Integrating Care for Diabetes and Hypertension with HIV Care in Sub-Saharan Africa: A Scoping Review

Introduction: Although HIV continues to have a high prevalence among adults in sub-Saharan Africa (SSA), the burden of noncommunicable diseases (NCD) such as diabetes and hypertension is increasing rapidly. There is an urgent need to expand the capacity of healthcare systems in SSA to provide NCD services and scale up existing chronic care management pathways. A scoping review mapped extant policy and evidence based literature on the feasibility of integrating NCD care with HIV in the region. Methods: A scoping review methodology was utilised to conduct a systematic search of peer-reviewed and grey literature published in English language and with no date limitation. A systematic search was conducted on PubMed, Embase, CINAHL, and the Cochrane library. The initial search identified 231 records considered for inclusion in this review. Twelve duplicate records were removed. The remaining 219 records were screened by title and abstract of which 165 records were excluded and 54 records were selected for full-text review. A further 16 records were excluded due to a lack of relevance or the unavailability of the full text article. Finally, 38 were charted and analysed thematically. Results: Thirty-eight studies were included. These comprised a range of different models to integrate NCD and HIV care in the region, reflecting differences in health system environments, and disease epidemiology. The studies provide a variety of evidence that integration of HIV and NCD care can be feasible and can improve clinical effectiveness and identify barriers and facilitators to integration and task shifting. The review confirms that integrated HIV and NCD care services is by-and-large feasible, being both clinically effective and cost-effective. Conclusion: The review may inform the understanding of how best to develop an integrated model of care service by reducing barriers to uptake, linkage and retention in HIV, diabetes and hypertension treatment in SSA countries.


INTRODUCTION
There is an increasing noncommunicable disease (NCD) burden globally, with an estimated 1 billion people living with hypertension and about 9.4 million NCD-related deaths annually [1]. NCDs are important contributors to the burden of disease in countries at all stages of economic development, posing a high priority threat to public health worldwide [2,3]. Over three-quarters of global NCD deaths (28 million) and the majority of premature deaths (82%) occur in low-and middleincome countries (LMICs) [4]. Among global NCDs, type 2 diabetes mellitus (DM) is especially common [5,6]. The International Diabetes Federation recently reported that the incidence of DM would increase from 415 million in 2015 to 642 million by 2040, with more than 70% of the cases in LMICs [7].
Although HIV-infection is a leading cause of premature death among adults in sub-Saharan Africa (SSA), global NCD trends are mirrored in the region. The trend is evident from the rapidly increasing burden of NCDs such as DM and hypertension in SSA countries, giving rise to a dual HIV-NCD epidemic [8]. The Global Status Report on NCDs emphasizes that the negative impacts of NCDs are particularly severe in poor and vulnerable populations such as those living in the SSA region [8], where poverty exacerbates many health conditions [9]. In SSA, the prevalence of hypertension is increasing with 78% of adults over 55 years living with hypertension [10] and prevalence of DM is anticipated to double between 2010 and 2030; with 28 million people in SSA predicted to be living with the disease [11].
Since 2003, significant global investment has facilitated the establishment of HIV care services in the SSA region. The expansion of and improvements in life-saving antiretroviral therapy (ART) has decreased HIV related morbidity and mortality, leading to an ageing population living with HIV who are more susceptible to common NCDs such as hypertension. As such, management approaches in the region have transitioned from acute and emergency care to chronic care, making HIV programmes large-scale chronic disease initiatives. However, prevention, care, and treatment services for DM and hypertension remain inaccessible for most in SSA, and health systems are rarely designed to provide the continuing services required to effectively identify patients at risk, engage them in care, and retain them for long-term treatment [12]. Recent research indicates that only 5-20% of people with DM or hypertension are thought to be in regular care [13]. Therefore, there is an urgent need to expand the capacity of healthcare systems in SSA to provide services for NCDs such as DM and hypertension. As populations are increasingly demonstrating multimorbidity, such as DM and hypertension with HIV [14], health care providers in SSA countries are now faced with an increasing need to manage HIV and NCDs simultaneously.
In the SSA region, as concern about the management of NCDs among people living with HIV (PLHIV) increases, the infrastructure and lessons learnt from the HIV chronic disease treatment model are important resources for a potential plan to expand and operationalise enhanced NCD prevention, care, and treatment [15]. These include health services which are stand-alone and vertically delivered and have been combined with decentralisation and task shifting, allowing primary health centres to treat large numbers of patients with almost 70% of people living with HIV-infection in regular care [16]. Given the similarities between chronic communicable and NCDs from the health system and program management perspective (their effects on health and individual functioning share common pathways and outcomes), the health care systems, tools to diagnose and manage patients, and implementation strategies developed to provide continuity of care for HIV in SSA can potentially be rapidly, efficiently, and effectively utilized to support patients and services for other chronic NCDs [15]. These include decentralised care, task-shifting, counselling, community engagement, drugs and diagnostics procurement and treatment adherence support for the management of HIV as a chronic condition [17,18]. However, though a number of models of integrated HIV/ NCD care have been established in recent years, the lack of evidence-based care models for scaling up makes it difficult for SSA countries to develop effective policy and implementation of appropriate integration strategies [19].
We conducted a scoping review to map and describe extant policy and evidence based literature in the field, examine the feasibility of integrating care for DM and hypertension with HIV care in SSA, and examine how best to leverage infrastructure currently in place for HIV care, for the treatment of DM and hypertension. The review is part of the 'INTE-AFRICA' European Union (EU) funded project [20,21] which is assessing the effectiveness and feasibility of two approaches in Tanzania and Uganda: scale up hypertension and diabetes integrated services alone, or in combination with HIV infection services.

