Comparison of Antiretroviral Therapy Adherence Among HIV-Infected Older Adults with Younger Adults in Africa: Systematic Review and Meta-analysis

Globally, there are 36.7 million people living with human immunodeficiency virus (HIV), of which 69% live in Africa [1]. UNAIDS estimated that in 2016 there were 5.8 (5.3–6.2) million people aged 50 years and older living with HIV, representing 17% of all adults aged 15 years and over living with HIV [2]. Their modelling data also predicts that the number of older adults living with HIV in low- and middle-income countries around the world will increase by 47% by 2020 and most of them will be living in sub-Saharan Africa (SSA).

Over the past decade there has been an increase in the access to life-prolonging antiretroviral therapy (ART), and factors such as improved HIV surveillance, clinician awareness of HIV, advancement of diagnostic techniques etc., have led to an increase in people aged 50 years and older living with HIV [3, 4]. As life expectancy in Africa is lower than developed countries around the world, WHO projects in Africa have categorized older adults as people aged 50 years and older [5]. This definition of older adults for reporting HIV incidence and ART adherence, in low and middle income countries, is well documented in the in published literature [6, 7].

With the increased life span, ageing population is becoming susceptible to other causes of ill health [8, 9], which in turn may affect their adherence to ART. Evidence shows that poor immunological response to ART is a surrogate of biological ageing [10]. The natural ageing process causes a decline in immunocompetence making older adults susceptible to both communicable and non-communicable diseases such as diabetes and hypertension when compared to young adults [11‐13]. Therefore, understanding adherence of older adults to ART is paramount to develop strong HIV treatment programs.

Optimal adherence has both individual health benefits as well as wider public health benefits such as prevention of HIV endemic and improvement of mental health [14, 15]. Whereas poor adherence leads to incomplete viral suppression allowing disease progression which ultimately increases mortality [16, 17]. In 2014, UNAIDS called for the ambitious 90-90-90 targets with the aim of ending the global AIDS epidemic by improving surveillance, ART adherence and viral suppression in infected individual [18]. The program was aimed to achieve its goals by 2020 and one of the targets was to have 90% of all people receiving anti-retroviral treatment (ART) achieve viral suppression by 2020. This target can only be achieved if there is a strong emphasis on ART adherence. Despite the growing recognition that older people are increasingly infected with HIV [6, 19] and are accessing treatment, the literature on ART adherence among older adults is limited [20, 21]. This is particularly true in SSA where the burden of HIV among older adults is highest, and treatment and adherence challenges are greatest due to overburdened health services that are not set up to address older peoples’ needs.

A recent review has shown that the adherence to ART among older adults is higher than that of younger adults in developed countries [7]. There are reasons to believe this in the context developed countries, where older adults are likely to be financially stable and medically aware about their health. However, older adults in SSA might experience different barriers to adherence when compared to younger adults. These might include high rates of poverty, limiting money for transport and care [22, 23]; multi-morbidity with various chronic diseases leading to poly-pharmacy [24]; social isolation [25]; ageism or elder abuse [22]; and little understanding among health providers of older persons’ needs [22, 24]. Another important determinant of adherence in elderly is neurocognitive impairment. It has been shown that decline in executive functioning, motor functioning, and processing speed leads to suboptimal adherence in older adults [26]. Moreover, in high morbidity/mortality and high poverty areas, older adults often lose their caregivers and must provide care to others while suffering from age-related illness and frailty [27, 28]. As a result they may feel lonely, disconnected and despondent eventually becoming hopeless and demotivated to seek treatment [29, 30]. Depression is a surrogate of this sequel which has been shown to be a predictor for non-adherence (OR 2.54; 95% CI 1.65–3.91) [31].

A critical first step to improving adherence among older adults in Africa and thus helping to achieve the 90-90-90 targets is to understand current levels of treatment adherence. Currently there is no synthesis in the literature on the comparison of adherence rates for ART between older and younger adults living in Africa. Therefore, we conducted a systematic review and meta-analysis to examine adherence to ART in Africa among older adults compared to younger adults.

