Developing learning diaries for action research on healthcare management in Ghana,Tanzania and Uganda

Abstract

Action research (AR) can be an effective form of ‘on the job’ training. However, it is critical that AR cycles can be appropriately recorded in order to contribute to reflection and learning. One form of recording is for coresearchers to keep a diary. We found no previous literature describing the use of diaries in AR in sub-Saharan Africa. We therefore use this paper to reflect on how diaries were used by district health management teams in the PERFORM project. We share five lessons from our experience. First, it is important to foster ownership of the diary by the people who are responsible for filling it in. Second, the purpose of keeping a diary needs to be clear and shared between researchers and practitioners from the very beginning. Third, diaries should be allowed to evolve. Fourth, it is a challenge for busy practitioners to record the reflection and learning processes that they go through. Last, diaries on their own are not sufficient to capture reflection and learning. In conclusion, there is no best way for practitioners to keep a diary; rather the focus should be on ensuring that an AR recording process (whether diary or otherwise) is locally owned and complements the specific practice setting.

Keywords

Diaries action research district health management teams Ghana Tanzania Uganda

Introduction

The lack of sufficient and high performing healthcare staff is the greatest barrier to scaling up services needed to achieve health-related Millennium Development Goals in sub-Saharan Africa (SSA) (The Task Force on Health Systems Research, 2005; Travis et al., 2004). The deficit may be addressed by training more new personnel, and, improving the performance of the existing and future workforce. Recent research and development initiatives have emphasised the former with a serious neglect of the latter (Marchal, Denerville, Dedzo, De Brouwere, & Kegels, 2005). At the same time, public health systems (HSs) in SSA are decentralising, meaning that authority for human resource (HR) planning, management and evaluation is being devolved to district health management teams (DHMTs). This opens up opportunities for DHMTs to take control of such functions to more effectively meet local healthcare needs. Given the increased authority of DHMTs, top down HR initiatives of central government become less relevant while bottom up, locally engaged and owned initiatives are required that will be of immediate practical use to DHMTs.

Action research (AR) can be an effective form of ‘on the job’ training in HR planning, management and evaluation (Mshelia et al., 2013). In AR, researchers and practitioners systematically work through cycles that describe and analyse the changing HR and HS situation that they face; identify and plan strategies to improve the situation; implement changes needed; observe and record the effects of doing so; and reflect on the processes and effects of such changes (Reason & Bradbury, 2001). Diaries are one way for busy DHMTs to record their journey through these learning cycles in order to quickly identify ways to improve their performance. Other ways of recording that have been used in published AR studies are learning diaries (Reason & Bradbury, 2001) and learning histories (Roth & Kleiner, 1998).

The purpose of this paper is to reflect on the use of diaries to record action and learning throughout an AR project conducted by DHMTs in Ghana, Tanzania and Uganda. The paper focuses on a practical AR question that was encountered early in the project – what is the most effective way for DHMTs to reflect on the ‘action’ for enhanced learning? The project team needed to know who should be recording, what, how, when and for how long. Substantial guidance is given in handbooks on the use of diaries, learning histories and other forms of recording of AR cycles (e.g. Draper, 2001; Munn-Giddings & Winter, 2001; Reason & Bradbury, 2001). However, these overwhelmingly focus on high income countries (HICs) and little similar guidance was found for low and middle-income countries (LMICs) where supportive technologies are less available, workloads can be higher, and continuous professional development for the health workforce is lacking. A literature review was undertaken to seek empirical exemplars from other AR projects in LMICs. The project team was surprised to find that most documentation of AR cycles had been undertaken by academics, with the non-academic co-researchers seemingly fulfilling passive or semi active roles of data providers. Only one article fully explored the issues in dealing with diaries in which the authors noted that “very little has been written on diaries as a tool in AR” (Buchy & Ahmed, 2007). We fill this gap in the literature by describing and reflecting on the project experience of introducing and using diaries within AR by DHMTs operating within the public HSs of Ghana, Tanzania and Uganda.

This paper is structured as follows. First, background on the HR challenges in Ghana, Tanzania and Uganda is presented. Next, the PERFORM project is introduced and how the project arrived at a decision to use diaries is explained. Thereafter our methods for this paper are described before reporting on DHMT experiences of using diaries. We finish by discussing the key lessons learned in order to share these with other AR practitioners working in LMIC settings.

Background

The PERFORM project (“Supporting decentralised management to improve health workforce performance in Ghana, Tanzania and Uganda”) was a four-year EU project (2011–2015) that aimed to enhance understanding of how, and under what conditions, AR could strengthen district health management. Stronger DHMT performance was assumed to lead to improvement in the performance of frontline healthcare staff.

Ghana, Tanzania and Uganda were selected for the PERFORM project because they are three LMICs that face common problems of health workforce shortage and maldistribution; are actively trying to address these problems through their health policies and plans; and have sufficiently decentralized management structures to support and make use of the AR approach.

HR challenges in Ghana, Tanzania and Uganda

Situated in West Africa, Ghana covers approximately 238,500 km², with an estimated population of 25,905,000 and children under-five years constituting 14.5% of the total population. Totally 53.4% of the population now reside in urban areas (Central Intelligence Agency, 2015a; World Health Organization, 2015b). Ghana is officially classified as a LMIC (The World Bank, 2015b). Healthcare services are provided by the Government (approximately 65%), Christian Health Association of Ghana (CHAG), private Islamic missions, private for profit, quasi-governmental and non-government organisations. There are several languages spoken in Ghana but the official language is English.

Tanzania is situated in East Africa. The mainland covers 947,300 km² and the population is estimated to be 49,253,000 and children under-five years constituting 17.9% of the total population. Totally 30.9% of the population now reside in urban areas (Central Intelligence Agency, 2015b; World Health Organization, 2015c). Healthcare services are provided by the Government (74%), Christian Social Services Commission (CSSC), private Islamic missions, private for profit and non-governmental organizations. Kiswahili and English are the official languages in Tanzania. Kiswahili is widely spoken and has recently been adopted as the language of instruction in all schools (Quartz, 2015).