METHODS
A methodologically rigorous scoping review framework comprising an iterative six-stage process developed by Arksey and O'Malley [22] was adopted to undertake a comprehensive search of extant policy and evidence based literature examining the integration of NCD care with HIV care in SSA. The six stages of the scoping review process are described below.

STAGE 1: IDENTIFYING THE RESEARCH QUESTION
The objectives of this review were focused on extant knowledge and gaps around the content, evaluation, and effectiveness of integrating DM, hypertension and HIV care in SSA. Three research questions were developed to guide the review:

STAGE 2: IDENTIFYING RELEVANT STUDIES
The comprehensive three-step search strategy recommended in standard Joanna Briggs Institute (JBI) systematic reviews [23] was utilized in this review in order to identify both published and unpublished (grey literature) evidence. The first step was an initial limited search of relevant databases such as PubMed, followed by an analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. A second search using all identified keywords and index terms was then undertaken across all included databases. Thirdly, the reference list of all identified reports and articles were manually searched for additional relevant studies. Only studies published in English were considered for inclusion in this review. No date limitation was imposed upon the search strategies. Electronic databases searched included: PubMed, Embase, CINAHL, and the Cochrane library. Search terms agreed upon, and the general search strategy are outlined in Figure 1. Citations were managed using the bibliographic software manager EndNote, with duplicates removed manually.

STAGE 3: STUDY SELECTION
The selection process consisted of two levels of screening: (1) a title and abstract review and (2) a full-text review. In the first level of screening, the titles and abstracts of identified articles were screened for inclusion against a set of minimum inclusion criteria (abstract relevant to the topic area). In the second screening, the full text of articles was assessed to determine if they meet the inclusion criteria (full text is relevant to the study title). The PRISMA flow diagram as illustrated in Figure 2 outlines the results of the literature search.
Consistent with the scoping review methodology, this study was broad in its inclusion of different types of literature [22,24,25]. and an assessment of methodological quality was not performed to include or exclude studies based on quality scores [22,24,26].
Both peer-reviewed and grey literature were searched, with no methodological requirement for study inclusion. This facilitated the inclusion of an array of literature, which included quantitative, qualitative and mixedmethod studies, as well as systematic reviews and metaanalyses. Protocols were excluded as they did not provide evidence required, but where relevant and available, studies published based on the protocols identified were screened for inclusion. The eligibility criteria for content were developed according to the JBI reviewer's manual (2015) [27], which suggests the use of the mnemonic PCC (population, concept, and context) to target the desired focus and scope for the review ( Table 1) and the following pre-determined exclusion criteria. All criteria had to be met for inclusion/exclusion in the final review.
• Not in English • Not related to the integrated management of NCDs and HIV • Not based in sub-Saharan Africa

