Behavioral and cognitive interventions to improve treatment adherence and access to HIV care among older adults in sub-Saharan Africa: an updated systematic review.

BACKGROUND
Approximately 14% of Africans infected with HIV are over the age of 50, yet few intervention studies focus on improving access to care, retention in care, and adherence to antiretroviral therapy (ART) in this population. A review of the published literature until 2012, found no relevant ART management and care interventions for older people living with HIV (OPLHIV) in sub-Saharan Africa. The aim of this systematic review is to update the original systematic review of intervention studies on OPLHIV, with a focus on evidence from sub-Saharan Africa.


METHODS
We conducted a systematic review of the available published literature from 2012 to 2017 to explore behavioral and cognitive interventions addressing access to ART, retention in HIV care and adherence to ART in sub-Saharan Africa that include older adults (50+). We searched three databases (MEDLINE, EMBASE, and Education Resources Information Center) using relevant Medical Subject Headings (MeSH) terms as well as a manual search of the reference lists. No language restrictions were placed. We identified eight articles which were analyzed using content analysis with additional information obtained directly from the corresponding authors.


RESULTS AND DISCUSSION
There were no studies that exclusively focused on OPLHIV. Three studies referred only to participants being over 18 years and did not specify age categories. Therefore, it is unclear whether these studies actively considered people living with HIV over the age of 50. Although the studies sampled older adults, they lacked sufficient data to draw conclusions about the relevance of the outcomes of this group.


CONCLUSIONS
These findings underscore the need to increase the evidence-base of which interventions will work for older Africans on ART.


Background
In 2015, an estimated 25.5 million people were living with HIV in sub-Saharan Africa [1]. The global HIV epidemic shows a growing number of people aged 50 years and older who are living with HIV [2][3][4]. According to a report by UNAIDS [5], 9% of the people living with HIV (PLHIV) in sub-Saharan Africa were over the age of 50 in 2012.
The increasing proportion of older people living with HIV (OPLHIV) is related to three main factors: (1) the success of antiretroviral therapy (ART) in prolonging the lives of PLHIV; (2) decreasing HIV incidence among younger adults, which shifts the disease burden to older ages; and (3) the fact that people aged 50 years and older exhibit similar risky behaviors as younger people [5]. Although the increasing numbers of PLHIV over 50 years has important implications for HIV responses, very few HIV strategies in low-and middle-income countries (LMIC) explicitly focus on this previously hidden population [5][6][7].
An exploration of the factors affecting the general health-seeking behaviors of the 50+ age group in sub-Saharan Africa shows several barriers. Some of these barriers to health care include lack of money for transport and care services, unkind treatment by health workers and health staff who have few resources and little understanding of older persons' needs [8][9][10]. Because of the increased vulnerability of older people seeking health care services within the health systems of LMIC in general, these barriers are possibly exacerbated for those who are experiencing any sort of ill-health [10].
The barriers to access to care and treatment services experienced within the general older population, including those who are unwell, are potentially compounded among those who are HIV-positive. For example, because of cultural norms around respect for older persons and against intergenerational discussions of sex and sexuality, older persons may be reluctant to honestly discuss their symptoms and diagnosis with younger health care staff [8,11]. In addition, older adults who live with HIV are more likely to have co-morbidities, resulting from the aging process and/or because of their HIV infection [12,13].
Evidence suggests that OPLHIV may have greater access to health services than those without HIV [14][15][16]. This is because, the presence of both HIV and non-communicable diseases makes it particularly important to pay attention to the effect of concurrent health conditions. Consequently, emphasis is laid on access to care and treatment adherence in this population. This phenomenon is described as the ART advantage [17].
There are also emerging data from high-income countries that older adults' access and adherence to ART are distinct [18]. Although it appears that adherence is better among older populations [19], cognitive impairment can reduce adherence and treatment outcomes [20]. Regardless of this, the preponderance of research around the factors affecting adherence and access to care and related interventions in sub-Saharan Africa to address these problems focus exclusively on adults under 50 years of age [7,[21][22][23][24][25][26]. It is, therefore, necessary to consider the development of context-specific interventions to address the specific needs of older Africans living with HIV.
Despite the clear need, there have been a limited number of intervention studies that aim to improve access to care and ART adherence among OPLHIV [20]. A review published in 2014 of the literature on ART care and adherence published up until 2012 found no relevant interventions for OPLHIV in Africa [7]. The evidence for interventions in high income countries is relatively limited with three reviews published in 2014 identifying only a handful of interventions directly addressing access to ART or adherence and as noted by Negin et al. [7] even these have many limitations [7,27,28]. Although not within the scope of this review, a search of the literature after 2014 suggests that this pattern has continued with only a few relevant publications on interventions about OPLHIV from high income countries found [29][30][31]. There is, therefore, a dearth of evidence for interventions that address the needs of OPLHIV in Africa especially within sub-Saharan Africa where almost 10% of PLHIV are those over 50 years [32].
To this end, we sought to update the original systematic review of intervention studies published by Negin et al. [7] with a focus on evidence from sub-Saharan Africa. We explored whether, in the recent upturn in research into OPLHIV [3,[33][34][35], there has been an increase in evidence for effective behavioral and cognitive interventions to address older Africans' access to care and adherence to ART. Since evidence suggests no specific ART behavioral and cognitive interventions in sub-Saharan Africa exclusively focuses on older adults, we also explored interventions in sub-Saharan Africa that include older adults to assess insights into which types of interventions might benefit OPLHIV.

