Implementing evidence-based task-strengthening strategies for hypertension control within Ghana’s Community-based Health Planning Services: a concept mapping analysis

As efforts are made to systematically implement an evidence-based task-strengthening strategy for hypertension control (TASSH) in Ghana, understanding their content can help national, regional and district health stakeholders within Ghana’s Health Services (GHS) identify strategies for enhancing intervention uptake and sustainability in their local context. This study sought to describe national, regional and district health stakeholders’ perspectives and characterize the array of strategies needed to enhance the uptake of evidence-based TASSH within Ghana’s Community-based Health and Planning Services (CHPS) zones. Methods This qualitative study was conducted among national, regional and district health stakeholders within GHS serving patients who utilize CHPS zones. The CHPS initiative is Ghana’s agship strategy for achieving universal health coverage. Semi-structured interviews were conducted with 22 national, regional and district health stakeholders identied via purposive sampling. Interviews were conducted in 2018. Data analysis was conducted between July 2019 and December 2019. Data was analyzed using group concept mapping systems to identify major themes and subthemes.

strategies for enhancing capacity and training with implementing evidence-based task-shifting hypertension interventions in Ghana.

Background
According to the World Health Organization (WHO), 75% of deaths in Sub-Saharan Africa (SSA) will be attributable to hypertension (HTN) by the year 2020 [1]. Thus, interventions targeted at blood pressure (BP) control are vital to reducing hypertension-related morbidity and mortality [2,3]. However, socioeconomic barriers, lack of insurance coverage, uncoordinated care, and shortage of physicians limit the capacity of SSA countries to implement Cardiovascular Disease (CVD) prevention at the primary care level [4][5][6][7][8][9]. Of these, systems-level barriers have a more adverse impact on health outcomes [10]. Thus, poor access to care and limited availability of laboratory tests are major barriers limiting the capacity of countries in SSA to manage a chronic disease like HTN [5].
The situation in SSA is compounded by the acute shortage of health workers available to implement CVD prevention at the primary care level. For example, 36 of the 57 countries that the WHO identi ed with a critical shortage of health workers are in SSA [11]. Although SSA harbors 11% of the world's population, it bears over 24% of the world's disease burden and only has 3% of the global health workforce [11]. Shortage of physicians is a major barrier to HTN control in Ghana, whereby in 2017, the density of doctors, nurses and midwives were 2.65 per 1000 population [12,13]. The shortage of healthcare providers hinders Ghana's capacity to control HTN at the primary care level, where the majority of patients receive care [14]. Thus, there is a need to implement and integrate bold innovative strategies into already existing infrastructure within the Ghana Health Service such as the successful Community-Based Health Planning and Services programme (CHPS) [15]. The CHPS concept is Ghana's key strategy for changing the focus of primary health care from the district and sub-district health facilities to community and doorstep locations [15,16].
Studies have demonstrated the effectiveness of using the evidence-based Task-strengthening Strategy for high blood pressure control (TASSH) model tailored to the needs of community health nurses to deliver hypertension care [17,18]. In 2015, we established in a cluster RCT in 32 district hospitals and community health centers in Ashanti Region, Ghana, that an evidence-based Task-Shifting Strategy for HTN Control (TASSH) based on the WHO Cardiovascular Risk Package and the provision of health insurance coverage (HIC) to patients, delivered by community health nurses (CHNs) led to a 20.4 mmHg mean reduction in systolic blood pressure (SBP) for TASSH + HIC group and 16.8 mmHg mean reduction in SBP for HIC group at 12 months [17].
Understanding how to systematically implement evidence-based TASSH interventions in Ghana, and its content can help national, regional and district health stakeholders within Ghana's Health Services. This study utilized Group Concept Mapping (GCM) to examine factors likely to in uence the adoption of TASSH within the CHPS zones. Group Concept Mapping (GCM) is a mixed-methods design that engages participants through the research process in well-structured qualitative and quantitative methods [19][20][21][22][23][24].
The approach allows researchers to capture, organize, and rate conceptual data from individual and groups. It as well creates a meaningful conceptualization of the issues under investigation [20][21][22][23]. We thoroughly used GCM to evaluate national, regional and district health stakeholders' perspectives and characterize the array of strategies needed to enhance uptake of evidence-based TASSH within Ghana's Community-based Health and Planning Services (CHPS) zones.

