Abstract
Introduction In Benin, as in many African countries, there is a gradual increase in physicians practising at the primary care level. A literature review showed that these primary care physicians (PCPs) have great potential for improving the quality of care. However, several conditions are necessary for this potential to be unlocked and for the PCPs to contribute effectively to strengthen their local health system. This study aims to understand the factors that underpin the performance of PCPs in Benin.
Methods We conducted a multiple case study in which we analysed eight contrasting PCP practices spread across five health districts in Benin. The study used a combination of quantitative and qualitative methods. We collected quantitative data through direct observations of the services provided by the PCPs. A descriptive statistical analysis of these data helped to assign performance scores to each PCP’s practices (for the technical quality of services and communication with patients). For the qualitative data collection, we conducted 40 in-depth interviews and 16 focus group discussions with PCPs, other health workers, health authorities, local authorities and other community members. This qualitative data was subjected to thematic content analysis to identify the factors that might explain the observed performances. We then constructed a matrix presenting all the cases studied and the potential performance factors. Finally, we carried out a cross-case analysis to identify the most critical factors supporting the performance of PCPs in Benin.
Results The PCPs’ performance scores ranged from 14.7 to 19.3 (out of 20) for communication and from 68% to 88% (out of 100%) for technical quality of care. Out of 14 potential performance factors retained after the first phase of the qualitative analysis, cross-analysis revealed nine factors that appeared to be essential in supporting the performance of PCPs in Benin in both the public and private sectors. These were (i) the values supporting the PCPs’ practices, (ii) the PCPs’ preparation to practice at first-line, (iii) the support provided to the PCPs from the hierarchy, peers or professional associations, (iv) the leadership mandate given to the PCPs and the degree of autonomy allocated to exert it, (v) the modalities of the financing of the PCPs’ practices, (vi) the accountability mechanisms in place to support the PCPs’ practices, (vii) the PCPs’ relationship with the rest of the primary care team and their leadership style, (viii) the PCPs’ collaboration with community leaders and other public officials, and (ix) the context in which the PCP’s practices are embedded.
Conclusion This study has enabled the identification of a series of factors that could act as levers for improving the practice of PCPs in Benin.
Introduction
Like the rest of the international community, African governments have committed to ensuring healthy lives and promoting well-being for all their citizens (1). This goal cannot be achieved without well-trained, motivated, available, and properly distributed human resources for health (2). While the shortage of health workers is the most visible and debated issue when it comes to human resources for health in Africa (2), the sub-optimal utilization and performance of the existing workforce is an important, yet often unnoticed, problem. Notably, recent years have seen a growing presence of physicians at African primary care facilities (3,4), whereas the dominating policy choice (still) favors task-shifting and provision of primary care services by nurse-practitioners (5,6). This situation results from a combination of physician training efforts by African universities, governments’ limited capacity to recruit and position them at higher healthcare levels, limited access to specialized training, and expansion of the private sector (7). This growing presence of primary care physicians (PCPs) in Africa is mainly observed in the private sector (and in urban areas) as a spontaneous, opportunistic, and poorly planned movement. However, there are a few cases where the PCPs are positioned as part of the academic development of family medicine or as part of a governmental choice to improve health systems’ outcomes (4). For example, the Beninese government recently launched a program to post physicians in some of the largest primary care facilities that were previously staffed by non-physicians. The aim of this program is, among others, to improve quality of care and financial viability of these facilities. However, little is known on the ability of these physicians to contribute to these goals.
Several scholars have indicated that PCPs have a good potential for improving the quality of primary care in sub-Saharan Africa and the well-being of the population (4,8–10). In many settings, PCPs improved the technical quality of care (11,12), the access to care for complex conditions (11,13,14), and the clinical processes through training and supervision of non-physicians (13). They were also able to innovate during the COVID-19 pandemic and support the country responses (8,10). However, there is also evidence that PCPs in African countries sometimes perform poorly, especially when it comes to key features of good primary care, such as patient-centeredness, (financial) accessibility (15,16), technical quality of care (16–18), continuity of care (9,19), comprehensiveness, coordination, and community engagement (13,17). Moreover, the bulk of African PCPs work in urban areas and in the private sector, which threatens the equitable access to the services they provide (4,8). These performance issues are alarming as they may lead to inefficiencies, given that PCPs remain a relatively rare and expensive source of care.
Fortunately, the current evidence indicates that, under certain conditions, the potential of PCPs can be unleashed to improve primary care in African countries. Indeed, the PCPs’ practices are diverse across various settings, and the literature report positive outcomes for the practice models that are relatively better organized and guided by a set of clear principles and governance arrangements. This is for instance the case for family physicians who receive a 2 to 4 years post-graduate training in family medicine, or for the “Médecins Généralistes Communautaires” who are community-based physicians serving in rural areas of some francophone Africa countries (4) or even physicians without a specific postgraduate training but who have received some degree of governance guidance (20). Therefore, understanding which factors influence the PCPs performance in a given context could help to better guide and improve their practices. This is especially important in countries, like Benin and many other African countries, where primary care delivery largely relies on non-physicians and where the PCPs still remain relatively scarce.