Uganda is also located in East Africa. It covers 241,038 km² and has an estimated population of 37,579,000, and children under-five years constituting 19.4% of the total population. Totally 84.2% of the population live in rural areas, largely practicing subsistence agriculture (Central Intelligence Agency, 2015c; World Health Organization, 2015a). Uganda is officially classified as a low-income country (The World Bank, 2015a). Health services are provided by government (63%) and non-state providers, including faith-based and private practitioners. English is the official language in Uganda with Kiswahili as the country’s second official language (Ka Tutandike, 2008).

The public healthcare systems in these countries are hierarchical. District services (the lowest level of the healthcare system) report to a regional department of health in Ghana and Tanzania, and the regions report to the central Ministry of Health (MOH). In Uganda, there is no regional level so the districts report directly to the MOH. Under decentralisation, some of those previously centrally or regionally held responsibilities and authorities are being devolved to DHMTs.

Table 1 provides a summary of the main health and HR statistics for each country.

Table 1.

Key statistics for each country.

Indicators	Latest year available	Ghana	Tanzania	Uganda
Total number of districts in the country	2012 (G, T), 2015 (U)	216	169	136
Infant mortality rate (per 1,000 live births)	2013 (T, U), 2014 (G)	41	38	42
Under-five mortality rate (per 1,000 live births)	2013 (T,U), 2014 (G)	60	53	60
Maternal mortality ratio (per 100,000 live births)	2013	380	410	360
Health worker density
Doctors per 1,000 population	2008 (G), 2006 (T), 2005 (U),	0.11	0.01	0.12
Nurse/Midwives per 1,000 population	2008 (G), 2006 (T), 2005 (U),	0.97	0.24	1.31

G: Ghana; T: Tanzania; U: Uganda.

Data sources: Country statistics and global health estimates by WHO and UN partners.

At the district level in each country, a DHMT, typically chaired by a District Medical Officer (DMO), is responsible for planning and management of health services. The number of healthcare professionals in such teams and their professional composition varies between the countries. In addition to the DMO (who is invariably a medical doctor, often a public health professional, and is typically based in a district hospital carrying both clinical and management duties), a DHMT may consist of heads of nursing, environmental and dental health, supply chain and data management, HRs and finance and other district functions. The terminology and location of DHMTs within local government structures varies between the countries. In this paper, we use the term DHMTs to mean a decentralised health management team responsible for planning and managing health services provision within a health district (including sub-district health facilities) as well as the implementation of national and regional plans and policies.

The DHMTs typically cope with several challenges. Finances, equipment and infrastructure are usually scarce; implementation of national policies may not have been fully delegated to regional or district levels; and teams can be overwhelmed by workload due to insufficient staff or ad-hoc duties passed down from central MOH as well as participating in different international aid projects. Healthcare providers feel discouraged, overworked and undervalued resulting in low motivation, high absenteeism and poor retention (Gormley & McCaffery, 2013).

In each country, three districts were selected to participate in the project using three broad criteria. First, owing to the collaborative nature of the project, it was important to have a motivated and reasonably staffed DHMT. A second criterion used in Uganda and Ghana was the district’s level of performance based on a national performance league table. It was broadly intended that one well and one less well performing district be included so that differences in project effects in these different settings could be examined. Finally, districts with broadly differing characteristics including a mix of rural and urban were sought. The selection of districts was undertaken under the guidance and with the approval of the MOH in each of the three countries. No ‘control’ districts were selected in any of the countries, because the project did not seek to attribute potential changes in management processes exclusively to the AR approach.

Table 2 summarises key characteristics of each district in 2011 when the project began.

Table 2.

Key characteristics of the study districts.

Feature	Ghana			Tanzania			Uganda
Feature	Akwapim North	Upper Manya Krobo	Kwahu West	Iringa Urban	Kilolo	Mufindi	Jinja	Luwero	Kabarole
Population	134,590	75,152	199,604	172,130	233,727	317,760	501,300	472,300	415,600
Area (km²)	450	658	414	162	7,881	7,123	768	2,577	1,844
Setting	Mostly rural	Rural	Mostly rural	Mostly urban	Rural	Mostly rural	Mostly rural	Rural	Rural
Number of doctors/ 1,000 people	0.07	0.03	0.09	0.02	0.01	0.01	<0.05	<0.05	<0.05
Number of core DHMT members	8			8			6

PERFORM Consortium partners and their roles

The PERFORM Consortium is made up of six partner institutions – three from Africa and three from Europe. Each partner from an African country was paired with a partner from Europe. Table 3 shows the paired partners.

Table 3.

Paired partners.

Country	African research partner	European paired partner
Ghana	School of Public Health, University of Ghana	Swiss Tropical and Public Health Institute, Switzerland
Tanzania	Institute of Development Studies, University of Dar Es Salaam, Tanzania	Nuffield Centre for International Health and Development, Leeds University, United Kingdom
Uganda	School of Public Health, Makerere University, Uganda	Liverpool School of Tropical Medicine, United Kingdom

Table 4.

Comparison of diary process and structure in the three countries.

Question	Ghana	Tanzania	Uganda
When was the diary introduced?	February 2013	August 2013	February 2013
What was the diary mainly used for?	Checklist for activities undertaken/ reflection to a lesser extent	Record of activities undertaken	Checklist for activities undertaken/ reflection to a lesser extent
How frequently did the CRT review the diaries (visit to the district)?	Weekly/fortnightly, depending on workload	One visit every two months	One visit every two months and ad-hoc
What format did the DHMTs keep the diary?	Paper	Computer and paper	Paper
One communal copy or individual copies or both?	Both	One communal copy	One communal copy
How did the format of the diary change over the lifetime of the project?	Individual diaries (without prompts) did not change; communal diaries (with prompts) replaced with a “documentation template” (with prompts)	Column linking activities recorded in the dairy with Comprehensive Council Health Plan (CCHP) activities was added	Format did not change; content and style of diary keeping changed

Each African partner, known as the Country Research Teams (CRTs), was located in one of the participating countries and was responsible for facilitating AR in the selected districts. Each CRT was made up of at least three researchers. In Tanzania, a focal CRT member was assigned to each district. However, in Ghana and Uganda, all members of CRT oversaw activities in each district. Each European Partner (EP) provided support and research advice to their respective CRTs. The EPs, although participated in inter-district workshops, did not engage directly with DHMTs in facilitating the implementation of AR in the respective districts. The CRTs were to be a critical friend to the DHMTs through repeated contact by scheduled visits, inter-district workshops, email, skype, telephone calls and texting.