Methods
We conducted a systematic review with a content analysis approach of the published literature from 2012 to present in MEDLINE, EMBASE, and Education Resources Information Center. We supplemented the search of these databases with a reference chase of the identified articles. Because this is an update of a previously conducted systematic review, we did not register it in PROSPERO.

Selection of studies
The following Medical Subject Headings (MeSH) terms and combinations were used to search the identified databases [antiretroviral therapy/agents, Anti-HIV Agents, Highly Active Antiretroviral Therapy, medication adherence/adherence, Patient Compliance/or Medication Adherence/"Patient Acceptance of Health Care," "middle aged (45 plus years)," (aging or seniors or older), Texting or Text Messaging/Motivational Interviewing/or Counseling/Behavior Therapy/or behavioral intervention, Patient Education]. We also conducted a manual search of the reference lists of the articles identified from the databases to complement the electronic search. We defined a search strategy to retrieve from the databases and to manually search, papers that were peer-reviewed, from sub-Saharan Africa, that assesses behavioral or cognitive interventions to address challenges of access to care, treatment, adherence to ART, and retention in care among older adults (50+). No language restrictions, which is in line with Negin et al. [7]. The inclusion/exclusion of the identified articles was guided by the Population, Intervention, Comparison, Outcome and Time (PICOT) mnemonics (Table 1).
Only papers that reported outcomes related to access to care, treatment adherence, and retention in care interventions conducted in sub-Saharan Africa, and including older adults were selected for the final review (Fig. 1). Our timeline started from 2012 to 2017 because the review conducted by Negin et al. [7] covered the previous years. Our search was meant, therefore, to complement what had been previously done by Negin et al. [7]. Data from relevant articles were extracted into an excel spreadsheet.
The NIH-NHLBI Quality Assessment of Systematic Reviews and Meta-Analyses were used to rate (good, fair, and poor) the quality of studies (Table 2) [36]. Six of the eight articles were rated as 'good' and the remaining two were rated 'fair'.
Thematic analysis was used to identify major themes in the chosen studies. We followed the PRISMA guidelines to report the study.

Data collection
The extraction of the data was conducted by FCM and LK. ES checked the extraction and resolved any 'disagreements' that emanated from the extraction process. Extraction of data from the identified papers was done under the following topics: (1) study citation and setting, (2) intervention type, (3) focus of intervention, (4) study design, (5) outcome measures, (6) study quality, and (7) detailed description of outcome. Refer to Table 2 for the data extraction process.

Data analysis
We reviewed the identified studies using content analysis approach-a family of procedures for the systematic, replicable analysis texts. Content analysis is a process that involves the classification of parts of a text through the application of a structured, systematic coding scheme from which conclusions can be drawn from the message content [37]. Our study entailed quantifying the number of studies describing behavioral and cognitive interventions that included OPLHIV. Therefore, we coded for intervention type and focus, charateristics of the sample with a focus on the age category of the study participants, study design, measures used, quality of the study, and the outcomes of the interventions (significance).
After the coding process, we searched through the selected articles to identify the consideration of age in the selected studies, specifically the inclusion or specification of those older than 50. For those articles that did not explicitly identify the PLHIV over the age of 50 in their sample, we contacted the authors inquiring if they had an age breakdown of their sample. The codes related to outcomes and their relationship to age were also reviewed to assess whether any of the interventions included in the review have lessons for future interventions to address ART adherence and access to HIV care among older people in sub-Saharan Africa.