A review of the PCPs’ practices in sub-Saharan Africa, taking a health system lens (4), helped identify potential factors that could influence PCPs’ practices positively or negatively. The professional identity of the PCPs includes the values underlying their practice, their motivation to work as PCP, and their educational and professional trajectory, which is likely to influence the way they perform in relation to primary care. Governance factors such as the existence of policy guidance, clear role definitions, resource allocation modalities, incentives and other accountability mechanisms may also influence PCPs’ performance. The way PCPs’ activities are organized and their relationships with other primary care team members, local health authorities, community members and other key stakeholders are additional factors that can influence their performance. The work environment (tools, equipment, infrastructure, etc.) and the overall context (rural, urban, poverty, overall health systems performance, etc.) can also influence PCPs’ performance.
The literature examining PCPs’ performance outside Africa also suggest similar factors, but it usually regroups these factors into physicians’ personal and professional characteristics on the one hand, and organizational and systemic factors on the other (21). The physicians’ characteristics most often pertain to their training and professional experience, and to a lesser extent to their age and gender. The organizational and systemic factors include the organization of the physicians’ practices (role and task distribution with other staff for instance), their workload, the availability of adequate infrastructure and equipment, and the features of the local health system in which the PCPs’ practices are embedded. The organizational and systemic factors are reported to play a greater part in physicians’ performance than their personal characteristics (21). The specific factors (or combination of factors) that influence PCPs’ practices in sub-Saharan African countries are barely known. Most of the publications on PCPs’ practices in Sub-Saharan Africa refer to the Republic of South Africa and a limited number of other anglophone countries such as Nigeria, Kenya, or Ethiopia (4).
Our study aims to better understand the factors that support the PCPs’ performance in Benin (West Africa). We define performance as the PCPs’ ability (individually and together with the rest of the primary care team) to provide care that is in line with the key features of good primary care and to contribute to the good functioning of the local health system in which they practice. The research questions are: (1) What is the performance of selected PCPs’ practices in Benin?; (2) Which factors are perceived by a range of stakeholders to influence performance?; (3) What are the factors highlighted by the cross case-analysis most likely to contribute to PCPs offer of quality primary care in Benin?
Methods
Study design
This is a multiple-case study, using a combination of quantitative and qualitative methods. For this multiple-cases study, we defined the case as the performance of a given PCP (see the objectives at the end of the introduction section).
Selection of the study sites
To address the research questions, we needed to contrast various performance levels (good, average and poor) and observe multiple potential influencing factors. Therefore, we purposively selected 8 cases. For each case, a study site was selected using following the steps described below.
Step 1: Defining sets of variables to use as criteria to appraise PCPs’ performance
Because our objective was to include a range of study sites with different performance levels, we needed criteria to help us predict the performance of the PCPs in our database. As a result of previous studies (8,28), we had access to a database of 150 PCPs. A factor analysis from this database helped us to identify three criteria. Each criterion comprises a set of variables (see supplementary file 1). The first criterion is related to the practice of quality improvement and public health activities. The second criterion is related to the availability of essential health services (i.e. those defined in the Benin national guidelines as high-impact interventions) at affordable cost. The third criterion related to the self-rated satisfaction of the PCP. We wish to emphasize that we did not aim to construct a scientifically validated tool to assess PCPs’ performance but rather to find a pragmatic approach to establish a level of contrast between individual PCPs deemed sufficient for our specific study purpose.
Step 2: Predicting the PCP performance based on each criterion
For each PCP, we calculated a score for each criterion, integrating the values of the variables composing this criterion. A score was considered low when below the mean score for a given criterion minus the standard deviation (score < μ-σ). A medium score was a score situated between the mean minus the standard deviation and the mean plus the standard deviation (μ-σ < score < μ+ σ). A high score was higher than the mean plus standard deviation (score > μ+ σ).
Step 3: Purposive selection of the PCPs to be included
We purposively selected 8 PCPs aiming for a large mix: 2 PCPs with high scores for at least two criteria, 3 PCPs with medium scores across the criteria, and 3 PCPs with low scores for at least 2 criteria. In order to have a sufficiently large variation of cases, we also considered the gender and postgraduate training, if any, of PCPs, the institutional ownership of the facility where the PCPs practice, and the geographic position and the urbanization level of the facility’s location.
Table 1 presents the characteristics of each of the 8 study sites, representing the 8 cases studied.
Study settings
The study sites are spread across five health districts in Benin (figure 2). Three districts are urban: Cotonou 2-3 (Cot 2-3), Cotonou 5 (Cot 5) and Parakou-N’Dali (PN); and the two others are predominantly rural: Ouidah-Kpomassè-Tori (OKT), and Nikki-Kalalé-Pèrèrè (NKP).
Conceptual framework
The conceptual framework (figure 1) used for this study was iteratively constructed based on the Donabedian model (23), a scoping review previously conducted by the research team (4) and relevant literature on primary health care and health systems (24,25). We also integrated in the framework the factors described in the literature (see above) as influencing PCPs’ practices.
Benin’s health districts where the study was conducted
In this framework, the outcomes box regroups the range of performance dimensions of PCPs and their teams that we assessed in this study: technical effectiveness, patient-centeredness, accessibility of the PCP’s services, continuity of care, community-engagement, the PCP’s capacity to strengthen the primary care team and the facility as a whole, and the contribution of the PCP’s practices to strengthen the local health system. The first five dimensions are widely used to assess primary care performance (26,27) and we have chosen them because the ultimate goal of this work is to improve the contribution of general practitioners to quality primary care in Benin and Africa. These dimensions were also perceived by Beninese stakeholders as key to judge the PCPs’ performance. The PCPs’ capacity to strengthen the primary care team and the facility and their contribution to strengthen the local health system are less commonly used to assess the performance of the primary physicians or primary care workforce in general. However, they were deemed important by the stakeholders in Benin because the PCPs are usually seen as the leader of their teams and because their work is meant to support the performance of the local health system, which role is to improve the health well-being of the population withing their boundaries.
conceptual framework
It is important to note that all these performance dimensions, apart from the PCPs’ capacity to strengthen primary care teams and facilities, include not only the individual performance of the PCPs, but also that of the primary care team. Indeed, as PCPs rarely work alone in Benin, it would not be appropriate to assess their individual performance in isolation.