Participants
Semi-structured interviews were conducted with 22 national, regional and district health stakeholders identi ed via purposive sampling. The stakeholders who participated are national policymakers with the Ghana Health Service and Ministry of Health , Brong Ahafo regional policy implementors with the Ghana Health Service, as well as district-level policy implementors within the Kintampo North and South districts. Participants were identi ed through the steering committee group formed to guide the implementation of community-based management of hypertension. Information on context as well as ways to improve practice capacity for the uptake of the study intervention was collected from the identi ed stakeholders. All procedures as described were approved by three independent ethics review boards. These ethics review boards are the Ghana Health Service Ethics Review Committee, the Committee on Human Research Publication and Ethics, Kwame Nkrumah University of Science and Technology as well as the Kintampo Health Research Centre Institutional Ethics Committee. Written informed consent was obtained from all participants.

Procedure
To better understand the context and identify practice capacity for the uptake of TASSH in CHPS compounds the study utilized the six steps of GCM [25] which are described below: 1) Preparation Stage: In this phase, stakeholders were identi ed a priori, and the focus prompt to guide the brainstorming exercise was developed. The focus prompt stated the following: "For the CHPS compounds nurses to successfully use TASSH ("identify, treat, refer") for hypertension control, they need?" This exercise was done face-to-face such that the project's goal was explained to participants before the concept-mapping exercise commenced.
2) Generation Stage (brainstorming stage): In this phase, key stakeholders participated and brainstormed a set of statements related to the focus prompt developed in the preparation stage. The stakeholders identi ed across the various levels of leadership provided statements in response to the focus prompt. Demographic data were collected for each of the respondents.
3) Structuring Stage (sorting and rating): In this phase, each participant sorted the statements into piles based on similarity and then rates each statement based on its perceived importance (i.e. how important is this strategy) and feasibility (i.e. how feasible is this strategy) of the statements generated in the previous phase.
After the participants sorted the statements, they rated each statement on a 4-point Likert scale based on importance and feasibility. 4) Representation Stage (concept-mapping data analysis): The concept mapping data analyses were conducted using the Concept System software. The sorted data was used to conduct multidimensional scaling (MDS) analysis with a two-dimensional solution as well as hierarchical cluster analysis [26,27].
The MDS analysis was based on the measurement model that assumes that the relative similarity of objects can be represented in terms of the relative distance between pairs of points [27]. To indicate the goodness of t, a "stress value" of the point map was developed to determine how well the MDS solution maps the original data [27]. With concept mapping, a lower stress value indicates a better t and re ects a stronger relationship between the optimal and actual con gurations [26,27]. Furthermore, hierarchical cluster analysis (HCA) was conducted using the two-dimensional x-y coordinate data obtained from the MDS analysis as input and applying Ward's algorithm as the basis for de ning clusters [27,28]. This approach forces the cluster analyses to partition the MDS con guration into non-overlapping clusters in two-dimensional space [26,28]. This technique also grouped the outcome statements on each map such that statements placed in the same cluster were located at contiguous areas of the map. [26,28] The resulting output was a "cluster map" which revealed how the statements generated, as represented by points, are grouped [27]. Also, mean ratings for importance and feasibility, respectively, of each statement and cluster were estimated. A pattern-matching graph was used to graphically show the agreement in mean cluster ratings between what is feasible versus what is important across the various stakeholders [22]. Another output from the pattern matching graph is Pearson's product-moment correlation coe cient (r) which is an indication of the strength of the relationship between the variables of interest (in our case feasibility and importance) [22]. Also, the t-test is often used to compare the mean ratings. [22] Lastly, a two-dimensional Go-zone graph was generated to visually plot the ratings of each statement on both importance and feasibility concurrently. [22] This graph puts statements that were rated above the mean for both importance and feasibility positioned in the top right quadrant (Go-zone). [22] This shows high priority statements (activities/actions) that will facilitate the uptake of evidence-based task-shifting strategies for hypertension control in Ghana.

