Access to sexual and reproductive health commodities in East and Southern Africa: A cross-country comparison of availability, affordability and stock-outs in Kenya, Tanzania, Uganda and Zambia

Background: Access to sexual and reproductive health services continues to be a public health concern in Kenya, Tanzania, Uganda and Zambia: use of modern contraceptives is low, and unmet family planning needs and maternal mortality remain high. This study is an assessment of the availability, affordability and stock-outs of essential sexual and reproductive health commodities (SRHC) in these countries to inform interventions to improve access. Methods: The study consisted of an adaptation of the World Health Organization/Health Action International methodology, Measuring Medicine Prices, Availability, Affordability and Price Components. Price, availability and stock-out data was collected in July 2019 for over fifty lowest-priced SRHC from public, private and private not-for-profit health facilities in Kenya (n = 221), Tanzania (n = 373), Uganda (n = 146) and Zambia (n = 245). Affordability was calculated using the wage of a lowest-paid government worker. Accessibility was illustrated by combining the availability (≥ 80%) and affordability (less than 1 day’s wage) measures. Results: Overall availability of SRHC was low at less than 50% in all sectors, areas and countries, with highest mean availability found in Kenyan public facilities (46.6%). Stock-outs were common; the average number of stock-out days per month ranged from 3 days in Kenya’s private and private not-for-profit sectors, to 12 days in Zambia’s public sector. In the public sectors of Kenya, Uganda and Zambia, as well as in Zambia’s private not-for-profit sector, all SRHC were free for the patient. In the other sectors unaffordability ranged from 2 to 9 SRHC being unaffordable, with magnesium sulphate being especially unaffordable in the countries. Accessibility was low across the countries, with Kenya’s and Zambia’s public sectors having six SRHC that met the accessibility threshold, while the private sector of Uganda had only one SRHC meeting the threshold. Conclusions: Accessibility of SRHC remains a challenge. Low availability of SRHC in the public sector is compounded by regular stock-outs, forcing patients to seek care in other sectors where there are availability and affordability challenges. Health system strengthening is needed to ensure access, and these findings should be used by national governments to identify the gaps and shortcomings in their supply chains. The study was designed as a cross-sectional survey. Data collection comprised a health facility survey in which the availability, price, and stock-outs of SRHC were measured. Ethical approval was granted by the Amref Ethics and Scientific Review Committee in Kenya, the National Institute for Medical Research in Tanzania, Makerere University School of Health Sciences in Uganda, and the National Health Research Authority in Zambia. Letters of introduction to health facilities were provided by County Directors of Health in Kenya, and Ministries of Health in Tanzania, Uganda and Zambia. This survey was conducted in ten counties in Kenya, twelve counties in Tanzania, six regions in Uganda, and ten provinces in Zambia. The provinces selected included each country’s main urban region and five or more other regions, using a random sampling strategy. Each survey area within a province covered a population of 100,000 to 250,000. Health facilities were identified for inclusion, using a stratification method, as public-, private-, and private not-for-profit (PNFP) facilities. Within each stratum, four health facilities were randomly sampled from rural and urban areas. In this study urban areas were defined per country according to the definition held by the corresponding National Bureaus of Statistics: an urban area was defined in Kenya and Uganda as an area with a population of 2000 or higher, in Zambia with a population of 5000 or higher, and in Tanzania with a population of 10,000 or higher [29]. In each case, one of the selected urban areas included the main public provincial health facility. The inclusion criteria for the other health facilities were that facilities had to be within 3 h travel from the main public provincial health facility, and all selected health facilities had to provide SRH services. A data collection tool, adapted from the standardised World health Organization (WHO)/Health Action International (HAI) Medicine Prices Monitoring Tool and validated in many countries, was used for collecting data [30–34]. The ‘basket’ of commodities assessed was developed by combining the WHO’s Essential Medicines for Reproductive Health, the Interagency List of Essential Medicines for Reproductive Health, the Interagency List of Medical Devices for Essential Interventions for Reproductive, Maternal, Newborn and Child Health, and the United Nations Commission on Life Saving Commodities for Women and Children: Commissioner’s Report [35–38]. In combination with in-country expertise via a specialist advisory group and after piloting the methodology, after which slight alterations were made to the commodity basket, the commodities list