From the literature (4,21,24,25), we identified a number of factors that can potentially influence PCPs’ performance. We regrouped them into three categories: first, elements pertaining to the professional identity of the PCPs; second, governance features of their practice; and third, the nature of activities conducted and roles played in the local health system (see figure 1).
Data collection
Quantitative data collection
The quantitative data served to respond to research question 1. We collected this data through direct observations of the curative out-patient consultations conducted by each of the 8 PCPs. The consultations were observed by a physician present in the consultation room, using a structured observation grid. This grid was developed based on two tools: the Global Consultation Rating Scale, which is a validated scale for assessing physician-patient communication and patient-centeredness (29), and another tool for assessing the quality of the clinical examination based on care protocols used in Benin (30). The observation grid helped us to quantitatively assess two essential primary care dimensions: patient-centeredness of care and its technical effectiveness. We pre-tested the grid with five PCPs working outside of our study area to assess the relevance of these items. Feedback from these physicians was used to further improve the grids. Feedback was also given by a biostatistician. The physician who observed the consultations was trained for that purpose and conducted 10 observations together with the principal investigator (KB) prior to the actual data collection. They compared their scores for each of the 10 consultations. Discrepancies were discussed to ensure common understanding of both tools. Each PCP was observed for three consecutive days, selected as much as possible in the busiest period. Our aim was to observe at least 20 consultations. We achieved this goal for all PCPs except in site 7. The number of consultations observed for each PCP is displayed in table 2.
Qualitative data collection
The qualitative data was collected through individual in-depth interviews (IDI) and focus group discussions (FGD), using individual and group interview guides. These guides included questions pertaining to all our research questions. For research question 1, we asked key informants’ opinion on the performance of the PCPs and their respective teams. For research question 2, we explored the respondents’ perceptions on the factors that influenced (either positively or negatively) the performance of the PCPs studied. For research question 3, the interviews provided information on the factors that could possibly influence the PCPs’ performance: their professional trajectory; their relationships with other team members, local health authorities, and community members; their activities; the organization of the primary care team; the type of support PCPs received; the accountability mechanisms, etc. For research question 3, we complemented the interviews with field observations. The latter provided information related to the PCPs’ work environment, the equipment and infrastructure available, and the work organization of the PCPs and their teams.
For each PCP, key informants were selected from the following groups:
- All 8 PCPs (IDI)
- Other primary care workers operating in the health facility where the PCPs are based, including other physicians, nurses, midwives, nurse aids, and where available, non-clinical staff (IDI and FGD)
- Local district health authorities (IDI)
- Patients and other community members, including local and religious leaders, community health workers, and representatives of community-based organizations (IDI and FGD).
We respected the principle of maximum variation so as to have a diverse sample. The key informants shared information related to each of the three research questions. In total, we conducted 40 IDIs and 16 FGDs. The latter included 8 FGDs with 54 primary care workers and 8 FGDs with 63 community members (table 2).
To ensure a deep understanding of the PCPs’ practices and performance, the principal investigator (KB) and the research assistants spent two months to study the PCPs’ practices in their specific context. Moreover, KB kept in touch with the PCPs several months after the data collection for follow-up questions, when relevant.
Data analysis
Research question 1
We determined the performance of each PCP, including the individual performance of the PCPs as well as the performance of the primary care team as a whole.
For the individual performance of the PCPs, we analyzed the quantitative data using descriptive statistics. For each PCP, we then determined the mean communication score following the instructions of the Global Consultation rating Scale (29), and we determined the percentage of recommended clinical actions effectively performed in line with the prevailing care protocols in the country.
For the performance of the primary care team as a whole, we analyzed the qualitative data. Most IDIs and FGD were recorded and transcribed verbatim. Summaries were immediately produced after the interviews that were not recorded because the participant did not feel comfortable with the recording. The research team kept a daily log, using a tool developed in Microsoft Forms, reporting all research activities performed, and storing observation notes. We uploaded the transcripts in the Dedoose software (32) and performed a thematic content analysis (31) on the qualitative data. For research question 1, we coded the respondents’ opinions on the performance of the PCPs and their teams into good, average or poor performance regarding the following dimensions: technical effectiveness, patient-centeredness, accessibility, continuity of care, community-engagement, PCP’s capacity to strengthen the primary care team and facility, and the case’s contribution to strengthen the local health system. We then summarized the performance of each PCPs along each of these dimensions into good, average or poor performance, based on the majority of opinions. Finally, we integrated quantitative and qualitative results in a table presenting an overall view of the cases’ performance.
Research questions 2 and 3
We developed a coding tree, based on the factors indicated in our conceptual framework and on the basis of our field notes. This coding tree was iteratively refined, and new codes were progressively added during the data coding process. The codes were then regrouped into themes, based on semantic similarities and co-occurrence.
