Introduction

Over 15 million HIV-infected individuals were on antiretroviral therapy (ART) by March 2015 [1]. However, adherence to treatment remains a key challenge for HIV programs to achieve optimal health outcomes and viral suppression [2, 3]. The World Health Organization (WHO) defines treatment adherence as “the extent to which a person’s behavior-taking medications, following a diet and/or executing lifestyle changes corresponds with agreed recommendations from a health care provider” [4].

ARV adverse drug reactions are an important factor that influences adherence to ARV [5]. WHO defines an adverse drug reaction as “a response which is noxious and unintended, and which occurs at doses normally used in humans for the prophylaxis, diagnosis, or therapy of disease, or for modification of physiological function” [6]. Serious and/or long-term adverse drug reactions can affect adherence to the treatment regimen [7]. There is limited knowledge of how HIV-infected individuals perceive and experience adverse drug reactions to ARVs and how these perceptions and experiences may influence ARV adherence.

Qualitative synthesis focuses on in-depth interpretive explanation of a phenomenon and therefore enables deeper analysis of the complex, multi-faceted experiences of adverse drug reactions and its relation to ARV adherence. The aim of this qualitative synthesis was to systematically review and synthesize the qualitative literature examining how perception and experience of ARV adverse drug reactions influence drug adherence among HIV-infected individuals.

Methods

Search Strategy and Selection Criteria

In accordance with guidance from PRISMA checklist [8], ENTREQ [9], and meta-synthesis guidance from the Cochrane group [10], we used a comprehensive search strategy to identify all relevant studies in English (Search strategy available in supplement). The review protocol (CRD42015017265) was registered (www.crd.york.ac.uk/PROSPERO/). The following electronic journal and dissertation/thesis databases were searched from January 1st, 2000 until February 21st 2015 (to limit to recent ARV drug regimens): CENTRAL (Cochrane Central Register of Controlled Trials), EMBASE, LILACS, PsycINFO, PubMed (MEDLINE), Web of Science/Web of Social Science, CINAHL, British Nursing Index and Archive, Social science citation index, AMED (Allied and Complementary Medicine Database), DAI (Dissertation Abstracts International), EPPI-Centre (Evidence for Policy and Practice Information and Coordinating Centre), ESRC (Economic and Social Research Council), Global Health (EBSCO), Anthrosource, and JSTOR. Conference proceedings including the Conferences on Retroviruses and Opportunistic Infections (CROI), International AIDS Conference (IAC), and alternating year International AIDS Society (IAS) clinical meetings were searched from their inception dates (1993, 1985 and 2001, respectively). References of included studies were checked and authors were contacted to provide additional information as required.

Studies were selected that used qualitative research designs, including but not limited to ethnographic research, case studies, process evaluations, and mixed methods research. Studies also had to include qualitative data collection techniques (e.g., participant observation, in-depth interview, or focus group) and a qualitative data analysis approach (e.g., framework analysis, or thematic analysis). The review included studies that provided description and interpretation of the impact of adverse drug reactions on adherence for all HIV-infected individuals. We excluded studies that only used quantitative methods to investigate the same phenomenon.

Data Extraction

A health sciences librarian conducted the electronic searches and excluded duplicates. All titles were reviewed to examine relevance to the meta-synthesis topic. Two authors (HL & GM) independently reviewed abstracts and made a list of categories for exclusion, and then independently reviewed full text manuscripts and noted reasons for exclusion. Discordance between the two authors was resolved by a discussion with a third author (QM). Studies were reviewed for relevance based on adverse drug reaction’s impact on adherence, study design, main findings and interpretations of findings. (Fig. 1).

Fig. 1
figure 1

Prisma Flowchart

Full size image

A data extraction form was designed specifically for this review. The characteristics extracted from each study included authors, year published, study design, location/orientation of study, key populations and subgroups, age range, main findings, ARV drug regimen, and adverse drug reactions described.

Data Analysis

Prior to conducting the synthesis, each paper’s quality was assessed by both HL and GM using a seven scale criteria tool adapted from the critical appraisal skills programme (CASP) [11], which has been used in other reviews [12]. The CERQual approach was used to evaluate the certainty of the review finding [13]. This approach includes an assessment of methodological limitations, coherence, relevance, and adequacy (see the supplement for details), based on a guidance document from the Cochrane Collaboration Qualitative Methods Group [10, 13]. Thematic synthesis was adopted in the data analysis. A line-by-line coding of the findings of primary studies was conducted by HL and GM prior to organizing into related areas to construct descriptive themes. Analytical themes were developed based on the descriptive themes [14]. Pre-defined subgroup analyses were conducted among specific groups of adults, children and adolescents, and pregnant women. EFV was specifically chosen in the analysis because it has known adverse drug reactions and yet remains first-line therapy in WHO ARV guidelines.