presented was believed to be a selection of the most essential SRHC within the study region. Commodity strengths and dosage forms were based on the national essential medicine lists (NEMLs) [39–43]. Commodities cover family planning, maternal and child health, and STI management, and when listed with multiple dosage forms or strengths, all the formulations were included in the survey (see Additional file 1 for a complete overview of surveyed commodities). Previous cycles of the research took place in 2017 and 2018 in Kenya, Tanzania, Uganda and Zambia. Data collection took place in July 2019 using a mobile data collection application. In each country, local data collectors were trained by the authors (GIO and DK) on how to use the data collection tool during a two-day workshop organised by Health Action International, which included a field test. During the workshop the data collectors were provided with one tablet each and taught how to use the mobile application through a step-by-step walkthrough. During the field test they practiced the use of the mobile application. Data collectors worked in pairs, supervised in each country by a survey manager. Data on availability, patient prices, brand information and stock-out days was only collected when commodities were visibly present. Product name, name of manufacturer, actual pack size and pack price were recorded for the lowest price for each commodity available. Stock-outs were only recorded if a stock card was available and seen. Stock-outs were noted for the 6 months prior to the day of data collection. After completion of data collection, data was uploaded to the server and downloaded into an excel spreadsheet. Data entries were double-checked for accuracy by the survey managers and researchers. If data was incompletely or incorrectly entered, such as if a wrong product or pack size was noted, or a wrong unit price was calculated, the data was rectified after verification with the data collectors or or an ‘X’ was noted to denote only the availability of the commodity when pricing information could not be verified. Thereafter, analysis was completed in a previously developed Excel analysis tool using descriptive statistics. The availability of a commodity was calculated as the mean of the sampled facilities where the medicine was found at the time of the survey, expressed as a percentage. Mean availability of SRHC per sector and country was calculated in a two-step manner: firstly, the mean availability per commodity across the sampled facilities was calculated, after which the mean of these mean availabilities was calculated. For each commodity, availability was only measured when the level of care at which a commodity should be available corresponded with the surveyed facility. For example, calcium gluconate should be available at hospital levels and up in Kenya, Tanzania and Zambia, and from health centre III level in Uganda. In the PNFP sector, availability of family planning commodities was only calculated if family planning services were provided by the facility. Availability was calculated per commodity, as well as in groups for similar use (the birth control pill, injectable contraceptive and implant) or for different formulations of the same medicine (i.e. for magnesium sulphate, amoxicillin, clotrimazole, ferrous salt, folic acid, zinc and ORS sachets). When availability was calculated for a grouping of commodities, it was an aggregate of the availability and calculated as the mean percentage of sampled facilities where either of the formulations or commodities with similar medicinal use were available. Availability of 80% or higher was considered acceptable as per WHO guidelines [44]. Two-sample F-tests for variance were computed to test for normal distribution and independence, after which two-sample t-tests were calculated to test whether significant differences existed between means, using a significance cut-off value of 0.05. Stock-outs were calculated longitudinally as the mean percentage of facilities that reported a stock-out of a commodity any time in the 6 months prior to the day of data collection. Stock-out days were also calculated longitudinally over a six-month period and were calculated as the average number of days a commodity was stocked out per month. Stock information was surveyed only for medicines, not for medical devices. Affordability was calculated using the median price of a commodity, and the number of days a lowest-paid government worker (LPGW) needs to work in order to pay for a standard treatment regimen for a commodity. The daily wage of an LPGW was 449.40 Kenyan Shillings (Kenya), 3077.15 Tanzanian Shillings (Tanzania), 6169.65 Ugandan Shillings (Uganda), and 33.12 Kwacha (Zambia) [45–48]. According to the WHO/HAI methodology, treatment was considered unaffordable if it cost more than a day’s wage for an LPGW [30]. Affordability was calculated only for medicines, not for medical devices. Accessibility was illustrated combining the availability and affordability measures. This resulted in a categorical variable, in which accessibility was achieved when a commodity had an 80% or higher availability, and when a treatment regimen cost less than a day’s wage of an LPGW.