STAGE 4: CHARTING THE DATA
To facilitate comparison and thematic analysis the following data were extracted from the articles:

STAGE 5: CHARTING, SUMMARIZING AND REPORTING THE RESULTS
A thematic analysis was conducted of the charted articles for inclusion in the final review using the World Health Organisation (WHO) definition of integrated service delivery [28]. The initial thematic analysis was completed by the first author (GM) using Braun and Clarke's framework [29]. Steps 4 and 5 of the framework were then carried out together by the first, second and last author.
Two tables were developed to summarize findings from the included studies (Table 2 (Appendix 1) and

STAGE 6: CONSULTATION WITH STAKEHOLDERS
The 209 records identified in the initial database search were shared with members of the 'INTE-AFRICA' consortium [20,21] which resulted in an additional 22 records being suggested.

PATIENT AND PUBLIC INVOLVEMENT
As this was a scoping review it was not appropriate to involve patients or the public in the study.

SEARCH RESULTS
The initial database search identified 209 records. These, along with an additional 22 records suggested by the steering group which were not identified in the initial database search, gave a total of 231 records which were considered for inclusion in this review. After 12 duplicate records were removed, reviewers screened the remaining 219 records by title and abstract, during which 165 records were excluded for reasons showed in Figure 2. Fifty-four articles met the inclusion criteria and were selected for full-text review. Following fulltext review, 16 records were excluded due to a lack of relevance or the unavailability of the full text article (Figure 2), leaving 37 records which were identified to be relevant to the integration of NCDs management with HIV management in Africa. The search process, as guided

Population
The article had to focus on individuals with NCDs (DM or hypertension) and/or HIV, without restrictions on age or sex.

Concept
The article had to describe, evaluate, or propose how NCD management can be integrated with HIV management to improve patient outcomes.

Context
The article had to focus on the integration of NCD management with HIV management in the context of SSA. HIV-NCD integration Rationale, policy bases and models of HIV-NCD integration Models of HIV-NCD integration that were "tested" in the context of developing countries vary but all models indicated that the integrated approach was feasible, effective, efficient and acceptable. However, overall evidence is limited and context-specific evidence is lacking The disease dimension," the "health provider dimension," the patient or "person dimension," and the "environment dimension" of chronic diseases.

OUTCOMES MEASURED FINDINGS
Lessons from present care models for HIV/AIDs and DM show potential for cross-fertilization between models: rapid scale-up approaches through the public health approach by simplification and decentralisation; community involvement, peer support and selfmanagement strategies; and strengthening health services  and Table 3.

MAPPING THE FIELD -EXTENT AND NATURE OF EXTANT LITERATURE
Despite no limitations being placed on publication date in our database searches, the earliest included record was published in 2010. This is understandable, given that integrated HIV/NCD management remains a relatively new area of interest, with the emerging NCD epidemic in SSA only more recently being put under the national and global spotlight, in contrast to the longstanding HIV epidemic that has been recognized for over three decades.
With no limitations being placed on the study design of records included in this review, the included records adopted a wide range of methodologies. Of the 37 records, 13 were observational studies, including crosssectional studies, case-control studies, as well as retrospective and prospective cohort studies. Eight were interventional studies, including pre-post studies, nonrandomized trials, and randomized controlled trials. Two were mixed methods studies, two were qualitative studies, and 13 were reviews, including systematic reviews, scoping reviews and narrative reviews. Study designs of included records are summarised in Table 2 (Appendix 1) and Table 3.