More specifically for research question 2, we identified the factors that were most cited across the 8 cases and across stakeholder’ groups as important in shaping PCPs’ performance. We then summarized these factors in a table and provided examples of positive and negative influencing factors, as indicated by our respondents.
For research question 3, we performed a cross-case analysis. This method facilitates comparison of commonalities and differences in events, activities, and processes within the various case studies (33). For the cross-case analysis in our study, we displayed in a matrix all the cases and the main themes representing potential performance factors. We then described in more depth the key aspects of each theme and discussed how this relates to the case’s performance. Finally, we summarized the similarities and the common elements throughout the cases in relation to each theme and concluded on whether or not a particular theme would be a factor influencing PCP’s performance.
In this paper, we present a summary of the most important potential performance factors. The matrix used for the cross-case analysis and the data within it, are presented in supplementary file 2.
Ethics
This study received the ethical approvals of the Institutional Review Board of the Institute of Tropical Medicine of Antwerp in Belgium (N° 1545/21) and the ethics committee for biomedical research of the University of Parakou in Benin (N°0513/CLREB-UP/P/SP/R/SA). We obtained informed consent from all participants, and the data were treated with strict confidentiality.
Results
Performance of the 8 PCP cases in Benin
Figure 3 displays the mean communication scores and the mean percentage of recommended clinical actions performed for each of the PCPs’ observed.
Mean quality scores of the consultations observed for each PCP case
Table 3 below summarizes the performance of the 8 PCP’s practices, based on the qualitative analysis (stakeholders’ perceptions and observation of the research team).
From the quantitative and qualitative results above, cases 1 and 2 have the weakest performance for all dimensions assessed. They had the lowest communication scores (14.69 and 17.08 respectively) and the lowest percentages of recommended clinical actions (68% and 73% respectively). Similarly, based on stakeholders’ perceptions and observations, the performance of these two cases was systematically poor or average throughout all dimensions.
The other six cases were more difficult to categorize. When we look at the individual performance for technical effectiveness and patient centered care, assessed through the consultations’ observations, case 8 had low scores. However, the qualitative data indicated that it performed well in terms of community engagement and contribution to strengthen the local health system. Cases 3 and 5 appear to be the best performing cases when we consider individual performance during observations. Indeed, even if they did not have the best score for any of the two indicators, they both had high scores for both (see figure 3). The qualitative data also confirmed this finding. However, when we integrate the other dimensions and the performance of the primary care team as a whole, the performance of cases 3 and 5 are rather average. For instance, patient centeredness of care provided in case 3 is average because of poor welcoming and communication with patients by the primary care team (apart from the PCP). Moreover, the accessibility of the PCP is limited because the PCP is regularly absent. As for case 5, its accessibility is limited because of high costs of care. There is also lack of a genuine community engagement beyond top-down designed sensitization campaigns. The contribution of this case to strengthen the local health system is difficult to appreciate because of the absence of a clearly defined population of responsibility and limited coordination with other health system actors.
Case 6 had an average performance for both indicators (17.47 for the communication scores and 81% for the percentage of recommended clinical actions). The best performing case for communication was case 7 (score of 19.28) but it was only average for the other indicator (80%). Similarly, the best performing case in terms of complying with recommended clinical actions was case 4 (88%) but it was only average for the other quantitative indicator (score of 17,63). However, the qualitative data revealed that cases 6 and 7 performed better in more dimensions than the other cases.
Factors perceived by various stakeholders as key-influencers of PCPs’ performance
The factors that were most frequently mentioned by the stakeholders were related (in descending frequency order) to: (i) equipment, infrastructure, and utilities; (ii) support of the primary care team; (iii) PCPs’ intrinsic motivation and values; (iv) PCPs’ knowledge and competencies; (v) fulfilment of the PCP’ personal needs; (vi) support from health authorities, peers, local government authorities and/or community members; (vii) workload that PCPs face; (viii) general context elements of the locality where the PCPs’ practice is situated; and finally (ix) financial resources available. Table 3 provides illustrations on how these factors influence the PCPs’ practices, as reported by the stakeholders.
Factors highlighted by the cross-case analysis as likely to contribute to PCPs providing quality primary healthcare in Benin
Based on the stakeholders’ perceptions reported above, our conceptual framework and field observations, we identified a set of factors that can potentially influence the performance of the PCPs’ practices. We have then conducted an analysis across the 8 cases to explore how each of these factors played out for each case and what could be the link between these factors and the case’s performance. We concluded that a given factor probably influences a particular dimension of PCPs’ performance when the following conditions were met:
- the factor is present for PCPs who display good performance and absent for the PCPs who display poor performance for a given dimension.
- There is a logical explanation on how the factor influences the PCPs’ performance.
Based on this analysis, we concluded that the following factors are likely to influence the performance of the PCPs’ practices included in our study: (i) the set of values underpinning the PCPs’ practices; (ii) the PCPs’ preparation to practice at the level of the first line; (iii) the support provided to the PCPs by the hierarchy, peers or professional associations; (iv) the leadership mandate given to the PCPs and the degree of autonomy allocated to exert it; (v) the modalities of financing of the PCPs’ practices; (vi) the accountability mechanisms in place to support PCPs’ practices; (vii) the extent of collaboration with community leaders and other key stakeholders; (viii) the PCPs’ relationship with the rest of the primary care team and their leadership style; and (ix) the context in which the PCPs’ practices are embedded, especially whether this practice is in urban and rural area, and the ownership of the practice by the State or not.
