How to create more supportive supervision for primary healthcare: lessons from Ngamiland district of Botswana: co-operative inquiry group

Background Supportive supervision is a way to foster performance, productivity, motivation, and retention of health workforce. Nevertheless there is a dearth of evidence of the impact and acceptability of supportive supervision in low- and middle-income countries. This article describes a participatory process of transforming the supervisory practice of district health managers to create a supportive environment for primary healthcare workers. Objective The objective of the study was to explore how district health managers can change their practice to create a more supportive environment for primary healthcare providers. Design A facilitated co-operative inquiry group (CIG) was formed with Ngamiland health district managers. CIG belongs to the participatory action research paradigm and is characterised by a cyclic process of observation, reflection, planning, and action. The CIG went through three cycles between March 2013 and March 2014. Results Twelve district health managers participated in the inquiry group. The major insights and learning that emerged from the inquiry process included inadequate supervisory practice, perceptions of healthcare workers’ experiences, change in the managers’ supervision paradigm, recognition of the supervisors’ inadequate supervisory skills, and barriers to supportive supervision. Finally, the group developed a 10-point consensus on what they had learnt regarding supportive supervision. Conclusion Ngamiland health district managers have come to appreciate the value of supportive supervision and changed their management style to be more supportive of their subordinates. They also developed a consensus on supportive supervision that could be adapted for use nationally. Supportive supervision should be prioritised at all levels of the health system, and it should be adequately resourced.


Introduction
Primary healthcare makes an important contribution to achieving equitable access to comprehensive and costeffective health services (1). Effective primary healthcare services in many low-and-middle-income countries, however, have been hampered by severe shortages and inequitable distribution of the health workforce (2Á5). Migration of health workers has exacerbated the shortage and inequity in many of these countries (3,4,6).
Multiple innovative strategies have been tried to increase the number of healthcare workers and to distribute and retain them, especially in rural and remote areas (7,8). These innovations comprised educational, fiscal, and regulatory approaches as well as diverse incentives and have had varying degrees of success (5, 7Á10). Based on existing evidence of effective retention strategies, Buykx et al. (11) developed a framework to guide retention interventions. While the emphasis was on 'bundling' multiple strategies for efficacy, the framework underscores the centrality of effective management, collegial support, and supervision.
Botswana has a documented shortage of healthcare workers which is worse in primary care (2). The scarcity of primary healthcare workers has been attributed to inequitable distribution favouring towns and hospital care, an increased need for health workers because of the HIV pandemic, and high attrition (2,12). Poor working environments and living conditions as well as inadequate human resource management practices, including poor support and supervision, are the main contributors to poor health worker motivation and high attrition (12). The supervision was perceived to be either authoritarian or inadequate and incompetent (12,13).
Supportive supervision has been shown to improve health workers' satisfaction, motivation, performance, retention, and may also enhance competence and patient outcomes (14Á16). However, implementation of supportive supervision can be hindered by the entrenched authoritarian supervisory paradigm (14). A shift to supportive supervision demands a significant paradigm shift, which supervisors may not always find socially or even culturally acceptable (14,17). In addition, there is a dearth of successful supportive supervisory practices in low-and middle-income countries to use as benchmarks (15). The sociocultural barriers to supportive supervision may be accentuated in Botswana where many of the healthcare workers have migrated from other African countries (2). This article reports on the findings of participatory action research whose aim was to change the management and supervisory practice of the district health management team (DHMT) and cluster heads in order to discover how to create a more supportive environment for primary healthcare providers. Because of its focus on enhancing practice and expanding knowledge and understanding of the process of change, the co-operative inquiry group was deemed the most appropriate method to address the sociocultural barriers to supportive supervision (18).

Study design
The co-operative inquiry group (CIG) is a form of participatory action research that involves a cyclic process of planning, action, observation, and reflection ( Fig. 1) (18Á21). The CIG method is a way of 'working with professionals who would like to reflect on and change their practice' and of 'learning and reflection . . . [on] pertinent everyday practice related problems and attempting to resolve them as part of the research process' (18,21).