The DHMTs were regarded as co-researchers in the project. They collected data, took part in the analysis of that data and led each phase of an AR cycle. Since the capabilities and strengths differed from one DHMT to the other, there was room to negotiate the roles and levels of participation between the CRTs and DHMTs. The DHMTs from the selected districts came together in (separate) national workshops to share learning and progress on the project activities. Apart from the workshops and meetings mentioned in the paper, no specific arrangement was made for regular inter-district collaboration as part of the project. No monetary payments were made to DHMTs for their participation in the project. However, they were reimbursed for the participation in workshops. DHMTs were assumed to benefit from participation in the project through improved capacity to systematically identify strategies to address HR and HS challenges albeit with limited resources. Payments were made to all CRTs and to EPs. We return to this small but important point on payments later in the paper.

The DHMTs identified and selected HR and HS strategies (for the purpose of PERFORM, these strategies were referred to as HR/HS bundles) which were feasible to address within project timeframes, decentralised responsibilities and DHMT budget. Together, the DHMTs and CRTs were to work through at least two AR cycles as they developed (plan), implemented (act), measured the outcomes of the implemented interventions (observe) and reflected on the outcomes (reflect) of these HR/HS bundles over a period of 12–18 months. Further detail on the PERFORM project methodology can be found elsewhere (Mshelia et al., 2013). This paper specifically focuses on recording the action, outcomes, reflection and learning during AR cycles.

Substantial discussion within team meetings and by email between EPs and CRTs preceded the decision to use diaries. Early in the project, recording during AR cycles was expected to be undertaken through the use of learning histories (Reason & Bradbury, 2001; Roth, 1996). Learning histories were identified as useful because they potentially support the learning capability of an organisation. They also promised to capture informal processes of reflection, learning and change that may be happening within the DHMTs anyway, for example, at their weekly management meetings. Learning histories would merely formalise those processes. The learning history approach, however, relies on a learning historian – the person who has primary responsibility for the development of the learning history. The partners decided that it would be more practical if the learning historian was a member (or members) of the DHMT because they would be continually present in district locations and could capture the dynamics of action, reflection and learning that was required. But no partner had experience in the use of learning histories so a (presumed) simpler choice of diaries was made by the lead EP. Diaries were still seen as a means to both record activity taking place when implementing the HS/HR bundles and be a vehicle for encouraging reflection and learning experienced by the DHMTs.

Before describing how diaries were introduced, used, and whether they met project expectations, we report on the methods that underpin this paper.

Methods

The authors of this paper (referred to as ‘we’ throughout the paper) include members of different EPs and CRTs. As mentioned earlier, the EPs provided support and advice to CRTs, and the CRTs facilitated the implementation of AR in the study districts. Although no members of DHMTs are co-authors of this paper, their views are included in the analysis and reported in the paper.

The data for this article came from a review of ongoing project documentation; a peer reviewed literature review; and semi-structured interviews carried out in the three participating countries.

The project documents reviewed included minutes from monthly project management meetings on the topic of documenting AR cycles; the AR handbook prepared by the lead partner on methodology; and the visit reports prepared by the CRTs after they visited the districts. We sought instances where EPs and CRTs discussed how recording was carried out and whether local learning had been captured.

A literature review was undertaken to identify published literature on studies in which local practitioners or the community participated in recording AR cycles. Criteria for inclusion of literature for the review were: academic articles, which gave sufficient detail on methods to indicate how active participants had been in data collection and analysis for AR taking place in LMICs; reflection on methodological issues encountered and how they were overcome was a highly desirable second criterion. Searches were undertaken in the Web of Knowledge database using various combinations of key word search terms: learning history; diary; record*; document*; recording learning; reflect*. The search was restricted to English-language articles from LMICs. The research results were cross referenced for continent and random LMIC to check that relevant articles had not been missed. The citation and abstracts were downloaded into Endnote. Abstracts were scanned for indications of recording by local participants, and where this was found, the full article was scanned. Eight articles (Adjei-Nsiah et al., 2004; Ahari, Habibzadeh, Yousefi, Amani, & Abdi, 2012; Buchy & Ahmed, 2007; Faure, Hocdé, & Chia, 2010; Freudenthal et al., 2006; Gutberlet, 2008; Marincowitz, 2003; Schoepf, 1993) were retained and reviewed.

Given the paucity of the literature base on the use of diaries and to ensure the representation of views of both CRTs and DHMTs in the analysis, the authors undertook interviews to understand how diaries were used to guide reflection and learning in the PERFORM project. Thirteen semi-structured interviews with four CRT members and nine DHMT members were undertaken between October 2013 and April 2014. All interviews were conducted in English and undertaken by researchers of one EP. The key questions were how the interviewees perceived the purpose of the diaries; whether and if so, how, diaries were used in general, for reflection and learning; and whether the diary was perceived to be making a difference to workforce management practices. The interviews were audio-recorded, transcribed, double-coded and later thematically analysed by two researchers, one of whom conducted the interviews, using NVivo (version 10).

Initial results from each of these methods were shared with all co-authors through several stages of reflection to confirm, challenge and feedback into a cycle of reflection and learning on diary use for AR initiatives in LMICs. The views presented here are those of researchers, because the DHMT members were not directly involved in the analysis of interview data or the writing of this paper.