INTEGRATING FINDINGS
Studies included identified various aspects of integrating NCD care into HIV care, which can be broadly classified into three themes: (1)

Evaluation of HIV/NCD Care Integration
Of the 37 records, 24 evaluated various aspects of HIV/ NCD Care Integration. These can be further classified into three sub-themes: 'Current Policies and Integration Models: Summary/Review', 'Feasibility: Capacity, Clinicaland Cost-effectiveness', and 'Facilitators and Barriers to Integration'. These sub-themes are not mutually exclusive, with some records falling under two or indeed all of the three subthemes.

CURRENT POLICIES AND INTEGRATION MODELS SUMMARY/REVIEW
Seven of the 24 records under this theme summarized and reviewed various policies and models related to integrated HIV/NCD management that are currently in place in various African countries. Of these seven studies, one was an observational study [41], one was a mixed-methods study [48], one was a qualitative study [54], and four were review papers [19,39,40,42]. The seven records reported studies in various SSA countries, including Kenya, Malawi, South Africa and Swaziland. Two of the seven records summarized and reviewed current policies in place in SSA countries for HIV/NCD integration [41,54]. These studies highlighted National stakeholders were supportive of integrated policy development, however inconsistencies in implementation where evident between countries. Evidence gaps for costeffectiveness, effects of integration on key HIV and NCD outcomes and funding mechanisms for sustained implementation of integrated HIV/NCD care strategies, were among challenges identified. All seven records summarized and reviewed current HIV/NCD care integration models in SSA countries [19, 39-42, 48, 54]. These included integrated community-based screening for HIV and NCDs in the general population; screening for NCDs and NCD risk factors among HIV patients enrolled in care; integration of HIV and NCD care within clinics; differentiated care for patients with HIV and/ or NCDs; and population healthcare for all. The studies acknowledged integration models are in their infancy, resulting in a lack of clinical and process outcomes data, and a lack of cost-effectiveness data from the various models.  [33,48], two were qualitative studies [51,54], and six were review papers [19,35,39,40,42,43]. Four of the 24 records examined the capacity of HIV/ ART clinics to accommodate integrated NCD care into care delivery [33,41,42,48]. Studies indicated that HIV/ ART programmes have developed considerable systems innovations which could accommodate NCD care. Nineteen records assessed the clinical effectiveness of integrated HIV/NCD care. For the purposes of this study, studies focusing on patient-centred outcomes such as: population penetrance [35,40], disease control as measured using surrogate disease markers [34-36, 39, 42, 46, 53], linkage and retention in care [19,35,36,42,51,54,59,64], quality of care (wait-time, service delivery, comprehensiveness and continuity of care etc.) and patient satisfaction [18,19,31,39,40,42,44,54,56], treatment adherence [40,42,54], and diagnostic yield [19,35,39,40,42,45,47,53,59,64] are classified as studies assessing "clinical effectiveness". Most of the studies concluded in favour of the clinical effectiveness of integrated HIV/NCD care in view of improvements that were achieved in the above-mentioned patient-centred outcomes. These included screening for NCDs within HIV care programmes which improve the identification of undiagnosed NCDs among patients living with HIV and a reduction in duplication and fragmentation of services, which increase efficiency of resource use and help patients remain in care. Five records examined the costeffectiveness of integrated HIV/NCD care [34,36,38,40,59]. These studies support the cost-effectiveness of HIV/ NCD integration efforts while highlighting the additional time burden of NCD screening and testing along with human resources and drugs prices as the major driver of costs.

