. We recorded a crude death rate of 5.2/1,000 population. Factors affecting the fertility level were age, residence and educational status. We identified measurement and reporting errors on age, birth, and death information. The life expectancy declined to an average of 53 years (range 48–58 years), after substituting national and regional mortality estimates (Paper III).

Conclusion: This thesis shows a low and unequal use of health services among population groups and lower-level geographic areas (kebeles). The number of non-communicable disease diagnoses at the primary level of care was very low. The fertility in Sidama is lower than in previous studies and is affected by age, education and residence. Moreover, there were measurement and reporting errors in the community and health facility data. The mortality rates are higher and the life expectancy is reduced, after adjusting for reporting errors. The low service usage, coupled with poor diagnostic capacity, might have affected the mortality measures, while relatively higher family planning services usage might have reduced the fertility rates. The health professionals suggested working towards making the health services of good quality, accessible, and affordable to the community to enhance the service use and improve the health outcomes of the local community.