Results
We identified eight articles published from 2012 to 2017 that assessed the effectiveness of behavioral and/or cognitive interventions to improve adherence to ART medication and retention in ART care that include persons aged 50 and above in sub-Saharan Africa. All the studies identified adopted quantitative research methods. Table 3 shows a summary of study results. Five of the studies were randomized-control trials, and the other three used a pre-test/post-test with no control group research design.
The interventions evaluated in the studies included text-messaging (2) [38,39], counseling and patient education (3) [40][41][42], home visits (1) [43], and a combination of interventions (2) [44,45]. Five of the eight studies specified the age distribution of participants; nevertheless, only in one study [43] did those over the age of 50 make up a significant proportion of the study population. Three studies referred only to general age range (for example 18 years or older) and did not specify age categories and, therefore, it is unclear whether they actively considered those over the age of 50 [38,39,42]. We contacted the corresponding authors of the other three articles by email, and they provided a breakdown of the sample included in their studies.
Based on this additional information, we obtained the age distribution of all eight studies included in the review. The articles' review revealed that in most of the other studies, the size of the over 50s was about 10% or less. From the additional information obtained by emailing the corresponding authors, we had one study that had up to 21% of the participants within the age of 50+ [39] with the others conforming to the findings based on the paper reviews (10% or less of the sample). There were no studies that exclusively focused on older adults.
Only three of the selected studies had the results of the interventions broken down by age. The first, which was an assessment of the impact of a peer education intervention on viral load and adherence, found that adults over the age of 40 years were less likely to achieve viral load suppression compared to other age groups within the study [40]. The second study aimed to assess the impact of patient education on ART adherence. The authors found no significant difference in adherence before and after the intervention among those over the age of 55 years. However, for those aged between 36 and 55 years, there was a significant increase in adherence [41]. The third study assessed the impact of a combination intervention of counseling and home visits on attendance in pre-ART care. The results showed that those aged 45-70 years showed greater retention after the intervention [44]. A range of outcome measures was used to assess access to care and adherence to ART. Of all the studies included, only two studies assessing adherence to ART used a biological marker. In addition, the study by Bigogo et al. [43] assessed retention in access to care through the incidence of illness. Of the three studies reviewed above that specified differences by age, only Coker et al. [40] included a biometric measure: viral load. This study also collected self-reported adherence and cumulative pharmacy refill rates to understand adherence to ART (Coker et al. [40]). Kunutsor et al. [41] assessed adherence using self-reported adherence through pill counts at the clinic, and Lubega et al. [44] used attendance of pre-ART care visits over a period of 24 months to show retention in care; neither of these studies included biological markers.

Discussion
Even with the recognition that HIV among older persons in sub-Saharan is an on-going and a potentially growing problem as the numbers of individuals on ART and aging with HIV increases, our review and previous studies show that there is a dearth of evidence on interventions targeting older adults [7]. It is notable that in our review, not a single paper outlined an intervention specifically targeting OPLHIV to address their access to care or treatment adherence. In addition to failing to specify age or adequately consider older people, some of the studies did not include a control group, while others failed to specify any comparison by age limiting the number of studies that could actually be analyzed for potential outcomes vis-à-vis the age of the study participants. The results show that even those intervention studies that include older adults lack sufficient data to draw useful conclusions about the impact of the intervention on older adults.   Cognitive behavioral 104 were purposefully (via announcement) selected for participation. Selected participants were randomly assigned into two groups.
Each group was allocated 52 participants.

Good
Using the intention to treat analysis, the results showed that 76.9% of those in the intervention group achieved adherence to ARVs compared to 55.8% in the control group. The authors concluded that combining counseling with text message reminders significantly improves drug adherence.  The studies included in this review considered a range of outcomes. However, only two included bio-markers limiting the studies to the reliance on self-reported and subjectively measured outcomes, which could introduce potential flaws. In addition, the broad range of behavioral and cognitive outcomes assessed within the selected articles limits the true comparability of the interventions. It also limits the clear evidence of the ability of interventions to change behavior.
The Coker et al. [40] study suggests that the peer education intervention was not successful for those over the age of 50, but the poorly defined age range and composition in the oldest group (40 years plus) makes it hard to draw any clear and significant conclusions. While Kunutsor et al. [41] provide sufficiently specific results to draw conclusions about the intervention's potential to meet OPLHIV's needs, the results show that there was no significant change as a result of the intervention. Therefore, only in the Lubega et al. [44] study is there a suggestion that the intervention may have been successful in retaining those between 45 and 70 years in pre-ART care. Nevertheless, the generalizability of this study is limited and with the current move to Universal Test and Treat and to supporting ART adherence, there is less relevance of a study of pre-ART care. In addition, the age range is relatively large and the composition is not clear, therefore, the results cannot clearly show an impact on persons over the age of 50 specifically.

Limitations
The use of only three databases for the search of the relevant literature poses the potential risk of missing studies that could have met the inclusion criteria. The risk of missing relevant studies was reduced by conducting a manual search of the relevant studies from the reference list of the papers that were identified from the databases. This study also assumed that the previous review on which it is built was thoroughly conducted as any studies that were found prior to 2012 were not included.

Conclusion
The findings of this study underscore the need to increase the evidence-base of interventions that will assist older Africans to access ART, remain in ART care, and adhere to treatment. Ideally, future studies to assess and improve ART access and adherence in sub-Saharan Africa should (a) target older persons living with HIV explicitly, (b) include control groups and sufficient sample sizes for statistical testing, and (c) include bio-markers and validated behavioral and cognitive outcomes. With this evidence, health policies and programs are more likely to address the needs of those aging with HIV.  The corresponding authors of these articles were contacted by email to provide the sample breakdown of their studies