Arguments for Integrated HIV/NCD Care
Eight of the 37 included records put forth arguments for integrated HIV/NCD care. Of these eight records, four were observational studies [32,37,49,65], and four were review papers [19,60,61,66]. The four observational studies [32,37,49,65] and one review paper [60] examined the prevalence of either NCDs [32,49,65], CVD risk factors [37,65] in PLHIV in SSA countries such as Malawi, Nigeria, South Africa and Zambia. All five studies found a high prevalence of these conditions in PLHIV, and proposed integrated care as a solution to provide holistic chronic care, enhance detection and management of the diseases, enhance NCD prevention, as well as to reduce cost and increase efficiency in a resource-limited setting. The three remaining review papers considered the consequences of the rising dual epidemics on the organization of care and potential bridges between the care of these diseases [60,61,66]. Overall, studies indicated HIV/NCD care integration is a feasible approach to health system strengthening in SSA but must be context specific. Njuguna et al. used case studies to demonstrate a variety of HIV/NCD integration models which programs can adopt and adapt based on available resources, needs and priorities. These included the creation of curricula to produce integrated chronic care providers instead of specific staff dedicated only to HIV or NCD roles and decentralisation to bring care closer to patients to minimize logistical barriers to access, while reducing congestion at 'central' health facilities.

Effectiveness of Task-Shifting/Sharing for NCD Care Delivery
Six of the 37 included records evaluated the effectiveness of various task-shifting/task-sharing interventions for NCD care delivery. Of these six records, two were observational studies [52,55], one was an interventional study [50], and three were review papers [30,57,62]. Task-shifting is defined as the rational movement of primary care duties from physicians to non-physician health-care workers (NPHWs), whereas task-sharing is a planned strategy in which a team of health-care professionals work together to deliver a service, accompanied by training or certification and support for health-care workers [30]. These strategies have been employed in developing countries, in particular in the SSA region, with compelling evidence on feasibility and indicators for success for various disease entities and health sectors including HIV, TB, mother and child health [62], and to a lesser extent, NCDs [52], and have been recommended by WHO for low and middle income countries, due to their low health care worker to patient ratio [52].

SUMMARY OF KEY FINDINGS
The review sought to answer three research questions:

What is the existing policy and evidence based literature regarding the integration of DM and hypertension with HIV care in SSA?
Several studies identified different models of HIV-NCD service integration [19,60,61]. These include NCD prevention and control incorporated into existing HIV services; HIV prevention and control added to primary healthcare already providing NCD services, and simultaneous introduction of integrated HIV and NCD services. In terms of existing policy, studies show that National stakeholders, led by Ministries of Health (MOH), are supporting HIV/NCD integration policy development and implementation through evidence generation and coordination activities [41,54]. However, the degree to which the push for HIV/NCD integration is reflected in national level policy documents varies geographically.

Is it feasible to integrate care for DM and hypertension with HIV care in SSA?
The review indicates that integrated HIV/NCD care in the SSA region is challenging but potentially feasible. However, studies highlight there remain various country specific, contextual, human resource, logistical and infrastructural barriers that need to be considered in order to best achieve efficient and effective integrated care delivery [33,35,36]. One way of overcoming the manpower and infrastructural strain due to the acute shortage of trained health staff in SSA, would be through task-shifting, which has proven feasible as an alternative to rapidly increase the health workforce [30]. This will require enhanced focus on NCD training cascade in integrated services. From the studies included in this review it is not possible to reach a definitive conclusion on whether it is feasible to integrate care for DM and hypertension with HIV care in SSA. Further research is required to generate evidence on the benefits, challenges and cost-effectiveness of

How can we best leverage HIV care infrastructure for the treatment of DM and hypertension in SSA?
While the review indicates multiple ways in which to leverage the lessons and resources of HIV programs when working to expand and enhance DM and NCD services in SSA, it also emphasises no single approach is likely to work in all settings. In some countries, integration of chronic disease services for HIV and NCDs at the point of care may be an effective approach [46]. In other settings this may be neither feasible nor desirable. Studies suggest that countries which have successfully scaled up HIV services have already learned profound lessons about the delivery of chronic care. Using these locally owned and contextually appropriate resources may be an efficient and effective way to kickstart NCD programs and to strengthen health systems to support longitudinal services for all [18,41,65].