Using diaries in AR

The EPs prepared guidelines on keeping a diary – content and format – that were emailed to each CRT in February 2013. The guidelines emphasised that the DHMTs were free to adapt the formats for their diaries but provided instructions on content (see Figure 1). All DHMTs agreed to create diaries; fill them in routinely; and to share their diaries with the CRTs, either in person during CRT visits to the district or via email. The DHMTs started filling out the diaries in March 2013 in Ghana, and August 2013 in Tanzania and Uganda. In Tanzania, the diary was kept in Kiswahili while it was filled in English in Ghana and Uganda.

Figure 1.

Instructions for keeping the diary.

The CRTs introduced the diary to DHMTs in their country. While the CRT received the same guidance on how to create and use the diary, the CRTs introduced the diary differently in each country, and the participating DHMTs took a unique approach to keeping their diary. To capture these unique differences, we next describe the evolution of diaries in each country, focusing on how the diaries were introduced, the purpose and value of diary keeping from the perspective of the DHMTs, how their format/structure and process of diary keeping changed over time, and last, how the diaries were used in reflection and learning within the participating districts.

During their visits to the districts which took place at least once every two months, the CRTs discussed the diary contents with the DHMT. The CRTs asked for clarifications and made suggestions on how to better capture the action in the diaries, as well as increase the frequency of recording. Examples of the diaries in each country can be found in the supplemental files.

Ghana

The CRT introduced the diary to the three participating DHMTs during an inter-district workshop which was held in February, 2013. All DHMTs opted for a paper-based diary and used a ruled notebook, separated into weekly segments. Each DHMT met once a week to review the recorded activities of the previous week, and plan for the week ahead. Each participating DHMT had one diary and any member of the DHMT could write in it. Initially, each member of the participating DHMT kept individual diaries as well as a common team diary. DHMT members recorded activities in their individual diaries, and then transferred the details into the communal version during DHMTs’ weekly meetings. This system meant that DHMT members often did not have time to transfer entries from their individual diaries to the communal diary. To deal with this, the CRT and DHMTs agreed, during a monthly CRT visit in April 2013, to choose a focal person amongst the DHMT members to ensure the transfer of entries between the individual and communal diaries. However, on occasions when the focal person was away from the district health directorate office, and therefore not available to make the transfers, other members of the team felt it was not their responsibility to make the transfers. This meant the communal diary was left unfilled. After this, in June 2013, all DHMT members agreed for each member to promptly transfer individual diary‘s recordings into the communal diary. The structure of the diary did not change during this time.

When introducing diaries, the Ghana CRTs emphasised to DHMT members that they would need to reflect on the management processes they went through during the implementation of project HR/HS bundles, and that such reflection was a key aspect of learning within AR. The DHMTs acknowledged the diary approach to recording management processes and activities as appropriate, since it helped them to reflect on their activities. Though reflection did take place, it was not recorded in the diary. Some DHMTs reflected as a group and some as individuals as they undertook an activity or wrote about it in the diary:

‘We meet every Monday morning as a team to consider issues which definitely might include PERFORM… so, as we sit down, if there are PERFORM objectives, there was an issue, and we look at it; in addition to other issues. And if the issues are so pertinent, that it can’t sleep over and wait till Monday, immediately we take a decision on it. And then on Monday, we look at it and see if our decision was okay; our ideas were exactly what we needed, and so forth… whether the appropriate action has been taken as well.’ (DHMT member)

DHMTs discussed their activities within the team, how they had been undertaken and whether they had achieved desired outcomes – but this was not recorded in the diary. CRTs and DHMTs recognised that the diary was rarely used as a means to record reflection.

‘As a group, (the DHMT) may have had a meeting and discussed a few things then, on reflection, taken some decisions… but then when we go into their diaries, nothing shows. We realised that the diaries were only capturing major activities and training.’ (CRT member)

The CRT decided to revise the format of the diary and changed the name to ‘documentation template’ in July 2013. These revisions were undertaken by the CRT and DHMTs together. The changes were focused on the inclusion of specific prompts which were intended to encourage reflection on causes and reasoning behind recorded activities. Each DHMT member still kept individual diaries on a daily basis. The documentation template replaced the communal diary. The DHMT saw the documentation template (with prompts) as a more convenient approach to recording and reflecting as a team. The individual diaries were useful for monitoring and daily tracking of activities. Between July 2013 and August 2014, the documentation template had been filled on average 39 times in each of the participating districts.

The project team assumed that learning meant a continuation between the information derived from experience and suggestions for change. In other words, when the project team sought indications of ‘learning’ in the AR cycles that learning was said to have taken place when the DHMT reported that they had considered alternative ways of acting that were different to what had happened before. The diaries did not capture this reflection and progression between considering options to an existing activity and reaching a decision to change that activity.

An example of this lack of alternatives was cited during the interviews when the interviewee described a request to repair a refrigerator.

‘So let’s say, we went to a facility, (and) found out that their cold chain system was down. Immediately, we have to report on that. For that one, we wouldn’t wait till the following Monday (during the DHMT meeting) to take action on that. Immediately when we come back, we have to write and send a request to the regional office for them to also forward to the headquarters for the fridge to be repaired (DHMT member)

Tanzania

The CRT in Tanzania introduced the diary to the DHMT members in participating districts during an inter-district workshop in August 2013. All DHMTs decided to keep the common diaries as a Microsoft Word document on their office computer and each appointed one or two ‘focal persons’. These focal persons were responsible for updating the diary and were the Health Secretaries in two districts and in the third district, the Council HIV and AIDS Coordinator (CHAC) and District Nursing Officer (DNO). All DHMT members met with the focal person(s) and the entry was made together with the focal person. All DHMTs used a table format, with each column having a different prompt at the top. The format did not change, although the CRT and DHMTs agreed to give an additional template to respective district health officers dealing with HIV in the Council (local government) to record any relevant information related to the PERFORM HR/HS bundles. During their monthly visits, the CRT noted challenges in the way the information was recorded. For example, there was confusion on how to complete two columns related to effect of bundles and reflection. Following discussions, both agreed that the effect column should contain information on the outcome of the activity conducted while the reflection column should contain information on circumstances that influence an activity to be successful or unsuccessful.