As illustration of this finding, the cases 1 and 2 were the least performing cases in our sample. In these cases, the PCPs had had little preparation to practice at the level of the first line, the mandate and roles assigned to them were limited to patients’ consultations, and they had a limited autonomy for influencing anything in their health facilities. They did not have any incentive or requirement to go into the community and they work in silo or in minimum collaboration rather than in team with the other primary care team members (especially the case 2). In contrary, in the cases 6 and 7, which performed better than others in several dimensions, the PCPs were very well prepared through a short post-graduate training and field coaching. As the owners of their health facilities, they had a leadership mandate and the necessary autonomy to exert this mandate. They also signed a charter with their professional association, the health authorities, and the local authorities through which they pledged to serve a well-defined population and they work together with the local stakeholders. In these cases 6 and 7, the PCPs lead the primary care teams but they exert a person-centered leadership with a good collaboration with the other primary care team members and a particular attention to the needs of these team members.
Regarding the values, we found in all cases a commitment to serve the community and a degree of professionalism. However, differences were found across the cases for other values, and this impacts their performance. For example, in the sites where altruism and commitment to help people were more developed (cases 1, 6 and 7), the financial accessibility was better. In contrary, sites where these values were lacking (case 5) or only officially stated (cases 2, 3,4, and 8), the financial accessibility was either average or bad. Similar patterns were found for the support provided to the PCPs, the modalities of the financing of the PCPs’ practices, the accountability mechanisms, and the context.
Even though equipment and infrastructure are the most often mentioned factors by all types of stakeholders, we could not find enough evidence to conclude that these influence the performance of the cases studied. Neither could we conclude that continuing education and fulfilment of the PCP’s personal needs would influence the performance of the PCP. We did not analyze the impact of the overall health regulation on the PCPs’ practices because this regulation is the same from one case to another and it was not possible to find differences among the cases.
The table below summarizes our key findings. A detailed analysis of the factors influencing the PCPs’ practices across the 8 cases is available in the supplementary file 3.
Discussion
Main results and comparison with international literature
Assessing the performance of the PCPs
In this study, we used seven dimensions to evaluate the PCPs’ performance: technical effectiveness, patient-centeredness, accessibility of the PCPs’ services, continuity of care, community-engagement, the PCP’s capacity to strengthen the primary care team and the facility, and the contribution of the PCPs’ practices to strengthen the local health system. This holistic view of the PCPs’ performance makes the evaluation complex, and simpler approaches could be advocated for routine monitoring. However, our holistic approach enabled us to get an in-depth understanding of the PCPs’ performance and to identify issues that might not be uncovered with the analysis of only one dimension. For instance, in our study, the two cases performing best for technical effectiveness performed averagely for other dimensions, which impeded the overall quality of the care received by the patients.
Measuring health workers’ performance in primary care systems should thus include various dimensions, reflecting the final outcomes expected from these systems. This multidimensional assessment is a strong recommendation of the PHC measurement framework published by WHO in 2020 (27)and other works (34,35). Unfortunately, many studies on health workers’ performance assessed one or very few dimensions, usually the technical effectiveness for specific health conditions (36–38).
Another important fact is that the PCPs’ performance can hardly be disentangled from that of their team and the health facility in general. This is particularly true in contexts like Benin and many other African countries where the PCPs usually work together or along with nurses-practitioners. The case study approach we used allowed us to analyze the PCPs’ performance within their context and understand how this performance relates to the performance of the rest of the healthcare team and to the patients’ experiences. For example, we found that some PCPs might have an excellent individual performance, but the overall performance of their practices (which includes their own performance the one of their team) was rather average. In other cases, individual average performance of the PCPs was compensated by a good performance of the primary care team, resulting in good patients’ experience and good contribution to strengthening the primary care team. Our approach to study the PCPs’ performance aligns with the recommendations in the literature, according to which one should not only focus on measuring and improving the performance of individual health workers, but also consider team and organizational factors (21,34,38).
Factors influencing the PCPs’ performance
Nine factors emerged from this multiple-cases study as likely to influence the PCPs’ performance in Benin:
- One factor is related to the professional identity of the PCPs’: the values supporting the PCPs’ practices.
- Five are related to the governance of the PCPs’ practices: the PCPs’ preparation to practice at the first-line, the support provided to the PCPs from the hierarchy, peers or professional associations, the leadership mandate given to the PCPs and the degree of autonomy allocated to exert it, the modalities of the financing of the PCPs’ practices, and the accountability mechanisms in place to support the PCPs’ practices.
- Two are related to the roles and activities of the PCPs: the PCPs’ relationship with the rest of the primary care team and their leadership style and the PCPs’ collaboration with community leaders and other public officials.
- One transversal factor: the context in which the PCP’s practices are embedded.
For the values, both the analysis of the stakeholders’ perceptions and the cross-case analysis showed that they influence various performance dimensions. Professional values are indeed cornerstones of the identity of any profession (39), and they are reported to be important for health workers’ performance (39). However, as clearly illustrated among some of the cases included in our study (table 5) and other studies (40), professional values should not be taken for granted and it is not enough to articulate them in official documents. It is important to ensure that the PCPs and other primary care workers adopt the values that would support the performance expected from them. The few existing postgraduate training programs for PCPs in sub-Saharan Africa (41,42) and some undergraduate programs (43) include socialization of the trainees for the core PHC and family medicine values, such as social justice, people-centeredness, community-orientation, etc. However, our study shows that values can also be shared through interactions with peers within professional associations or coaching from supervisors. This provides opportunities for PCPs already working and countries without a postgraduate training program yet.