5) Interpretation Stage:
Interpretation of results is a real-time, participatory process where stakeholders interacted with the ideas generated. [22] This activity included examining the cluster maps to determine which clusters of statements were rated most important to the focus statement, examining the patternmatching to determine key areas to target based on high ratings, and examining the go-zones to determine the area of most importance for each stakeholder group [26]. This activity ensured that all participants were involved in explaining the visual maps and graphs. 6) Utilization Stage: In this stage, we worked with the stakeholders to determine the best ways to use the maps and reports produced by the GCM procedures. The concept mapping output includes creating priority areas for improving practice capacity for TASSH uptake which in turn served as the basis for planning the implementation of TASSH at CHPS compounds.

Results
Characteristics of participants Table 1 describes the demographic characteristics of the participants. Of the 22 participants 68% were male, with a mean (SD) age of 40 (7.5) years, and had on average 9 (SD: 7.04) years of experience providing hypertension-related care within the Ghana Health Service. mean (SD) of 9 (7.04) years of experience providing hypertension-related care within GHS A six-cluster map was considered to be the most suitable with statements within each cluster rationally belonging together ( Figure 1 and Table 2). The conceptual map consisted of 46 strategies needed for implementing evidence-based TASSH, organized into 6 clusters: 1) Referral Systems; 2) Availability of Equipment; 3) Protocols and Guidelines; 4) Capacity Building/Training; 5) Policy Reform, and 6) Technical Support and Supervision. The number of statements in the clusters ranged from 3 (referral system) to 15 (Capacity building/training)  Go-Zone Figure 3 shows the Go-Zone ratings generated for all the 46 statements and the 6-cluster map.
Statements rated above the mean for both importance and feasibility are shown in the Go-Zone (the green shaded area of Figure 3). The Go-zone statements were drawn from all the six clusters: 1) Referral Systems; 2) Availability of Equipment; 3) Protocols and Guidelines; 4) Capacity Building/Training; 5) Policy Reform; and 6) Technical Support and Supervision.