The diaries were kept either in the focal persons’ offices or the DMO’s office where they could be easily read by any member of the DHMT. The diaries were usually emailed to the CRT in Dar es Salaam on a monthly basis for review and comments. The CRT used these diaries as monitoring tools for the implementation of the bundles and also formed the basis for discussion with DHMTs during the next field visit to the districts.

Over time, the CRT and DHMT noticed that they were not recording all their PERFORM related activities and so it was agreed towards the end of 2013 that PERFORM would be added to the agenda of the routine weekly DHMT meetings as a means of improving recording in the diary. Each district averaged about eight entries in their diaries per month.

The diary was seen as a tool to record “and report activities which are being done” (DHMT member) as part of the PERFORM project. To the CRT, it appeared that the DHMT members did not have a habit of reflection in writing at initial stages of the project:

‘(they) are not used to thinking, to observing in a reflective way. Normally they just do….’ (CRT member)

However, some other DHMT members appreciated the importance of the diaries. During a focus group discussion between the CRT and DHMT, one DHMT member noted that using diaries:

‘…helps us in keeping records, data and other activities that we implement in the month, week, etc…. it can make us succeed in self-assessment as to how we are supposed to move ahead, what are the challenges, so the benefits are immense…’ (DHMT member)

Some DHMT members went further to propose that the tool should be used in other aspects of their work:

‘…we must use the diary in the implementation of the [Comprehensive Council Health Plan] CCHP in order to track the day to day implementation of the plan. If a person comes now wanting to know what we did yesterday, we can show that we did from here to there. Therefore as DHMT we feel that we will go on using it [the diary], and it is something that we have learnt very well…’ (DHMT member)

Regular monthly CRT visits to the districts were used to discuss and reflect on diary entries. However, this did not appear to have initiated a culture of reflective practice partly because some DHMT members stated they were too busy with many ad-hoc activities and had little time to reflect on the effects of the implementation of the bundles during their weekly management meetings.

Uganda

The CRTs introduced the diaries to DHMTs during a workshop in February 2013. All DHMTs opted for paper based diaries using a ruled notebook. Two DHMTs kept their diary in a specific (usually the DMO’s) office and then any team member could write in it as desired by visiting the designated office to make an entry. One DHMT decided that the diary should physically move between DHMT members and, also, to the sub-districts. All DHMTs selected a focal person to be responsible for the safekeeping of the diary and to coordinate entries. Keeping the diary in one place meant that DHMT members knew where the diary was at all times and knew how to gain access to it. On the other hand, those who worked at a different location (e.g. in the sub-district) could only write in it when they came to the DHMT headquarters. When the diary moved locations, it was easier for the DHMT and sub-districts to gain access. However, it also meant waiting for up to one week before the diary came to their location. When this happened, there was a time lapse, especially when the diary was in another team member’s office or in a different location, so not readily accessible:

‘(Filling in the diary) is not as immediate as when I do an activity today, I will record it today. I will wait for their diary to come from where ever it has gone and record it in there… in your head you keep remembering 'I did this activity, when the diary comes, I will record in it.’ (DHMT member)

The diary in Uganda was therefore not seen as a complete record of activity, since as well as forgetting, workload could mean the diary was not filled in.

‘Sometimes the people write the reports after the work is done, sometimes they forget to record here from their work, and anyway I ask them have you updated book, the ‘PERFORM book’ and they say ‘eh, I've forgotten’. (DHMT member)

Over time, the communal diary evolved into duplicate copies. When DHMT members went for supervisory visits to sub-district health facilities, they used the diary to document the conversations they had with heads of those facilities, and to note the actions and changes that sub-district health facilities needed to make before the next DHMT visit. However, it was not possible to leave the communal diary with any one health facility so the DHMT members decided to start filling in the diary in duplicate, using carbon paper, so that they could give the head of the health facility a copy of the entry and retain the original. The DHMTs also experimented with using a Facebook group as a diary format, where they could share their experiences of using diaries and implementing the HS/HR bundles. This was not taken up because the costs of accessing the Internet were laid on DHMT members, and many of the DHMT members had never used Facebook before and were not confident in using it.

The written diary often acted as a checklist for activity monitoring in Uganda – as a DHMT member read an entry, s/he could check off activities undertaken from the annual district plan. DHMTs perceived the value of the diaries to be in focusing on monitoring and implementation issues, particularly prioritising activities and improving accountability to pre-planned objectives:

‘…we initiated a district health disciplinary committee…to look through the issues that are affecting human resources performance issues like late coming, absenteeism, planning and discipline. So when we establish that committee we started summoning the staff, …we summon that person, who comes to the committee, we discuss and make this person commit himself or herself to improving on his area of service and it has had an effect’ (DHMT member)

While DHMTs appreciated that the written record was not complete, they felt the diary did allow a new form of sharing that promoted learning:

‘It gives a learning experience from one health sub-district to another, from a district headquarters to the lower (health system) levels. So basically, it is there to help us run our day to day activities and keep referring to what worked and see whether what worked can work again.’ (DHMT member)

Although DHMTs noted that they were unfamiliar with a ‘diary culture’:

‘Our reading and writing culture, documenting culture is poor’ (DHMT member)

The guidance and support from CRTs was seen as essential to facilitate diary use and overcome some of the constraints of time, workload and culture mentioned above.