Role of Funding Source

The WHO was a funding source for this study. The authors had full access to all the data in the study and had final responsibility for the decision to submit for publication. The content of the article is solely the responsibility of the authors and does not necessarily represent the official views of WHO or the National Institutes of Health.

Results

Thirty-nine studies published after 2000 were included in the review. In the CASP assessment, most of the studies reached good quality with the highest score of 7 (17 studies) or 6 (14 studies), while some studies had medium score of 5 (4 studies) and low score of 4 or less (4 studies). Thirty-three out of 39 studies reported that perceived adverse ARV reactions decreased adherence [12, 15–46], whilst six studies [47–52] described no apparent relationship. Nineteen studies [12, 15–27, 29, 32, 36, 44, 47] were conducted in high-income countries, eight [30, 31, 33, 34, 41, 42, 46, 49] in middle-income countries, and twelve [28, 35, 37–40, 43, 45, 48, 50–52] in low-income countries. (Countries—see Tables 1, 2).

Table 1 Summary of extracted data from selected full texts
Full size table
Table 2 Summary of study characteristics
Full size table

The majority of the studies (34/39) sought the perspectives of HIV-infected adults. Two studies considered the views and experiences of children (aged 3–12) and their caregivers [24, 46] and one study of adolescents [45]. Two studies focused on pregnant women [16, 36]. Three studies explored experiences of coping mechanisms for adverse drug reactions as a way of improving adherence [31, 38, 45]. Study characteristics have been summarized in Tables 1 and 2. Eight themes emerged in the thematic synthesis, which included three barriers (i.e., visually noticeable adverse drug reactions, psychological suffering, and Efavirenz (EFV)-containing regimens), three facilitators (i.e., prior knowledge of adverse ARV reactions, coping strategies, and self- perceived effectiveness of ARV), and subgroup analyses among populations of children, adolescents, and pregnant women.

Barriers

Visible Adverse Drug Reactions

Visible adverse drug reactions (e.g., body changes, a buffalo hump, excess sweating, darkening of the skin, body odor, hair loss, weight loss/gain, and skin rash) increased the risk of unintentional HIV disclosure and disturbed HIV-infected individuals’ daily routines. It led to low self-esteem and self-stigmatization among some HIV-infected individuals, and was felt to contribute to poor ARV adherence [15, 19, 21, 23, 25, 28, 29, 32, 35, 38–40, 42, 45, 49, 50]. In several persistent cases, ARV adverse drug reactions caused HIV-infected individuals to undertake long treatment free vacations or quit treatment completely [15, 21, 23, 32, 49]. This finding has a CERQual high level of confidence.

For many participants on ARVs, adverse drug reactions were perceived to be detrimental to adherence [29]. Primary data from the reviewed articles include the following: “I have asked my doctor not to put me on any medication that would make me look like a freak. I would simply not take the drug if I developed a buffalo hump on my back” [23]. “I then had a skin rash to a point where I was so uncomfortable and sick that I stopped taking the medication” [21].

Psychological Suffering

Psychological suffering associated with adverse drug reactions (e.g., being reminded of being sick, delusional, a feeling of being killed by taking the drugs, and getting mad) were occasionally considered to be severe enough to contribute to poor adherence [17, 19, 20, 37, 40, 47, 49]. This finding has a CERQual high level of confidence.

Primary data showed that some participants stopped taking ARVs because they realized that they were not psychologically prepared to take ARVs. For example, participants described: “Just the fact that I have to take it [medication(s)] and knowing that I’m going to have to take it for the rest of my life is having psychological effects, it’s causing trauma to my life right now” [19]. Participants generally used adjectives such as depressing, cranky, moody, sad, angry, guilty, irresponsible, nervous, anxious and afraid to describe their negative psychological problems related to ARVs [20].

Adverse Reactions Associated with EFV-Containing Regimens

Some people receiving EFV-containing regimens reported that neuropsychiatric adverse reactions (e.g., intense body heat, delusions, anxiety, intense dizziness, and nightmares) contributed to poor adherence [47–52]. This finding has a CERQual moderate level of confidence.

Participants described their pain and suffering related to EFV-associated adverse drug reactions: “The numbness crept up from my feet and spread out over my body. Just like with an electrical shock, the feet feel it the worst. During the night, my feet are so sore, numb, and swollen that I can hardly sleep. The pain is so constant. I can hardly close my eyes” [49]. “I can’t say that they have worked well because as you see now my face it has changed from its normal shape, I don’t know whether it is a type of drug or it is because I have spent some time on them” [50].

