Background: Tanzania is one of many low income countries committed to universal health coverage and Sustainable Development Goals. Despite these bold goals, there is growing concern that the country could be off-track in meeting these goals. This prompted the Government of Tanzania to look for ways to improve health outcomes in these goals and this led to the introduction of Payment for Performance (P4P) in the health sector. Since the inception of P4P in Tanzania a number of impact, cost-effective and process evaluations have been published with less attention being paid to the experiences of care in this context of P4P, which we argue is important for policy agenda setting. This study therefore explores these experiences from the perspectives of health workers, service users and community health governing committee members. Methods: A qualitative study design was used to elicit experiences of health workers, health service users and health governing committee members in Rufiji district of the Pwani region in Tanzania. The Payment for Performance pilot was introduced in Pwani region in 2011 and data presented in this article is based on this pilot. A total of 31 in-depth interviews with health workers and 9 focus group discussions with health service users and health governing committee members were conducted. Collected data was analysed through qualitative content analysis. Results: Study informants reported positive experiences with Payment for Performance and highlighted its potential in improving the availability, accessibility, acceptability and quality of care (AAAQ). However, the study found that persistent barriers for achieving AAAQ still exist in the health system of Tanzania and these contribute to negative experiences of care in the context of P4P. Conclusion: Our findings suggest that there are a number of positive aspects of care that can be improved by Payment for Performance. However its targeted nature on specific services means that these improvements cannot be generalized at health facility level. Additionally, health workers can go as far as they can in improving health services but some factors that act as barriers as demonstrated in this study are out of their control even in the context of Payment for Performance. In this regard there is need to exercise caution when implementing such initiatives, despite seemingly positive targeted outcomes.
The study was carried out in Rufiji, one of the seven districts in the Pwani region. Rufiji is a rural district and according to the 2012 national census, the district had a population of 217,274 [33]. Administratively, Rufiji is divided into 26 wards [33]. The main economic activity of the district is agriculture and 78% of the inhabitants actively participate in this sector. The main cash crops in the district are cashew nuts, coconut and simsim [34]. The district has a total number of 64 health facilities, including two hospitals, five health centres, and 57 dispensaries [34]. Like many rural districts in Tanzania, Rufiji district faces significant shortages of staff and of the 583 positions in the district health sector, only 301 are filled, a total shortage of approximately 49% [34]. Shortages of qualified staff is more pronounced at low level health facilities, i.e. dispensaries and health centres [35] and among all cadres except the medical attendant category which is overrepresented. Rufiji has a delta zone and during rainy season access to health facilities in the delta is difficult and these health facilities periodically face huge problems in procuring medical supplies and in maintaining regular communication with the district health offices located at the district centre Utete. A qualitative study approach was adopted using focus group discussions (FGDs) and in-depth interviews (IDIs). Data was collected by the first author with the help of a research assistant for a total of six months period, between January 2013–June 2016). Data was collected at 11 health facilities, including two hospitals, two health centres and seven dispensaries. Of these 11 health facilities, four were church-run and seven were public. As the study aimed to elicit the experiences and perceptions of health service users on access and quality of care, it was considered important to conduct FGDs with community members. Under this category of participants, we had discussions with mothers and health facility governing committee (HFGC) members. A total of nine FGDs with 44 participants were conducted, seven with mothers and two with HFGC members, for an overview, see Table Table1.1. All focus groups had an average of five participants, a number encouraged in the literature to increase the range of participation of each participant [36]. FDGs with HFGC members had least participants as these committees have up to five members in total. Mothers were targeted primarily for their in-depth experiences with maternal and child health services. In order to maximize the relevance and quality of our data from this group, we used two criteria to recruit them. Firstly, we targeted women who had at least two children with the older one being at least 5 years old. Secondly, we recruited women who had been residing in the same area (catchment area of a particular health facility) for at least 5 years. The participants were recruited with the help of health workers at health facilities. In practice, we ended up including some informants who could not meet one of the criteria; however the majority of our participants were able to meet the criteria. Overview of FGDs with health service users The second group of FGD participants was the health facility governing committee members. These committees represent community interests at health facilities. The committees were established in Tanzania in 1999, alongside Community Health Fund (CHF). When P4P pilot was introduced, the committees’ duties were expanded to include monitoring if P4P was implemented in ways protecting the interests of the community. The committees consisted of up to 5 members and many of them were very active in community activities and some even hold local political positions. In addition, the members are knowledgeable about the health status profile of their villages and the district. Typically, the person in charge of a health facility acts as the secretary of the committee. To avoid conflicts of interest the secretary was not invited into the FGD with the committee members. In FDGs, a topic guide was used with great flexibility to allow the discussion of emergent issues. The topic guide covered the following issue: experiences with MCH services, quality of health care, access and utilization of health services. To complement the data from the FGDs with community members, we conducted 31 IDIs with health workers of different cadres ranging from medical officers, assistant medical officers, clinical officers, nursing staff, laboratory staff and medical attendants were conducted. Table Table22 provides an overview of study informants. Overview of IDIs All FGDs and interviews except two interviews with Medical Officers were conducted in Swahili. The first author speaks colloquial Swahili, while the research assistant is a Tanzanian citizen with experience in qualitative health services research. All interviews and FGDs were recorded, transcribed in Swahili and then translated into English. In addition, rapid note taking was used. Translations were error checked. Qualitative content analysis was used as the mode of analysis [37], Table Table3,3, gives an overview of the analysis process. The transcripts were subjected to a thorough review before the coding exercise began. Meaning units at manifest level were identified and coded and from these sub-themes and themes were identified. OpenCode 3.6 software [38] was used for data management. Overview of P4P indicators of the Pwani Pilot, Tanzania % of facilities included in the HMIS monthly reports exported through District Health Information System (DHIS) to RHMT in timely manner (by 14th of the following month) % of facilities receiving a copy of a Quarterly District Health Profile report, based on DHIS Overall performance along Examples of moving from meaning units to themes in content analysis of health service users’ FGDs
N/A