‘(the CRTs) give us guidance throughout, even if it’s just through their responses because sometimes we get stuck how do we go about this and we then discuss with DHO (District Health Officer) and then he (the DHO), calls and sometimes they (CRT and DHO) mail each other like that… so that information and even the regular meetings with them have helped us so much’ (DHMT member)

Discussion

Table 4 provides a summary of how the structure and process for keeping the diaries varied between the three countries. From the results presented above, it is clear that diaries were actively taken up by DHMTs, which make them unusually active in recording AR cycles when compared to other AR initiatives conducted in LMICs. Based on the DHMT and CRT experience of using diaries in the PERFROM project, we see five key learning points to share with other AR initiatives when recruiting busy local healthcare personnel to use diaries and record AR cycles.

Ownership by practitioners encourages use of diaries

There is no consensus in the literature on the best method for local practitioners to record AR cycles. For AR projects in LMIC contexts, written recording is most often undertaken by academic researchers who use standard qualitative research tools (such as interviews, field notes, personal reflective diaries, workshops) that do not appear to be significantly altered by their use in an AR project. Ahari, Habibzadeh, Yousefi, Amani, and Abdi (2012) provide some examples. Learning histories have, so far, not been used in LMICs while the use of diaries has been reported once (Buchy & Ahmed, 2007). In PERFORM, the DHMTs were not directly involved in the discussions around the choice of using diaries in the project; rather the decision to use diaries was reached between the EPs and CRTs. It is therefore notable that diaries were actually used – DHMTs took ownership of a recording tool and brainstormed on how to make it work locally. The actual use is an important finding and contrasts with the results reported by Buchy and Ahmed who reported no such local ownership of diary use by community NGOs in their AR project, despite repeated encouragements from researchers including payments to the local practitioners as an incentive to diaries (Buchy & Ahmed, 2007). No payments for implementing activities were made to DHMTs during the PERFORM project.

A possible reason for the uptake and ownership of diaries may be the context that practitioners are operating in. The diaries were embedded into the routine practices of the DHMTs rather than being a separate ‘new’ mechanism. This was significant as healthcare staff operate in a text-based work culture of annual plans, health information monitoring and reporting against targets. Therefore, the diaries were similar to other recording processes familiar to the DHMTs. In Tanzania, for instance, the teams routinely read and discuss their forward plans and reports. These kind of recording and monitoring against targets are compiled typically in relation to the budget spent on each activity. Such recording is required not only by the public HS (to Departments of Health, for instance) but also by external project donors. PERFORM was one of a number of externally funded projects taking place in the participating health districts during 2011–2015. Hence, the DHMTs were familiar with recording their activity and the concept of continuous monitoring against plans. These concepts form an important part of their routine management practice. The communal diary and individual diaries were suited to the specific context and circumstance of the districts. We propose, then, that there is no single best way for busy local practitioners in various contexts to keep a diary, rather the best way of recording should be context-specific and be determined by the practitioners themselves.

While PERFORM was successful in producing diaries that were kept by DHMTs, more detailed diary entries kept more consistently may have resulted if the DHMTs had been involved at the earliest stages of choosing and developing the ideas for the diaries.

A clear and shared purpose is necessary for effective diary keeping

Although the diary was intended to support recording, reflection and learning, the DHMTs focused on recording. There are two possible explanations for this. It may be that the CRTs placed more emphasis on recording when the diaries were introduced and during the regular visits to the district thereby inadvertently neglecting reflection and learning. In addition, DHMTs were less familiar with active reflection on their daily practice and preferred to default to familiar behaviour, being recording.

Our experience on PERFORM is that it is important to obtain agreement from all the parties on the purpose and design of the AR recording tool. It is equally important to ensure consistent and balanced reiteration of the purpose of the recording tool over the course of a project to ensure that it can be met.

Allow diaries to evolve

Conspicuously, the PERFORM diaries evolved over several months of use. The format changed from being individual to becoming communal and from being kept solely by the DHMT members to copies being given to heads of health facilities. In Ghana, an additional reflection tool was seen to have added value, while the Ugandan DHMTs experimented with the use of Facebook to support the diary. The diaries became ‘live’ documents with local practitioners experimenting with various formats and ways of keeping them. Allowing the diaries to evolve was beneficial because it increased ownership of the diary by the DHMTs and encouraged integration of the document into their routine practice.

The CRTs reviewed the diaries when visiting the DHMTs and by email at other times. This meant that the DHMTs received regular support and guidance on the content and format of the diaries. Feedback from the CRTs led to discussions on how the diaries were kept and their contents. This in turn led to modifications in the way the diary was kept. The DHMTs were willing to accept feedback from the CRTs and act on it. This compares to Buchy and Ahmed (2007) who reported that feedback on the field diary was sent once to community partners but no further versions of the diary were thereafter made available for review. This highlights the importance of not only fostering ownership of a diary, but enabling a strong and supportive relationship that allows space for practitioners to respond freely to comments/feedback from academic partners without fear of damaging their relationship. The PERFORM experience is that frequent face-to-face visits to the districts were a key element in strengthening the relationship between the CRTs and DHMTs.

Recording reflection and learning processes is challenging

In interviews, the DHMTs and CRTs said that the concept of ‘diarising’ was relatively new and that DHMTs had not yet “mastered the art of writing”. These statements were made in reference to the mode of keeping a diary to record reflective processes and therefore to make a learning cycle explicit. From the interviews, it was clear that diaries were used to record and monitor project/plan management outputs and whether these were or were not achieved, while not reflecting on the processes for getting to those outputs per se. Again, this may have been because of the work contexts the DHMTs were operating in, as discussed earlier.

Buchy and Ahmed (2007) suggested that the lack of recorded learning in their project was because the NGOs and academics differed conceptually on what was meant by the term ‘learning’. NGO staff felt it meant exchange of information based on field experience; the academics felt it meant a critical loop of activities in which awareness preceded engagement and hence change. However, we do not see these activities as mutually exclusive, since one can inform the other.