The setting
The Ngamiland health district (population: 74,792), which has 29 primary healthcare facilities (clinics and health posts), is in the extreme north-western part of Botswana and is both rural and remote. The district has many areas with very difficult terrain which are inaccessible, especially during the rainy season. The headquarters of the district is in Maun, where the DHMT and cluster heads are based.
The Ngamiland DHMT was headed by a public health specialist, who was responsible for the overall leadership of the district health services. The head of preventive services, a senior nursing officer, had oversight of the primary healthcare and public health services. The head of curative services, a doctor, was mainly responsible for hospital-based clinical care. The head of corporate services, an administrator, was in charge of all the support services. The one principal nursing officer was mainly accountable for hospital-based nursing services while the subordinate principal nursing officer was the direct supervisor of the cluster heads (Fig. 2). The primary care clinics and health posts were functionally organised into seven clusters, each feeding into a larger 'mother' clinic. Each cluster was managed by a cluster head who was a nursing officer based at the larger clinic in the cluster. These were the direct supervisors of the healthcare workers in the clinics within their clusters.
At the time of this research, there were no guidelines for supervision of primary healthcare workers. However, all government employees, including primary healthcare workers, are mandated to sign an annual performance contract with their supervisors. The supervisors are therefore expected to meet at least quarterly with their subordinates to review progress and address challenges, but these meetings rarely took place (12). The interaction between the supervisors and their subordinates was largely limited to the submission of mandatory statistical reports, passing on directives from the National Ministry of Health  and crisis management when problems arose (12,13). Both the DHMT and cluster heads had scheduled regular meetings: monthly for the DHMT and weekly for the cluster heads.
The CIG process Members of the DHMT and cluster heads were invited to a meeting in Maun where the findings of prior research to assess the deficit in human resources for primary health in Botswana were presented. One of the important findings of the research was that there was inadequate support and supervision of primary healthcare workers, which had led to poor motivation, high attrition, and poor patient care (12). They were invited to participate in a co-operative inquiry that would address the question 'How can managers change their practice to create a more supportive environment for primary care providers?' Although they were initially sceptical about the validity of the CIG as a research method, all the managers consented to commit a year to participating in the study. Each group member was given a journal to record their actions and reflections. The group was trained to reflect using the adapted reflectivity 5-step method (21) ( Table 1), as well as to give and receive feedback (22).
The CIG meetings were conducted in quiet venues where refreshments were served, but there were no financial incentives for participation. The discussions were conducted in English, the official language of organisational transaction. The CIG went through three cycles between March 2013 and March 2014, each lasting 2Á6 months ( Table 2). ON facilitated three while RM facilitated two of the CIG meetings. At the meetings, each group member presented their personal actions, reflections, and new knowledge by responding to these questions: What did I do? What happened? What was different from what I expected? What did I not do? What did I do instead? What have I learnt from this? What is the issue now? What specific action should I take? (22). Other members of the group then gave feedback to the presenter. After individual presentations, the group reflected on overall experiences and what they had learnt as a group. They then planned individual and collaborative actions for the subsequent cycle (18,22,23). The facilitator reminded the group of the research question at each meeting to ensure alignment of actions and reflections with the purpose of the inquiry.
At the meeting on 11 September 2013, in addition to the usual feedback, the results of a survey on the perceived  current and desired organisational culture in Ngamiland health district were presented to the group. The results of the survey have been fully described elsewhere (13). The cultural values survey gave insight into how the district health service was experienced by the primary healthcare workers and what kind of organisational and leadership transformation would create a more enabling and supportive environment.