Discussion
Context matters in implementation efforts [29,30]; thus, it is important to understand the key factors that will enhance or limit the effective implementation of evidence-based interventions in resource-constraint settings like Ghana. The main objective of this study was to identify ways to describe national, regional and district health stakeholders' perspectives and characterize the array of strategies needed to enhance the implementation of evidence-based TASSH within Ghana's Community-based Health and Planning Services (CHPS) zones.
Using the GCM, 46 statements grouped under six clusters highlighted stakeholder's perceptions of factors likely to in uence uptake of TASSH including contextual issues that should be considered and developed with key stakeholders explicitly for the implementation of a task strengthening approach to community-based management of high blood pressure. These ndings are relevant particularly in LMICs seeking to implement community-based strategies to ameliorate the burden of high blood pressure.
These six clusters include: 1) Referral Systems; 2) Availability of Equipment; 3) Protocols and Guidelines; 4) Capacity Building/Training; 5) Policy Reform, and 6) Technical Support and Supervision. Cluster 1 highlights how useful a good referral system can facilitate the implementation of the task strengthening strategy to improve hypertension outcomes. Cluster 2 highlights the essence of providing the needed equipment to implement an intervention like this such as the sphygmomanometer, weighing scales as well as height measuring devices. Cluster 3 and 4 while distinct share similar factors for the intervention's implementation such as capacity building, provision of protocols and guidelines suitable for a community-based facility (such as the CHPS zones) to undertake the mandate of screening and referring individuals with hypertension to the next level of care (health centers or district hospitals) for the required treatment. Cluster 5 and 6 also highlight key health systems strengthening strategies that facilitate this community-based approach. The stress value of 0.225 reported by this study lies within the 95% con dence interval (0.205-0.365) reported by a meta-analysis of concept mapping projects which is an indication that nal cluster was a representative of the stakeholders perspectives [20,31].
A pattern match compared importance and feasibility ratings by the cluster. The cluster "availability of equipment" had a high mean rating (mean 4.80 out of 5) for importance followed by "capacity building" as well as both clusters had a high proportion of statements placed within the Go-zone. Also, the cluster "capacity building" received the high mean rating (mean 4.20 out of 5) for feasibility followed by the cluster "availability of equipment". This is an indication that these two strategies were regarded as the most feasible and the most important strategies for the successful implementation of any communitybased hypertension management program. These ndings are similar to a study conducted by Blackstone et al., 2017 in the Ashanti Region of Ghana, that found that personnel training was the most feasible and important intervention component [32]. Although important (mean 4.40 out of 5), Policy reform was rated as the least feasible strategy to address. In this study, the overall correlation between the ratings for importance and feasibility was moderately positive (r = 0.67). This is an indication that participants opinions on importance aligns with what is considered feasible. The degree of slope of the lines (in Fig. 2) between importance and feasibility demonstrate this alignment. For example, there was much alignment between importance and feasibility to implement indicators within the cluster protocol and guidelines domain. Also, participants agreed on the relative low importance and feasibility of policy reforms as a strategy required for the successful implementation of a community-based hypertension control programme like uptake TASSH.
The statements in the cluster "availability of equipment" offers a rich repository of ideas for identifying the needed logistics and immediate next steps for the successful implementation of a health system strengthening initiatives for hypertension control. Furthermore, statements in the cluster "capacity building" highlights the need to develop a capacity building curriculum that includes educational and training materials for the health workers which will be used in the task strengthening training modules. These statements were considered and combined to design the interventions and training package that re ects speci c targets and contexts. Although, policy reforms had the least mean rating for both importance and feasibility, the statements in the cluster provides pragmatic implementation strategies that can be implemented within the primary care.
There are two key messages from this formative phase activity. To begin with, for a successful implementation of a community-based hypertension control programme requires multicomponent, multi-stakeholder action and cooperation. Outputs from this study signals 6 broad areas for interventions as well as speci c requirements for action across the various level of care.
Secondly, our work underscores the importance of incorporating the perspectives of healthcare leadership (different stakeholder groups) in highlighting strategies that will be useful for the implementation processes of community-based task-shifting strategies for hypertension management and control. Encouragingly, some of the identi ed strategies mimic the Ministry of Health strategies sets out to prevent, control and manage non-communicable diseases (NCDs) including hypertension. For example, the policy document as part of its strategic areas of implementation specify health system strengthening; particularly in terms of capacity building and provision of logistics as a priority for preventing and managing NCDs at the community level. [33] Strength and Limitations Key contributions of this study are the identi cation of key strategies for the implementation of community-based strategies for hypertension control. The stakeholders who participated in this study work at GHS and are quite knowledgeable about the inner works of the health system in Ghana. The understand the capacity for success of an intervention and their numerous years in the eld provide a keen insight that will help with the modi cation of a tasks strengthening intervention. The diverse group of stakeholders who hold positions at the national, regional and district level provide a varied scope of recommendations that can be targeted at each tier of the health system in Ghana when considering the scale of implementation of the intervention in the health system. Statements grouped into clusters are a useful source of information for researchers, policy makers and policy implementing agencies to select and combine to design and implement community-level strategies for hypertension control across different context and various cascade of healthcare. The statements suggested and grouped into clusters provides other implementation strategies which can be considered for future research for its effectiveness, Despite these strengths this study, the team note the following limitations. Although the study sampled and included various stakeholders across different levels of healthcare leadership, the small sample size of 22 may limit the generalizability of the study ndings. Albeit this limitation, the study draws on the perspective of key individuals who formulate policy as well as implement policy across the Ghana healthcare system. Therefore, the views shared cut across the various levels of healthcare and are relevant to the existing system. Additionally, we did not include the community health o cers who were going to implement the task strengthening interventions. However, the participants included leaders of these community health o cers which implies that their views may not be different from the community health o cers.

Conclusions
Through an established concept mapping methodology, our study highlights health systems and healthcare leadership perceived factors pertinent to the design and implementation of community-based task-shifting strategies for hypertension management and control. Our ndings point to a prioritized set of speci c initiatives for the effective implementation of a community-based hypertension programme in Ghana across 6 thematic areas. Furthermore, ndings from this study will inform implementers and policymakers as to the best strategies for implementing community-based hypertension interventions in resource-constraint settings to optimize feasibility and acceptability in the local context. Data presented by this study will be useful in chatting a common agenda for a successful implementation of any form of community-based hypertension control programme. Availability of data and materials