Facilitators

Prior Knowledge of Adverse ARV Reactions

When people receiving ART had prior knowledge of ARV adverse drug reactions, or had been informed in advance of potential adverse reactions and how to manage them, they felt these reactions were expected, normal and manageable, and therefore reported less non-adherence [35, 38, 48, 50]. This finding has a CERQual moderate level of confidence.

Participants described their experiences: “We were warned that some people would get drug reactions. This helped me not to worry when I fell sick” [38]. “That’s normal stuff” [48]. “Those things occur, and after a time they go away” [48]. For those participants who were informed about potential adverse effects of ARVs prior to treatment initiation, the reactions were not reported as reasons for non-adherence [48]. Participants wished they had been pre-informed about the adverse effect profile at the initiation of treatment, so that they know how to manage these effects [35].

Coping Strategies

When people receiving ART developed coping strategies to cope with adverse events (e.g., drinking a lot of fluids and resting to reduce dizziness, or eating a snack or meal before swallowing pills to control nausea), they expressed good levels of adherence [31, 38, 45]. This finding has a CERQual moderate level of confidence.

Participants described how they dealt with adverse drug reactions: “I have abdominal distention. I felt so uncomfortable even if I only ate a little. I felt that the food was stuck there and was not digested. Taking some herb digestion medicine makes me feel better” [31]. A number of strategies were also employed to disguise visible adverse effects. For example, women reported wearing dark nail polish to cover darkened nails and wearing long sleeved clothes, ankle reaching skirts and trousers to cover body sores [38].

Perceived Effectiveness of ARVs

When people on ART experienced individual benefits of taking ARVs on their health (e.g., regaining physical strength, being more independent, and able to help with domestic work again), they achieved greater adherence independent of the adverse effects [30, 31, 45, 48]. This finding has a CERQual moderate level of confidence.

Participants described benefits of ARVs as a reason to adhere to medication as follows: “I say ARVs are medicine [e.g., effective medicine, not fake]. Because basing it off my health, I had lots of sores all over my body, on my arms, in my mouth, on my head, on my private parts, you understand? …After being started on the drugs, all those sores went away, my health status started to improve, and I was able to eat well. Until now after coming to test recently I had a CD4 of like 181. So it has risen well. It’s these meds [that do it], man” [48]. “I cannot live without these meds. If I did not take them, I might die very soon. The benefit of the meds is they enable me to survive a few more days. As patients, our lives have been shortened by HIV, but ART can extend our lives a little bit longer. Taking meds on time can prolong my life” [31].

Subgroup Analysis

Subgroup Analyses Among HIV-Infected Children and Adolescents

The experience of drug attributed adverse events (e.g., queasy feeling, vomiting, diarrhea, nausea, insatiable hunger, and skin sores) resulted in children’s refusal to take ARVs (e.g., coughing, gagging, and throwing it up) [24, 45, 46]. This finding has a CERQual moderate level of confidence.

For example, an adolescent said, “There is this one [pill] at night that makes you feel dizzy, that one bothersome a lot. I keep taking them, and then sometimes they […] provoke these scabs. Those others provoke wounds […], they damage your face. So these ARVs […] and my body don’t get along” [45].

Subgroup Analyses Among HIV-Infected Pregnant Women

Pregnant women perceived that the likelihood of their unborn baby suffering from ARV adverse effects was greater than the likelihood of their unborn baby acquiring HIV infection [36]. ARVs also worsened their pregnancy-associated nausea, hence they felt that ARVs did more harm than good [16, 36]. This finding has a CERQual low level of confidence.

For example, a woman described: “When you are pregnant, the thing about taking zidovudine (AZT) is that, it is so strong. Taking AZT is like taking drugs, it’s like taking marijuana or smoking crack…it’s like the baby is gonna be born with that in his system” [16].

Discussion

This synthesis identified 39 qualitative studies from diverse countries. Most of the studies supported that occurrence of ARV adverse drug reactions are associated with ARV non-adherence. This is consistent with quantitative research demonstrating that adverse drug reactions contribute to poor adherence with treatment interruptions [53–56]. This synthesis formally evaluated the qualitative evidence and extended the literature in three ways: (1) proposing barriers that likely contribute to non-adherence (2) exploring facilitators to improve adherence, and (3) uncovering challenges among specific groups.