As for the PCPs’ preparation, providing to the PCPs the necessary knowledge and skills to practice in primary care settings was among the top five factors cited by the stakeholders we interviewed. Scholars also pinpointed the physicians’ knowledges and skills as a key factor determining their performance (34,43). However, knowledge and skills should help the PCPs to achieve the core primary care functions (27) and be geared towards the people’s needs at primary care level (as opposed to a hospital setting) (44). Other aspects of the PCPs’ preparation include preparing them psychologically to work in resource constrained environments and improving their soft skills (leadership and communication skills for example) to embrace the various roles expected from them, to work within a team and to deal with the community leaders (5,34,45). Our study showed that the PCPs who got such preparation tend to perform better. Alike the values, this preparation can be provided through various channels including undergraduate training, postgraduate training, or coaching by professional associations or supervisors.
Another factor highlighted by the stakeholders’ interviews and the cross-case analysis is the support received by the PCPs. This support helped the PCPs to have a better understanding of the roles expected from them, to have access to some necessary resources, to get feedback on their performance and to have a psychological resilience in difficult situations. This in turn improved their performance. PCPs received support either from the hierarchy, or their peers or professional associations. Studies in both hospital and primary care settings reported that support from the health facility managers helps to implement best-practices(38), motivates the health workers for good performance (34,35), and facilitates access to necessary resources to provide care of good quality (38). Peer support is also reported in the literature as a powerful means to enhance health workers’ performance (46).
Regarding the leadership mandate, the PCPs in our sample who were given this mandate make visible efforts to improve the quality of care and the availability of resources in their health facilities. We also found that among PCPs with a leadership mandate, those with the necessary autonomy were more able to take the necessary decisions and hold the primary care team accountable for a good performance. Leadership is a key function in health system (27,47). In a primary care team this function can be fulfilled by any cadre. However, in the African context, many stakeholders see the PCP as the leader of the primary care team(5,45,48), and our study confirmed this. Consequently, giving the PCPs the official mandate and skills to effectively exert this leadership mandate appears important.
Moreover, ensuring autonomy over resource utilization and daily management at the first line facility is shown to improve performance because managers can be more responsive to local needs and react better to daily issues (34).
Concerning the financing modalities of the PCPs’ practices, the cross-case analysis showed lack of financial support is likely to lead to increased costs for the patients and poor quality. Providing financial support to the PCPs’ practices can thus contribute to improve their performance (and their financial sustainability) (49). For instance, we found in this study, as in other work (34,35), that the opportunity to receive funding (or risk to lose it) was an incentive to be more responsive to the patients. However, this financial support (and the PCPs’ payment mechanisms) must be carefully designed and geared towards the results expected from the PCPs (34,50). It should avoid the issues related to financial incentives by integrating a mix of provider payment mechanisms (50) and being combined with other types of incentives. We also found that professional financial management can help the PCPs to mitigate their costs and enhance their financial sustainability. This professional facility management (different from the leadership) is also recommended by the WHO (27) and does not need to be done by the PCPs themselves.
The accountability mechanisms in place to support the PCPs’ practices strongly appeared in the cross-case analysis as influencing the PCPs’ practices. Effective accountability mechanisms include a clear definition of the PCPs’ roles (this was for example the case for the PCPs in the public sector in our sample), a clear definition of the results expected (including the population for which they are accountable), effective ways to assess and share these results and clear consequences when these results are achieved or not. The literature largely reports the need for the PCPs and the primary care teams in general to be accountable (34,35,45). Our study also highlighted that accountability mechanisms can exist and be enforced but not support the right results. For instance, the PCPs in the public sector in our sample were held accountable for the income generated by their health facility, but this does not necessarily support the core functions of primary care. Therefore, the accountability mechanisms for PCPs in Africa should be built around the end results expected from African primary care systems (34,51).
The PCPs’ relationship with the rest of the primary care team (mentioned by the stakeholders as the support of effective and available primary care team) is another performance factor found by this study. Like in other research (27,38,50,52), we found that the PCPs are likely to perform better if they work in cohesion with other primary care professionals (especially nurse-practitioners). When such cohesion exists (rather than working in silos or within hierarchical relationships), there are better chances for the primary care team to have shared values, to share knowledge and to work together towards a shared objective (38). Moreover, in the cases where there is a teamwork, we found that the non-physicians were able to provide good quality of care to the patients, with the support of the PCP, even when the latter is not present. This teamwork requires a leadership style from the PCP that is person-centered and favors the development of shared objectives, understanding and motivating team members, and building a team spirit (35,45).
The collaboration with community leaders and other public officials appeared in the cross-case analysis and stakeholders’ interviews as a factor that supports the PCPs’ performance in terms of community orientation and contribution to a good local health system. Indeed, when such collaboration exists, the community leaders usually help the PCPs to understand the needs of the communities they serve and to build the necessary trust for the use of the health services by these communities. We also found some cases where the community leaders helped the PCPs to access financial or material resources or provided feedback on their practices, thus contributing to the overall performance of these PCPs. Engaging with local stakeholders to understand key issues, design and implement solutions for these issues has long been proven to be as an essential factor for improving performance and ensuring the success of health interventions (27,35). However, this factor is more commonly cited when referring to health for health system governance or facility management (27,35). Our study shows that engaging key stakeholders in the community is also important for the delivery of good quality primary care services and the performance of the primary care teams.