Construction of new knowledge
The individual contributions by group members and the group discussions were audio recorded using digital voice recorders and then transcribed verbatim. After each meeting, ON summarised the transcripts and circulated, via email, the synopsis of the individual feedback and reflections, group's learning, and further plans for action during the subsequent cycle. Key insights and learning from the group's reflections during the CIG process are presented in the results section with quotations extracted from the transcripts to support and illustrate the learning in the words of the CIG members (18,22).
The final meeting of the CIG group in March 2014 constructed a consensus of the group's learning using the nominal group technique to identify and prioritise the key issues (24,25). Each member of the CIG was given an A4 sheet of paper and asked to quietly write their responses to the question: How can managers change their practice to create a more supportive environment for primary care providers? After about 15 min, the group came together and each member in turn gave one answer to the question until all the answers were recorded on a flip chart. The group then clarified the items and those that overlapped or were duplicates were combined. Each member was then given a voting paper and asked to write down five items that were the most important for them and score them from five (most important) to one (least important). The scores were then summated and the items ranked from what the CIG considered most to least important in terms of their learning.

Ethical consideration
Approval to conduct the study was given by the University of Botswana Institutional Review Board, the Ngamiland

Results
All the 12 (100%) district health services managers, which included five DHMT members and seven cluster heads, participated in the CIG. Two of the members were male doctors, one was a female administrator, and nine were female nurses. Two mangers left the district during the 12 months of the inquiry, and one was replaced by a new member who also joined the inquiry group.
The key actions carried out by the members of the CIG included developing their supervisory visit schedules and sharing them with their subordinates, conducting supervisory visits, addressing the difficulties identified by the supervisees during the visits and optimising the use of transport. The managers also adapted a supervisory checklist to use during their visits.
Key insights and learning Important lessons and key insights emerging from the CIG are discussed below.

Current supervision practice
Managers recognised that they had not had a culture of regular supervisory visits to their supervisees in the workplace and had tried to supervise remotely from their offices in Maun. They also realised that their current supervisory style was top-down and was characterised by demanding information and giving directions:

Perceptions of healthcare workers' experiences
The managers reported that the healthcare workers appreciated the new supervisory visits: The new caring supervision style encouraged healthcare workers to share personal problems which were then addressed: . . . a lot of our staff members . . . were bringing sick leaves . . . we found out that they didn't have any money . . . they were living above their means. . . . didn't have money even for transport . . . they ended up going to get a fake sick leave . . . then it was agreed by the management that there should be some financial training . . . So this workshop was conducted last month for each and every person . . .. Barriers to supportive supervision and efforts to overcome them Hindrances to supportive supervision included inadequate supervisory skills, lack of transport, and lack of support from the National Ministry of Health. A number of managers felt inadequately prepared to deal with difficult and uncooperative subordinates: . . . I have realized that even in one mother clinic the nurses there they are having problems with communication . . The other big challenge was unscheduled meetings, workshops, and visits by the National Ministry of Health, which significantly interfered with the managers' ability to The managers realised though that the shortage of transport was exacerbated by a failure to utilise the existing vehicles and drivers optimally to benefit all the clinics: . . . When the one clinic did not have a vehicle, drivers from clinics with vehicles just passed it without going to check if there was anything they needed transported to Maun . . .. (Cluster head 7) [03/07 /13] The managers also recognised that there were opportunities to piggyback their supervisory visits onto other activities in order to ameliorate the transport challenge: . . . I realized that the anti-Retroviral [programme] has regular transport . . . there is also space for me to go for supervisory visits . . .. (Cluster head 3) [11/09/13] The managers also started collaborating amongst themselves to solve complicated transport challenges:

Appreciation of professional growth afforded by the CIG process
The managers appreciated the new skills of reflection, giving and receiving feedback and listening actively, which Table 3. Consensus on how to strengthen supportive supervisory practice of district health managers

Priority level
How can managers change their practice to create a more supportive environment for primary care providers they viewed as important to their personal development as leaders and which were different from the usual theoretical workshops. They also discovered their ability to influence some of the challenges that plagued them: . Consensus on the findings of the CIG Members of the inquiry group collectively contributed 20 items they considered key to supportive supervision in terms of their learning ( Table 3). Availability of resources and training of managers were considered the most important enablers of supportive supervision.