HOW THIS REVIEW RELATES TO OTHER LITERATURE
Although this topic has been studied before [37,49,65], the best practices in integrated care provision for the prevention, identification and treatment of hypertension, diabetes and HIV in SSA have yet to be identified. While the potential benefits of HIV-NCD integration for the health system and patients are well documented [67,68], consistent with this review, research indicates differences in health systems preparedness to reconfigure the health system environment when it comes to chronic care delivery among the diverse SSA countries make implementing an integrated care model for HIV-NCD challenging [69]. The majority of models of HIV/NCD integration that have been trialled or implemented in SSA in this review and the wider literature have highlighted the need for integration programs to be context-specific, with no single approach likely to work in all settings [70]. Therefore, an enhanced contextual understanding of integration is further warranted. Service experience and patient outcomes are very dependent on understanding cultures and contexts surrounding integrated service design, and the inputs and actions of policy makers, healthcare providers, clinicians, patients, communities and international donors With these in mind, the policymakers at national and regional levels can then devise a contextually appropriate strategy that maximises existing resources and leverages upon the strengths of the local health system while preventing, and prevailing cracks from hurting hard-earned HIV gains and quality of clinical care.
Implications for research and policyWhile the quantity and quality of research on integrated HIV/NCD care are and have been steadily increasing over the past decade, there remains limited evidence about integration both in terms of scope and generalizability of the evidence [16]. Existing evidence is limited to small scale feasibility studies in largely different contexts, lacking large scale or randomized and/or controlled evaluations [71]. There is also a general lack of clinical and process outcomes data, and of cost-effectiveness data from the various models, which are often required prior to large-scale HIV/NCD intervention programs being rolled out. In addition, context-specific evidence needed to adapt an integrated care model that is best suited to the needs of a specific country is lacking. These evidence gaps are often cited as a barrier to integration in SSA countries, suggesting an urgent need to fill these gaps to further inform implementers and policymakers. Hence, further research is warranted to gather not just more, but higher quality evidence on outcomes and cost through randomized controlled trials, as well as country-specific evidence at operationalisation and scale up over time. Evidence on regional, national and subnational contextspecific variations in burden of NCDs in PLHIV should inform policy decisions, priority-setting and resource allocation, while evidence on the impact of HIV/NCD integration on the existing HIV care model should inform policymakers on the likely effects of integration on HIV targets, indicators and systems. Efforts should further concentrate on both HIV and NCD awareness raising tactics and prevention activities. Additionally, further research is required to understand public understanding of the HIV/NCD burden of disease and integration of services, understand and respond to the scaleup process and determine the sustainability of integrated care models. It is the aim of the 'INTE-AFRICA' project [20,21] to address these evidence gaps, along with the various barriers to integration identified in this review by conducting a randomised control trial, health economic and process evaluation in the operationalisation and scale up of integrated clinics in Tanzania and Uganda

METHODOLOGICAL CONSIDERATIONS
Several limitations should be considered when interpreting the findings of this review. Whilst we adopted the rigorous scoping review methodology and used a comprehensive search approach, there is a possibility that not all publications relevant to the inclusion criteria were identified by the searches or databases used. Scoping reviews do not include an assessment of study quality as the focus is on covering the range of work that informs the topic rather than limiting the work to studies that meet particular standards of scientific rigour. Finally, only articles published in English were considered for inclusion in our review, which could have resulted in the exclusion of equally relevant literature published in other languages used in the SSA region such as Portuguese, French and local African languages.

CONCLUSION
Our review illustrates that HIV/NCD care integration in sub-Saharan Africa is potentially feasible, with several models available that countries can adopt and adapt based on available resources, needs, and priorities. However, it is important to note that existing evidence is largely limited to small-scale feasibility studies in varying contexts, with a paucity of higher quality research measuring clinical outcomes, cost-effectiveness and the sustainability of such integration. Conducting high quality trials and implementing their findings will enable the optimization of existing resources and enhance the outcomes of NCD detection, treatment and care for all patients in a manner that is both cost-effective, and which does not weaken the well-functioning HIV efforts in the SSA region.