The last factor emphasized by both the cross-case analysis and the stakeholders’ interviews is the context in which the PCPs’ practices are embedded. We found that this context influences the relative importance of other factors on PCPs’ performance. For instance, a PCP with a person-centered leadership style appears to be even more important in the public sector than in the private sector. Indeed, the PCPs in the public sector don’t always have the necessary decision space to enforce orders given in a command-and-control way. Also, lab tests, branded medicines and private health insurances were more available for the cases located in urban areas. This increases the availability of services, but it also provides more incentives for over prescription. The sociocultural and economic context indeed influence the practice of any profession, and this should be considered when designing interventions to improve health workers performance (21,34,35). The level of urbanization warrants a particular attention as the share of the urban population is quickly growing worldwide. In Benin, the urban population constituted 49.5% of the country’s population in 2022 and this proportion is projected to be 65.45% in 2050. In Western Africa, 63.80% of the population is expected to live in urban areas by 2050 (53). However, some positive factors seem easier to get in rural areas (for instance a good relationship with community leaders). Therefore, planning for primary care services and PCPs’ practices should carefully consider how to not lose the advantages linked to the rural areas and avoid or work around the shortcomings inherent to urban settings.
Factors with insufficient evidence regarding a potential influence on the PCPs’ performance
Even though they were cited by the stakeholders as influencing factors, we did not find sufficient evidence at the cross-case analysis to conclude that the following factors influence the PCPs’ performance in Benin: the equipment and infrastructure available for the PCPs’ practices, the continuing education, the fulfillment of the PCPs’ personal needs and the workload of the PCPs.
The lack of equipment and infrastructure and the quasi-inexistence of a well-structured continuing education system are common situations found across Benin’s health system. In our study, we did indeed observe this lack of resources and continuing education in most of our cases, but we realized that those who performed well managed to overcome these issues. For example, in some sites PCPs managed to ensure good continuity of care despite the lack of digital patient records, which could have facilitated their work. Also, almost all the cases that had good results with patient-centered care managed to do so despite physical spaces that were not always appropriate. Finally, PCPs who had a leadership mandate, the necessary will or creativity, the necessary autonomy, or the right relationships with community leaders managed to gain access (even if temporarily) to resources for their work. Similarly, some of the cases in our sample were able to compensate for the lack of continuing education, thanks to peer support. Therefore, despite the importance of an adequate equipment (21) and continuous training (21,37), our results show that much can be done even with limited resources and other factors can help to have a good performance. Other studies have also reported successful experiences of primary care provision in resource-limited settings and even in situations of fragility (35,54). This opens great possibilities in countries similar to Benin, with urgent need to improve the quality of care for their population but where the socio-economic situation will probably not allow the PCPs to get the same resources as PCPs in better-resourced countries.
Regarding the workload, some studies found that a high number of patients negatively influences the physicians’ performance(21). In our study, it was difficult to appreciate such influence, as the average numbers of patients seen daily by the PCPs was relatively low. Also, the PCPs with a high non-clinical workload in our sample were less available for patient care. This was compensated by nurse-practitioners, and we did not find enough evidence to conclude that it influences the PCPs’ performance.
Implications for practice
Contribution to international knowledge
Most of the factors found in this study are consistent with previous knowledge. However, subtleties emerged from our study which are linked to the Benin context and to similar contexts in Africa and other regions with limited resources. In terms of strengthening knowledge and competencies, several approaches can be used: basic training and postgraduate training, of course, but also coaching and on-site support from peers and supervisors. Therefore, in countries where postgraduate training does not exist yet, actions can already be undertaken to have effective PCPs while waiting for postgraduate training programs. Similarly, we found that limited availability of key resources such as equipment, infrastructure, or continuing education can, to some degree, be compensated by a strong leadership, a good support from peer communities or health authorities to the PCPs.
The study also highlighted structural factors (preparation of the PCPs, shared values adapted to the context and to the expected results, etc.) which go beyond the individual characteristics of doctors but have great potential to influence their performance. Addressing these structural factors could help to tackle the root causes of the underperformance of the PCPs and, more broadly, primary care systems. Finally, this study provided empirical evidence that can be combined with the existing literature to build a theory on how the PCPs’ performance can be supported in Benin and refine the conceptual framework used for this study. This refined conceptual framework can be used as a basis for a policy framework or to guide future efforts to analyze and improve the PCPs’ performance in Benin and similar contexts.
Building a theory and a refined conceptual framework
A deep analysis of the factors identified in this study showed that they are interlinked and pathways to performance can be defined. These pathways include intermediary factors (highlighted by the present study and previous literature) that can play a role in how a given factor influences the PCPs’ performance (figure 4).
Pathways to PCPs’ perfromance in Benin
The following theory can summarise the pathways shown by figure 4.