This qualitative synthesis explored potential neuropsychological mechanisms of subjective perceptions and experiences influencing ARV adherence. This is consistent with cross-sectional data correlating subjective perceptions of ARVs and adherence [55, 56]. Cross-sectional and cohort studies showed that depression, anxiety, negative cognitive and psychosocial functioning were frequently associated with poorer ARV adherence [57, 58]. This synthesis suggested that the experiences of taking ARVs were associated with significant psychological trauma and emotional suffering in some HIV-infected individuals. In particular, the trauma and constraints were usually related to being on a lifelong therapy associated with HIV infection. Also, visually noticeable adverse drug reactions could disturb daily routines and threaten HIV disclosure and embarrassment in public, which might in turn increase psychological and emotional burden. This suffering may further exacerbate self-stigmatization and further decrease the self-esteem of HIV-infected individuals. When this suffering is not effectively addressed, it may contribute to ARV non-adherence.

Even though some ARV drugs (e.g., stavudine) has been recommended to be phased out [59]. This qualitative synthesis highlighted this concern and provided descriptions of adverse drug reactions related to EFV-containing regimens contributing to poor adherence. This is consistent with several quantitative meta-analysis demonstrating the association between EFV use and central nervous system (CNS) toxicity [60] and treatment non-adherence [61]. Shubber and colleagues’ systematic review described that around a third of people using EFV experienced CNS adverse reactions (e.g., dizziness, abnormal dreams) with variable degree of severity, which are usually mild and transient [62]. However, Shubber and colleagues’ review showed that compared to nevirapine, EFV was associated with a lower frequency of severe adverse reactions, particularly in treatment discontinuation among both adult and children populations [62]. A recent review [63] on comparative safety of EFV and impact of CNS adverse events showed 90 % of patients remaining on EFV (78 weeks). Even though the relative risk of treatment discontinuation due to EFV-related adverse events was higher than that related to other 1st line options (e.g., tenofovir disoproxil fumarate, abacavir, dolutegravir, raltegravir), the absolute differences were low (<5 %) [63]. Although the overall frequency of EFV adverse medication reactions remains rare, our findings suggest that EFV-containing regimens are perceived by PLWH as related to adverse reactions influencing their lives, appearance, and well-being.

Another important finding of this qualitative synthesis is that the HIV-infected individuals’ experience, perception, understanding, and knowledge of adverse drug reactions were related to ARV adherence. When people receiving ART experienced ARVs maintaining their health effectively, had knowledge of adverse ARV reactions, and practiced coping strategies to manage these reactions, they were more likely on adherence. This is consistent with previous quantitative studies that having knowledge about adverse drug reactions, adherence self-efficacy, self-management, and individual strategies for fitting medication into daily routines were associated with ARV adherence [64, 65]. However, researchers also argued that extensive communicating about potential adverse medication reactions may increase the likelihood of a patient developing those reactions [66]. A simplified and structured informing approach about potential adverse drug reactions to ARV is needed.

This qualitative synthesis found very limited studies considering perspectives of children, adolescent, and pregnant women, and the findings were in moderate and low confidence levels respectively. For children and adolescents, refusal of ARVs or doses skipping were related to pediatric adherence barriers, such as feeling sick and needing a break [67], palatability of medication, large medication volumes, and limited treatment options [68, 69]. For pregnant women and women with history of childbirth, some important issues were emerging that they perceived ARVs as medically powerful and full of potential adverse effects, which made these women felt that ARVs did more harm than good. For this specific group of women, ARV adherence became more complicated.

This qualitative synthesis is subject to some limitations. First, some of the reviewed articles are weak in research design or have incomplete reporting. However, we evaluated the quality of individual studies and this quality assessment was reflected in the overall confidence assessment using CERQual. Second, most of the reviewed articles emphasized adults with HIV and few specifically focused on children, adolescents or pregnant women. It is therefore difficult to generalize the findings to other specific sub populations. Third, our qualitative synthesis only provided subjective interpretation of the relationship between patient attributed adverse drug reactions and ARV non-adherence and causal pathways could only be inferred.

Despite these limitations, this synthesis found that visually noticeable drug reactions, emotional and psychological suffering, and EFV-containing regimens may be related to non-adherence for HIV-infected individuals. This synthesis also found that improved knowledge and understanding of adverse ARV reactions, experiencing the health benefits of ARVs, and strategies of coping and self-management were practically important in addressing the problem of ARV non-adherence caused by adverse drug reactions. It’s therefore important to provide HIV-infected individuals tailored counselling to enhance knowledge and understanding of adverse ARV reactions and to train and support people receiving ART how to adopt multiple strategies to deal with these reactions. Moreover, this synthesis provided directions for future studies. Vulnerable children, adolescents, and pregnant women are still far under studied in the qualitative literature on their experience and perception related to adherence of ARVs. Future qualitative studies should focus on exploring subjective experiences and perceptions of ARV adverse drug reactions and strategies being used to manage these reactions from the perspective of children, adolescents, and pregnant women.