AR was used in PERFORM because it was thought to make explicit a process of action and learning that already happened within DHMTs. We can see that AR did indeed capture this. Using the previous example of a fridge that broke down, the DHMT reported an observation (no cold storage); a reflection (we need it); and an action (request to get it repaired). Learning here is expressed as concrete practice and expertise. However, diaries did not record wider reflection – for instance, does the cold storage often break down? Why might that happen? How could breakdown be prevented? Such wider reflection is not found in the written diaries. Without the wider reflection, there is no indication that the HS/HR bundles were modified to create new ways to achieve healthcare objectives though we acknowledge that the duration of the AR cycles recorded by the project may have been too short so that learning and change were not captured.

Regardless, the DHMTs did see an added value in the diaries, particularly in the prioritisation and focus on outputs. This in itself was different to previous monitoring conducted within the routine work environment.

Diaries are kept when inserted into supportive research relationships

The DHMTs did use the diaries for reflection but when this happened, reflection was around (rather than in) the diary. The diary acted as a discussion tool but summaries of the discussions were not written down in the diary. Oral discussion as part of AR reflection process has been reported elsewhere (Faure et al., 1972). Faure, for instance, reports that farmers and their management boards were active participants in workshops and board meetings, and interpreted that participation as being active in reflection and learning (also Buchy & Ahmed, 2007). That is to say, reflection and learning can be rendered explicit through oral discussion as well as in text form, such as in a diary.

The DHMTs undertook reflection within discussion with the CRTs. In Ghana an additional documentation template was used to encourage this. The individual diary recordings were initially transferred into a composite diary (without prompts) which did not incite DHMTs to reflect. Hence, to stir-up reflection, CRT introduced a documentation template which had prompts eliciting observations and reflection on project intervention activities and processes during weekly or fortnightly meetings. The content of these discussions was captured by the CRT. Lessons learned across the participating districts were shared through the dissemination of CRT visit reports which were written after every visit by the CRT. Apart from CRT reflections, the visit reports also contained CRT-DHMTs reflections which enabled best practices to be shared. This implies that local practitioners require support to reflect on their activities, record reflections that lead to learning and translation of those lessons for practical use. For diaries to function as effective reflection and learning tools when kept by busy local practitioners, it is necessary to have facilitators with excellent facilitations skills who can encourage others to reflect. This may require training and ongoing development for both practitioners and researchers.

Conclusion

Ongoing changes in African HSs have opened up opportunities for local healthcare managers to take greater control over health planning and management. AR was identified as a potentially effective way to strengthen HR management. The PERFORM project pioneered the use of diaries to record AR cycles with nine different DHMTs in three different HSs in Africa. Diaries were actively taken up by DHMTs, co-facilitated by African research teams. We see five key learning points on the use of diaries by busy health practitioners within AR initiatives. First, it is important to foster ownership of the diary by the people who are responsible for filling it in. Second, the purpose of keeping a diary needs to be clear and shared between researchers and practitioners from the very beginning. Third, diaries should be allowed to evolve – there is no single best way for practitioners to keep a diary hence the format and structure can change over time so long as it continues to meet the intended purpose. Fourth, it is a challenge for busy practitioners to record the reflection and learning processes that they go through. Last, diaries on their own are not sufficient to capture reflection and learning. The diary needs to be inserted into a supportive relationship to help practitioners in their reflecting and learning processes. Facilitators, whoever they may be, will need training and time to be able to fulfil such a role.

Buchy and Ahmed ultimately recommended to “find a better documenting method” (2007). We argue that busy local practitioners can take ownership of a new recording tool and can find ways to use these that are congruent with their context. The PERFORM experience suggests that when developing an AR project where practitioners are at least partially active in recording AR learning cycles, it is not the tool (diary or otherwise) that is the necessary focus. Rather, a number of tools could work so long as they are introduced in ways that encourage ownership and emphasise the importance of reflection and learning in the local practice setting. Taking account of these factors will influence the ability of local people to take it on, make it useful for their own context, and still be able to generate useable lessons to inform their practices.

Footnotes

Acknowledgement

This document is an output from the PERFORM project: improving health workforce performance in Ghana, Tanzania and Uganda, funded by the European Commission. The project involved a consortium of six partners: Liverpool School of Tropical Medicine, University of Leeds, University of Ghana, Swiss Tropical and Public Health Institute, Institute of Development Studies, University of Dar-es-salaam, School of Public Health, Makerere University. The authors also thank Dr Joanna Raven for her contribution to the development of the Diary in the PERFORM project. The authors would also like to thank Svante Lifvergren for leading the review process of this article. Should there be any comments/reactions you wish to share, please bring them to the interactive portion (Reader Responses column) of the website: .

Author biographies

Comfort Mshelia is a Pharmacist and holds a masters degree in International Health and Management from the University of Aberdeen, United Kingdom. She has worked on several public health projects in the United Kingdom and overseas. Her research interest is mainly on strengthening health systems in Low- and Middle-Income countries as well as improving access to, and use of, medicines.

Gillian Lê is a research fellow in Health Research: Policy & Systems at the University of Leeds. Her research focuses on the health system strengthening needed to help lower and middle income countries achieve Universal Healthcare Coverage.

Tolib Mirzoev is an associate professor in International Health Policy and Systems at the University of Leeds. His expertise and research interests cover areas of Health Policy Analysis, Health Planning, Capacity Assessment and Strengthening, Health Management Information Systems and Human Resources for Health. He teaches at under- and post-graduate levels at the University of Leeds and guest lectures in other UK universities. Tolib is currently leading different health systems and policy research projects and has extensive international experience in strengthening capacity of national health systems in different Asian and African countries.

Samuel Amon, is currently research officer and masters of Public Health Student at the School of Public Health, University of Ghana, Legon – Ghana. He has seven years working experience in research on health systems, tropical diseases, HIV/AIDS, Ebola and Non-communicable diseases. He also has immense experience in action research and, monitoring and evaluation of health interventions.

Ambrose Kessy, PhD, is a senior lecturer at the Department of Political Science and public Administration, University of Dar es Salaam, Tanzania. Dr. Kessy has published chapters in books, referred journals and research reports. He has extensive teaching, research and consulting experience of over 14 years. He is a political scientist who specializes in local governance, policy analysis, strategic planning and management; public health and climate change and human security studies. He has carried out major research and consultancies funded by various local and international organizations.