“If we offer adequate undergraduate training and adequate postgraduate training and/or the right coaching, the right financing modalities, and effective accountability mechanisms, we will have PCPs with the right values and the right preparation, with the necessary support and with a leadership mandate and the autonomy to exert it. These PCPs are able to mobilize the necessary resources (education and CME) to do their work and ensure the sustainability of their practices. This in turn can allow the PCPs to perform their activities in good collaboration and teamwork with the rest of the primary care team members and in collaboration with the community leaders and other key stakeholders and with the appropriate workload. All this leads to good performance, i.e the PCPs and their team providing are that is technically effective, patient-centered, accessible, and community oriented. The PCPs would also be able to strengthen her team and the PCPs’ practices will effectively contribute to strengthen the local health system. Getting a professional facility manager (other than the PCP) can also help them in this reducing their workload, mobilizing resources and achieving the results. This pathway to performance will be influenced by the context and the right regulation can help maintaining it in a good direction”.
From this theory, we build a refined conceptual framework displayed in figure 5.
Refined conceptual framework for PCPs’ perfromance
Limitations and strengths
We used validated tools for assessing the performance of the PCPs but had only one observer to assess the consultations (instead of at least two as recommended for such assessment(29)). It was difficult to have more than one observer because of the small-sized consultation rooms and some reluctance from the PCPs and patients. This may have reduced the reliability of our findings. However, the triangulation between quantitative and qualitative data helped us to mitigate these insufficiencies. We also discussed the assessments at length between observer and principal instigator and enhance observer’s inter-consultation variation. Another limitation is the low number of cases included in the multiple case study. There is thus room to study the range of factors influencing PCPs’ performance on a larger sample and to test empirically the conceptual framework derived from these factors.
Despite these limitations, the study has several strengths. We achieved a reasonable level of variation in our sites in terms of performance, geographic localization, urbanization level, gender, and institutional ownership. This variation provided opportunities to observe similarities and differences across cases and note how a given factor would play a role (or not). This provides a certain level of confidence when concluding whether a factor was more likely than others to influence the performance of PCPs. We used an ethnographic approach with principal investigator and research assistants spending two months interviewing and observing the primary care teams and other key stakeholders in great depth. Finally, we succeeded in conducting 56 interviews allowing us to reach saturation of data.
Conclusion
By triangulating data related to the stakeholders’ perceptions, field observations of PCP practices and evidence from the literature, we can conclude that the following factors are key in shaping PCPs’ performance in Benin: (i) the values supporting the PCPs’ practices, (ii) the PCPs’ preparation to practice at the first-line, (iii) the support provided to the PCPs from the hierarchy, peers or professional associations, (iv) the leadership mandate given to the PCPs and the degree of autonomy allocated to exert it, (v) the modalities of the financing of the PCPs’ practices, (vi) the accountability mechanisms in place to support the PCPs’ practices, (vii) the PCPs’ relationship with the rest of the primary care team and their leadership style, (viii) the PCPs’ collaboration with community leaders and other public officials, and (ix) the context in which the PCP’s practices are embedded. Even if our methodology does not allow us to overgeneralize our findings, the factors highlighted in this study can further be tested in subsequent studies on PCPs’ performance in other African countries. Moreover, the study clearly shows that the factors affecting the PCPs’ and other health workers’ performance are closely linked to the performance expected of them and to the specificities of their working arrangements and conditions. This reinforces the idea of a context-adapted policy framework being important for strengthening physician performance. The empirical evidence on the factors generated by this study can be leveraged to improve the PCPs’ performance in Benin and similar contexts. It is a first step in the development of coherent policies with regard to the organization and regulation of PCPs’ practices in Benin, and beyond in other West African countries with relatively similar health systems.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
Declarations
Ethics approval and consent to participate
This study was conducted under the ethical approval N° 1545/21 issued by the Institutional Review board of the Institute of Tropical Medicine of Antwerp, and the ethical approval N° 0513 of the 1st March 2022, issued by the local ethic committee for biomedical research of the University of Parakou (Benin). Written consent was obtained from the participants, and we managed the data with strict confidentiality.
Consent for publication
Not applicable.
Availability of data and materials
The data supporting the finding are included in the tables in this published article and its supplementary information files. More information is available from the corresponding author (KB) if needed and on reasonable request.
Competing interests
The authors declare that they have no competing interests.
Funding
The study was conducted with funding from a PhD grant allocated by the DGD Belgium, through the Institute of Tropical Medicine of Antwerp. The authors also received funds from the KU Leuven for conducting a multi-stakeholders workshop for validating the finding and using the results to develop a policy framework for the practice of PCPs in Benin.
Authors’ contributions
KB co-conceptualized the study, coordinated the data collection, curated the data, performed quantitative and qualitative data analysis, and produced the first draft of the manuscript. JDL and BC co-conceptualized the study, supervised the research process, contributed to the data analysis, and reviewed the manuscript drafts. DMZ supervised the research process and reviewed the manuscript drafts. All authors have read and approved the manuscript.
Supporting information captions
Supplementary file 1: Factor loading table
Supplementary file 2: Cross case analysis simplified matrix.docx
Supplementary file 3: Details cross-case analysis final
Acknowledgements
List of abbreviations
- Cot 2-3
- Cotonou 2 and 3
- GP
- General practitioner
- IQR
- Interquartile ranges
- MGC
- Médecin généraliste communautaire
- NKP
- Nikki-Kalalé-Pèrèrè
- OKT
- Ouidah-Kpomassè-Tori
- PN
- Parakou-N’Dali
- PCP
- Primary care physician
- PHC
- Primary health care
- qual (in lowercase letters)
- Non-dominant qualitative strand in a mixed method study
- QUAN (in uppercase letters)
- Dominant quantitative strand in a mixed method study