Evaluation of the Direct Health Facility Financing Program in Improving Maternal Health Services in Pangani District, Tanzania

Background: Maternal morbidity and mortality remain significant public health concerns globally, with Tanzania reporting 398 deaths per 100,000 live births annually. While national level data provide some insights into the issue, a focus on sub-national levels is required because of differences in contexts such as rural-urban disparities in maternal mortality. This study examined Direct Health Facility Financing (DHFF) and its effects on the quality of maternal health services in Pangani, a rural district in Tanzania. Methods: This study was conducted in Pangani district of Tanga region in Tanzania. The study used both qualitative and quantitative methods, including 16 in-depth interviews with the council health management teams, facility in charges, maternity nurse in charge, and 5 focus group discussions with community health governing committee members. The number of deliveries that occurred in health facilities, as well as medical supplies, equipment, and reagents purchased by the facilities, were compared using descriptive statistics before and after the DHFF implementation. Results: Direct disbursement of funds from the central government through the Ministry of Finance and Planning to the primary health facilities reduced delays in procurement, improved community outreach services, and improved community leaders’ engagements. Deliveries occurring at health facilities increased by 33.6% (p < 0.001) one year after the HDFF implementation. Various medicines, delivery kits, and some reagents increased significantly (p < 0.05). However, the lack of computers and poor internet connectivity, an insufficient supply of medical equipment and unstable stock of the Medical Stores Department increased the difficulty of obtaining the missed items from the selected prime vendor. Conclusion: Overall, this study shows a positive impact of the DHFF on maternal health service delivery in Pangani district. Specifically, an increase in the number of medical supplies, equipment, and reagents necessary to provide maternal health services contributed to the observed increase in facility deliveries by 33.6%. Moreover, the system minimizes unnecessary delays in the procurement processes of required drugs, supplies, and other facility reagents. To maximize the impact of the HDFF system, lack of computers, unstable internet, limited knowledge of the staff about the system, and inadequate health workforce should be addressed. Therefore, strengthening the DHFF system and staff training in-service and on the job is essential for smooth implementation.