Discussion
This study has demonstrated that developing supportive supervision for primary care providers in Ngamiland required significant leadership and contextual transformation. The managerial leadership of the district gained insight into the need for personal development and to build more effective communication and relationships between multiple role players. In addition, it became clear that major contextual changes were required to provide an environment within which supportive supervision could flourish. The most essential changes within the context included the need for transportation and a culture of visiting the healthcare workers, even in remote rural settings. District health managers recognised the need to change their supervisory paradigm from authoritarian to supportive supervision (14). The participatory process enabled them, with assistance from colleagues, to challenge their assumptions and change their practice (20,21). The change in paradigm was a necessary step for implementing supportive supervision, which requires motivation, investment, and shift in behaviour (14). Moreover, dictatorial supervision demotivates healthcare workers, while supportive supervision has been shown, although inconsistently, to improve health workforce performance, productivity, quality of care, motivation, and retention (16, 26Á28).
Although the managers in Ngamiland health district were inspired to be supportive supervisors, motivation can only be sustained if identified obstacles are addressed and supervision is prioritised. The Ministry of Health has an important role to develop strategies that support supervisors and improve their performance (29,30). Despite formidable challenges, the managers made significant inroads in solving some of them and were optimistic that with adequate support, some degree of autonomy, and sufficient resources they could resolve many more.
The ideas expressed in the CIG's consensus of learning correlate very closely with the organisational values that the primary healthcare workers in Ngamiland and Mahalapye health districts would like to experience in the district health services. These sought after values were transparency, professional growth, staff recognition, shared decision making, accountability, productivity, leadership development, and teamwork (13). Similar organisational values were also called for by primary healthcare workers in Cape Town, suggesting that these challenges may be widely prevalent in district health services in the region (31).
Locally developed evidence on how to create more supportive supervision is more likely to be seen as relevant, applicable, and feasible to implement by policymakers in Botswana and the region (14,15). It is notable, however, that the consensus contains many of the tenets of supportive supervision that are recognised internationally, such as a shift from checklists to proactive planning, collaborative problem solving, promoting of high standards, teamwork, and two-way communication (14,30,31). In addition, the learning of the CIG recognised that supportive supervision should be prioritised at all levels of the healthcare system, and that it requires adequate resources and skilled supervisors, factors which are known to be integral to successful supportive supervision (14,30). Despite scheduling challenges, flexibility, excellent rapport, and empathetic communication facilitated value-based reflectivity with busy district managers in a resourcelimited setting. The CIG approach therefore offers a potential vehicle for the transformative process necessary to make the well-known tenets of supportive supervision a reality in district health services.
The CIG process remained well aligned with its purpose, and the members gradually developed a sense of ownership of the inquiry process. The range of possible actions by the CIG members was significantly limited by the context, particularly the lack of transport. Reflectivity developed well during the inquiry as demonstrated by the richness of the final consensus. Group meetings were difficult to organise due to the busy schedules of the management team, and more regular meetings might have enabled the inquiry to progress faster and to go deeper. The consensus on supportive supervision was developed in one of Botswana's 28 health districts which may limit its applicability nationally. This, however, is unlikely as previous studies by the team have shown that the primary health workforce in Botswana have almost identical experiences of the district health services (12,13). The consensus is also congruent with what is known to be important for supportive supervision and may thus be transferrable to wider settings (14,29,30,32).
There is a risk of desirability and confirmation biases, especially because of the hierarchical composition of the group, and some participants may have deferred to the learning of more senior members. However, the democratic nature of the CIG with emphasis on individual action, reflection, and learning significantly reduces these sources of bias (18).
This study was part of a quasi-experimental research with pre-and post-motivation and organisational culture surveys, and the results of the surveys are still to be published. However, further research is still required on the healthcare workers' perceptions of the changes and how these can be sustained.

Conclusion
Through participatory action research, district health managers in the Ngamiland health district have come to appreciate the value of supportive supervision and have changed their management style. The managers also became aware of the experiences of the primary healthcare workers in their district and the challenges to implementing supportive supervision. Furthermore, the managers developed a consensus on what is required to develop supportive supervision that can potentially be integrated into the management of district health services across the country and may be applicable in other similar settings. Supportive supervision should be prioritised at all levels of the health system and be adequately resourced and accompanied by training and mentoring of the managers.