Our systematic review shows that ART adherence among older people is similar to that of younger people in SSA. Despite a general neglect of programs for older adults living with HIV in Africa and despite issues of poverty, social exclusion, lack of support and mobility challenges, adherence is not statistically different to that of younger people. A 2013 meta-analysis by Ghidei and colleagues identified 12 studies from non-African settings that contained data on older adult adherence compared to younger counterparts [7]. They found that older age reduced the risk of non-adherence by 27% in the studies from the US, Italy, United Kingdom, Canada and Brazil. Our study found no such difference in their adherence. However, this review included studies that used a variety of different measures of adherence each of which have their own strengths and weaknesses [40]. For example, self-report is a practical tool in low-resource settings, but can overestimate adherence [41]. Pharmacy refill adherence measures such as pill counts and calculation of the medication possession ratio can be impractical and do not measure whether patients are actually taking their medication, although are less prone to reporting bias than self-report [42, 43]. Clinical contact measures require a clinician to classify adherence as good or poor, and thus subject to assessment bias and difficult to standardize [43]. All three included measures have been demonstrated to be significantly predictive of detectable viremia [44].

We found that when MPR or clinician counts were used as the measure of adherence, older adults were more adherent than of younger adults. The reasons for this is unclear; however, it could be argued that pharmacy refill-based adherence monitoring tools (e.g. MPR) are better adherence measures than self-report [45]. It is well accepted that there is no gold standard to measure ART adherence [46]. Yet a number of available adherence methods such as self-reports, pill counts, pharmacy records, self-report electronic monitoring, and therapeutic drug levels can be used depending on the setting or the available resources. Self-reports and pharmacy refill are the most common approaches used in Africa due to the low-cost and simplicity. Common limitations include poor sensitivity to detect non-adherence (e.g. social desirability bias for self-report or pill dumping for pill count) but good specificity for pharmacy refills [42]. Therefore, pharmacy refills may accurately capture older persons’ adherence behaviour, making it a better measure for this population.

Another interesting finding in our analysis was when two studies with the largest weights were discounted from the pooled analysis, younger adults were significantly more adherent than older adults p = 0.03. The key differentiating characteristic of these two studies was that their ‘length of follow up’ was greater than the other studies i.e. 24 and 60 months respectively. This may mean that longitudinally older adults may be more likely to be adherent to an ART program. Adherence in older adults is complex and the next challenge is to better understand the individual, interpersonal, community and structural factors behind adherence, and most importantly, the levers by which adherence could be improved. For example, Shubber and colleagues’ in a recent systematic review on patient-reported barriers identified a number of obstacles including depression, stigma and health-service related barriers such as distance from facility [47]. Cultural norms and personal beliefs have been acknowledged as key predictors for medication adherence in elderly [48]. In the indigenous culture illness is sometimes considered as fate, or a condition whose control is held by an external locus [49, 50]. These believes combined with the likelihood that older adults living in Africa are less likely to be exposed to contemporary western medicine and more likely to have deeper roots in their tradition, when compared to their younger counterparts, makes adherence a challenge. A solution to that can be the inclusion of patient centred pharmaceutical counselling, this may be an important tool to improve their adherence. However, the evidence for effectiveness of pharmaceutical counselling is from developed countries [51, 52], and may not be easily translatable in the context of low-resource countries in Africa. In these settings, the interaction between physical and mental health becomes particularly important [30], and future studies should have a special focus on this aspect.

In developed countries older adults show less attrition on ART [53], and the CD4 reconstitution is low [54‐57], thus increasing the importance of early access and strong adherence to reducing mortality. Past studies have explored the impact of depression on adherence in Africa [58] as well as exploring the impact of community-based adherence interventions [59‐61]. A number of peer-led interventions, such as medical companions or support groups for people living with HIV, have proven successful in improving ART access and adherence in SSA [62, 63] but nearly all exclusively target people under 50 years of age [64]. Therefore, prospective programs need to have a focus on older adults living in low-resource or developing countries.

Understanding reasons for late presentation and mechanisms for addressing that shortcoming would be critical to improving outcomes among older people. Overall, however, there have been very limited HIV-related behavioral interventions among older people [65] which has limited the development of appropriately and specifically targeted programs. In order to achieve global HIV goals, the health providers must strive to improve the quality of care and quality of outcomes among all people living with HIV. While older adults have long been neglected, there are promising signs that their specific needs are now being recognised and that appropriate programs will emerge to ensure that older people remain productive members of their communities.

This systematic review and meta- analysis found ART adherence levels were similar among older and younger adults in Africa. Further research might identify specific barriers to adherence in the HIV affected populations, and targeted interventions to improve clinical outcomes and quality of life when those affected by HIV grow older.