This study was conducted in Pangani district, geographically located in northeast Tanzania. The district is generally rural, and according to the 2012 Tanzania Population and Housing Census, it had a population of 54,025 people. The female population numbers 27,155, which is 50.27% of the total population, with an average household size of approximately 4 persons. The population density was 24 persons per square kilometre, with about 21% of the population living in a semi-urban centre. The main social-economic activities are cash crops, food crops, livestock keeping, poultry keeping, fishing, and small-scale industries. In February 2018, Tanzania officially launched the DHFF system policy.47 This policy aims to devolve planning, budgeting, and implementation responsibilities to primary health providers. The program was introduced to meet the following goals: improving the structural quality of maternal and child health services, increasing accountability and governance in the health system at the primary health-care facilities, facilitating increasing health system responsiveness to patients who receive health-care in the respective health facilities, and improving health-seeking behaviour and service utilisation at the primary health facility level.41,48 This is a marked departure from the previous system where planning and budgeting were done by DMOs. Another aim of the DHFF is to improve the sense of community ownership of health-care provision by including community representatives in health facility governing committees.47,49 Under the DHFF program, the Council Health Service Boards and health facility governing committees have been given the primary responsibility to undertake planning, budgeting, and implementing health service delivery at the local level.47,49 Before the DHFF, disbursement of funds to Primary Health facilities passed through 2 levels; from the central government through the Ministry of Finance and Planning to the President’s Office-Regional Administration and Local Government and then the respective Local Government Authorities.41,50 Under local government authorities, the DMOs received the health budgeted funds, and disbursed the funds to primary health-care facilities.51 Delays were sometimes encountered in disbursements of the funds by DMOs, which negatively impacted the delivery of health services, including those for vulnerable women and led to negatively impacted service delivery by consumers.47,52 The DHFF system acts as a bridge between central government (Ministry of Finance and Planning) and local government authorities the primary health facilities in which funds are directly transferred from the Ministry of Finance and planning to frontline primary health facilities.53 The shift aimed to increase service delivery efficiency by reducing leakages and enhancing predictability and timeliness of funds received at primary health facilities.41,54 PlanRep is a planning and reporting database used by local government authorities during planning and reporting, facility budget is governed by previous financial year annual revenue collection as budget sealing cut off point from central government to local government and partly is driven by consideration of population size of respective health facilities.53,54 In the four-year lifespan of the DHFF in Pangani, there has not been a thorough assessment of its impact on maternal health services.47 This study attempts to make such an evaluation, and we consider that the implementation of the DHFF is timely to improve the quality of and access to maternal health services. Under the DHFF, the facility therapeutic committee must consider previous stock and available funds when planning the procurement of maternal drugs such as folic acids, iron supplements, magnesium sulphate, and sulfadoxine-pyrimethamine (SP) for malaria treatment; medical reagents such as syphilis test, malaria rapid diagnostic test, and deep urine stick for protein; medical equipment like blood pressure machine, weighing scale machine, fetal scope, and delivery kit. The facility pharmacist prepares the Medical Stores Department (MSD) orders through an electronic logistic management information system. Once approved by the medical officer in charge of the facility, the order is sent to MSD. MSD prepares a pro forma invoice for available and out-of-stock items, which are then sent to the facility for payment of available items. Out-of-stock items are sourced from approved private vendors by the procurement management unit, which issues a local purchase order to the vendor.55 Once the facility therapeutic committee and community health governing committees approve items delivered by the vendor, payment is then made. This study used a mixed-methods approach, utilizing both qualitative and quantitative data. The qualitative data were based on interviews and focus groups, hence consistent with the Consolidated Criteria for Reporting Qualitative Research (COREQ).56 The quantitative data compared births occurred at health facilities before and after the DHFF intervention. The study the used financial year 2016/2017 as a source for baseline data for the period before introducing DHFF and the financial year 2017/2018 as the follow-up year after the intervention. To evaluate outcome on facility deliveries, availability of essential drugs, medical equipment, and reagents for maternal health and quality of care provided for maternal health, fidelity of implementation framework57–59 was used. The study was implemented based on the structured conceptual framework as shown in Figure 1. The framework was adapted from Ntuli’s47 to enable the establishment of the effect pathway relationship to determine the trend and progress of DHFF interventions. Theoretical framework for DHFF implementation in the study area. The study used a purposive sampling approach with stratification to select participants from five5 health facilities, namely, Pangani district hospital, Mwera health center, and three dispensaries: Masaika, Kwakibuyu, and Mkalamo. A reason for choosing these facilities was to have a representative sample of health facilities across different levels of service provision at the district level: hospital, health center, and dispensary. Also, the study selected these facilities because they had a high volume of deliveries compared with others of the same level within the district. Therefore, the choice of these facilities enabled us to access a large quantity of information related to our study objectives. The number of deliveries from the facilities before HDFF (2016/17) and one year thereafter (ie, 2017/18) was 1189 and 1589, respectively (Table 1). Number of units of different medical supplies, equipment and reagents are shown in Table 2. Status of Facility Deliveries Before and After the DHFF Implementation in Pangani District Notes: “A” represents financial year 2016/17 and “B” represents financial year 2017/18. P-value <0.05 indicates that the difference is statistically significant. Status of Medical Supplies, Medical Equipment, and Medical Reagents Before and After the DHFF Implementation in Pangani District Notes: “A” represents financial year 2016/17 and “B” represents financial year 2017/18. P-value <0.05 is statistically significant. Data collection procedures for both quantitative and qualitative data involved visiting each study facility and meeting all key participants and key informants to capture information relevant to the study objectives. This was consistent with the principles for process evaluation and mixed methods research.60 Quantitative data were collected to assess changes over time from facility registry for 2016/2017 as a baseline year before the introduction of DHFF and 2017/2018 as follow-up year after the DHFF intervention was implemented. Health-care workers and health governing committee members from each study facility were interviewed based on their experiences before and after introduction of DHFF to describe changes over time. The qualitative data was collected by interviewing the key informants, through in-depth interviews and focus group discussions (FGDs). The reason why we have conducted qualitative part of our study is to explore the reality of DHFF system implementation success and challenges direct from primary health facilities. During the development of the questionnaire and interview guide, different steps were taken as stated below: Participants were informed and consented to the publication of anonymised responses. Data collection tools were piloted for five days by the research team in Bagamoyo district. The district was selected for the pilot survey because it is similar in characteristics with Pangani district where the actual study was conducted. All challenges and observations noted during the pilot survey were addressed before the actual field data collection. Qualitative data analysis was undertaken using a framework grouping of relevant themes that answers key issues as per study objectives. Transcripts were managed and coded in NVIVO 11 by the use of the framework analysis approach.61 Coded data were reviewed, organized into charts according to research themes focusing on the study objectives and emerging issues from the interviews. After development of the questionnaire and interview guide, different purposive selections of respondents were selected in different sites. After obtaining their consent, we conducted interview using the entire respective questionnaire and interview guide. During this pilot/pre-testing we identified questions that needed to be adjusted and/or improved. At each facility, the study extracted retrospective information from the health facilities’ information system books: book number 4 – ledger book, book number 6 – antenatal care register, book number 12 – labors, and delivery book, and book number 13 – postnatal register. Data was also extracted from the facility Health Management Information System registry. The variables recorded were services provided, medical equipment, maternal drugs, and maternal medical reagents. This study conducted 16 In-depth Interviews (IDIs) with health-care workers and 5 Focus Group Discussions (FGDs), which comprised 8–12 participants. The participants were community health governing committee members from each facility. The principal investigator (ie, Samwel Tukay) used a guiding tool during the collection of qualitative data. There was no prior relationship between the participants and the research team. The investigator took notes during IDIs, and FGDs sessions and all sessions were audio-recorded to facilitate transcription of the information. Digital recordings of interviews were made and transcribed in Kiswahili and then translated to English. The study used digital voice recorder, pen, pencils, and notebooks to capture participants’ information and critical issues to answer research objectives. All the audio interviews were transported from the audio recorders to the laptop for further processing. All the interviews were transcribed word for word by the researcher. After transcription, all the transcripts were cross-checked in relation to the audio recording to ensure all the information have been captured correctly as it is in the audio without omission of any information. All the correct transcripts were then translated from Swahili to English by an experienced translator, and all the translated transcripts were also checked for clarity before analysis. The quantitative data analysis was done descriptively using Stata (version 12) statistical software. The number of deliveries and those of different medical supplies, equipment, and reagents were compared before and after DHFF implementation using a two-sample test of proportions. Qualitative data analysis was undertaken using framework analysis. We grouped relevant themes that answer key issues as per study objectives. Coded data were reviewed, organized into tables according to research themes focusing on the study objectives and issues emerging from the interviews. Further details about the coding are presented in the data analysis by framework grouping (Table 3). Data Analysis by Framework Grouping

N/A

• tukay, samwel marco
• pasape, liliane
• tani, kassimu
• manzi, fatuma