Sebastian Olikira Baine is a member of faculty at the Makere University School of Public Health, Uganda. He is a medical doctor and has a PhD in health systems. His research is mainly focused on health financing, health policy and health systems.

Reinhard Huss is a Senior Teaching Fellow at the University of Leeds. His expertise is in Governance and Decentralisation, Programme and project planning, Monitoring & Evaluation, District and Quality Management of Health Care Services, Rational Medicine Management and Sexual and Reproductive Health Services. He teaches at Undergraduate and Postgraduate level and is currently the Programme Director for the Masters of Public Health [International]. Reinhard is currently leading in several health systems research and related projects and has extensive international experience in different Asian and African countries.

References

Adjei-Nsiah

Leeuwis

Giller

K. E.

Sakyi-Dawson

Cobbina

Kuyper

T. W.

Van Der Werf

(2004) Land tenure and differential soil fertility management practices among native and migrant farmers in Wenchi, Ghana: Implications for interdisciplinary action research. NJAS – Wageningen Journal of Life Sciences 52(3–4): 331–348.

Ahari

Habibzadeh

Yousefi

Amani

Abdi

(2012) Community based needs assessment in an urban area: A participatory action research project. BMC Public Health 12(1): 161.

Buchy

Ahmed

(2007) Social learning, academics and NGOs: Can the collaborative formula work? Action Research 5(4): 358–377.

Central Intelligence Agency. (2015a). The World Factbook: Ghana. Retrieved from https://www.cia.gov/library/publications/the-world-factbook/geos/gh.html.

Central Intelligence Agency. (2015b). The World Factbook: Tanzania. Retrieved from https://www.cia.gov/library/publications/the-world-factbook/geos/tz.html.

Central Intelligence Agency. (2015c). The World Factbook: Uganda. Retrieved from https://www.cia.gov/library/publications/the-world-factbook/geos/ug.html.

Draper

(2001) Handbook of action research – Participative inquiry and practice, Oxon: Blackwell Science.

Faure

Hererra

Kaddoura

A.-R.

Lopes

Petrovsky

A. V.

Rahnema

Ward

F. C.

(1972) Learning to be, Paris: UNESCO.

Faure, G., Hocdé, H., & Chia, E. (2011). Action research methodology to reconcile product standardization and diversity of agricultural practices: A case of farmers’ organizations in Costa Rica. Action Research, 9(3), 242–260. doi: 10.1177/1476750310388056.

10.

Freudenthal

Ahlberg

B. M.

Mtweve

Nyindo

Poggensee

G.,

Krantz

(2006) School-based prevention of schistosomiasis: Initiating a participatory action research project in northern Tanzania. Acta Tropica 100(1–2): 79–87.

11.

Gormley

McCaffery

(2013) Human resources for health professional development at the district level: Recommendations based on the Ugandan experience , Washington, DC: CapacityPlus.

12.

Gutberlet

(2008) Empowering collective recycling initiatives: Video documentation and action research with a recycling co-op in Brazil. Resources, Conservation and Recycling 52(4): 659–670.

13.

Ka Tutandike. (2008). Facts about Uganda. Retrieved from http://www.katutandike.org/facts-about-uganda.

14.

Marchal

Denerville

Dedzo

De Brouwere

Kegels

(2005) Decentralisation, decision spaces and human resource management at hospital level: High commitment human resource management approaches used by the management team of Ghana’s Cape Coast, Central Regional Hospital Institute of Tropical Medicine, Antwerp.

15.

Marincowitz

G. J.

(2003) How to use participatory action research in primary care. Family Practice 20(5): 595–600.

16.

Mshelia

Huss

Mirzoev

Elsey

Baine

S. O.

Aikins

Martineau

(2013) Can action research strengthen district health management and improve health workforce performance? A research protocol. BMJ Open 3(8): DOI: 10.1136/bmjopen-2013-003625.

17.

Munn-Giddings

Winter

(2001) A handbook for action research in health and social care, (Routledge Ed. London: Routledge.

18.

Quartz. (2015). Tanzania dumps English as its official language in schools, opts for Kiswahili. Retrieved from http://qz.com/355444/tanzania-dumps-english-as-its-official-language-in-schools-opts-for-kiswahili/.

19.

Reason

Bradbury

(2001) The SAGE handbook of action research: Participative inquiry and practice, London: Sage.

20.

Roth, G. (1996, May 3). Learning histories: Using documentation to assess and facilitate organizational learning. Center for Organizational Learning, Sloan School of Management.

21.

Roth

Kleiner

(1998) Developing organizational memory through learning histories. Organizational Dynamics 27(2): 43–60.

22.

Schoepf

B. G.

(1993) AIDS action-research with women in Kinshasa, Zaire. Social Science & Medicine 37(11): 1401–1413.

23.

The Task Force on Health Systems Research (2005) The Millennium Development Goals will not be attained without new research addressing health system constraints to delivering effective interventions, Geneva: World Health Organization, pp. 65.

24.

The World Bank. (2015a). Data – Uganda. Retrieved from http://data.worldbank.org/country/uganda.

25.

The World Bank. (2015b). Data – Ghana. Retrieved from http://data.worldbank.org/country/ghana.

26.

Travis

Bennett

Haines

Pang

Bhutta

Hyder

A. A.

Evans

(2004) Overcoming health-systems constraints to achieve the Millennium Development Goals. The Lancet 364(9437): 900–906.

27.

World Health Organization. (2015a). Countries – Uganda. Retrieved from http://www.who.int/countries/uga/en/.

28.

World Health Organization. (2015b). Global Health Observatory data: Ghana Country profile. Retrieved from http://www.who.int/gho/countries/gha/country_profiles/en/.

29.

World Health Organization. (2015c). WHO African Region: Tanzania. Retrieved from http://www.who.int/countries/tza/en/.