Antiretroviral therapy (ART) adherence levels of ≥ 95% optimize outcomes and minimize HIV drug resistance and to optimize measures of patient outcomes . Previous studies in Ethiopia were using only self reported dose adherence as a measurement [23–25]. In our study we also used the time restriction (time adherence) and instructions related to food (food adherence) in addition to self reported dose adherence measurement. Our data suggest that adherence rates among patients in southwest Ethiopia were higher than adherence rates in most developed countries. In this study measuring adherence by patient self-report, 95% of the patients were adherent with ≥ 95% of prescribed doses in the last 7 days. Other studies conducted in developed countries demonstrated that the rates of adherence by self-report ranged from 40% to 70% [26–28]. Even in Botswana, fifty-four percent of patients in the study were adherent by self-report with 95% of prescribed doses . Other studies in developing countries have shown comparable or better levels of individual adherence than what is seen in North American and European populations [29, 30]. According to a prospective study in Southwest Ethiopia, 384 (96%) and 361(94.3%) of the study subjects were adherent based on self-report of missed doses (dose adherence) in a one-week recall at base line (M0) and follow up visit (M3) respectively. Three hundred eighty nine (97.2%) and 373 (97.4%) of the study subjects always followed the time restrictions (time adherence) agreed upon with their providers at M0 and M3 respectively. Three hundred thirty eight (84.5%) and 319 (83.3%) subjects followed instructions related to food (food adherence) all the time. Hence, the rate of self reported adherence in the study area based on the combined indicator of the three adherence errors was 79.3% at baseline and 75.7% at follow up visit . Similarly, two studies in Ethiopia reported 81.2% and 82.8% adherence to more than 95% of doses [23, 25]. This high rate of adherence showed adherence to ART in resource limited country can achieve a high level of adherence than those developed country. The overall rate of self reported adherence in this study based on the combined indicators of the three adherence errors was 72.4%. Similarly, consistent finding has been documented in similar set up . Some studies in resource-rich settings have documented less than 50% of patients taking all their antiretroviral medications on time and according to dietary instructions [31, 32]. Bonolo et al. review 43 articles on adherence to HAART. They found a mean rate of non-adherence of 30.4%, range from 5% to 67% . This was much lower than our report confirming that patients in developing countries can achieve good adherence despite limited resources. The possible explanation for the greater adherence in our study might be the majority of the participants started ART recently, the participants were given strict adherence counseling sessions before starting ART in the hospital.
Non-adherence takes the form of skipping a dose. In a study of southwest Ethiopia, they found principal reasons reported for skipping doses were most 38 (43.7%) simply forget, 17 (19.5%) felt sick or ill at that time, and 11 (12.6%) ran out of medication at baseline. During the follow up visit again the majority 14 (65.6%) simply forgot, 4 (19%) felt sick and 4 (18%) were busy . In our study the reasons given for missing drugs were running out of medication/drug 9(27.3%), being away from home 7(21.2%) and being busy with other things 7(21.2%) and the rest reasons included simply forgetting, having no food to take with the medication, fear of side effect and feeling sick or ill at that time. Forty-eight percent of patients asserted that they missed their doses due to finances, while 24% listed forgetting as a primary reason for treatment non-adherence. Other barriers to treatment included running out of medications (17%), travel/migration (13%), side effects (12%), and being too busy (12%) . Forty-one percent of subjects (71/173) stated they never missed a dose of ARV. The 102 patients reporting missed doses at baseline did so for a variety of reasons, the most common of which was 'forgetting' to take the medication (41%; 42/102). Other reasons included being away from home (9%), being busy with other activities (6%), and taste perversion (5%), or concern about toxicity (4%). Less commonly listed reasons (2%) included running out of ARV medications or anxiety related to the constant reminder of their HIV infection . Study subjects most commonly reported that they missed antiretroviral doses because they were busy or forgot, away from home, or experienced a break in their daily routine. Smaller proportions reported missing doses because they felt depressed or overwhelmed, were taking intentional drug holidays, or had run out of medication . This implicate that the reason for skipping a dose should be given due emphasis from clinical, dispensing visit as well as during ongoing adherence counseling, and follow up visit. Other interventions aimed at maintaining adherence, and thereby optimizing the benefit of effective therapies should be sought in detail by health care workers.
There is good reason to expect that sociodemographic, psychosocial, and clinical variables should be associated with antiretroviral adherence and thus HIV disease activity . In this study patients with average family income of middle and highest were more likely to have an overall adherence than the lowest average family income in bivariate analysis. The most common patient-related barriers were financial constraints [29, 35]. Among patients having the economic ability to receive their medication, there was an association between the annual income and adherence [36, 37]. Findings have also been inconsistent in defining the relationship of lower income [6, 8, 37, 38] to adherence. A monthly middle income was significantly associated with greater pharmacy adherence. Low or high incomes groups showed a higher risk for pharmacy non-adherence/economic status, in particular patients with the highest monthly income when compared with monthly middle income, was retained as a predictor of poor adherence only in the best case scenario . A recently published meta-analysis  examined the association between socio-economic status and adherence to antiretroviral therapy: out of 8 studies, only 2 prospective studies identified low income as a predictor of non-adherence. Other factors might be contributed for the difference between income and adherence like educational status. Other study also demonstrated that social support has a paramount important for adherence uptake. In our study patients who got family support were 2 times more likely to adhere than those who didn't get the family support. Another factor facilitated adherence was support from the family encouraging and helping to remind them to take the treatment. Social support, such as someone to help with the tasks of starting to rebuild a life, assistance with cooking and assistance to grow crops, all encouraged adherence . Similarly, it has been reported in other studies  as social support was a constant predictor of adherence identified at baseline and follow up visit, living in a couple could improve adherence because it increases the routinization of daily behaviors and activities (Wagner & Ryan, 2004)  and better social supports for using medications were all associated with better adherence . However, a recent meta-analysis of studies across multiple medical conditions determined that adherence was more strongly and consistently associated with functional support (i.e., practical/emotional support) than structural support (i.e., living arrangement/relationship status; DiMatteo, 2004) . Within the domain of functional support, the study found that the provision of practical support had a significantly greater influence on adherence than emotional support . Lacks of social support have been found to be associated with lower adherence [6, 26]. Social support  was associated with greater adherence. Lack of support has been associated with an increase in suboptimal adherence [45, 46]. Murphy and colleagues reported that those with greater social support for example having reassurance from family members, those having reliable alliances were more likely to be adherent over the past one month . This highlights that social support assist in reminding to take the drugs according to the prescribed schedule and time, hence, for adherence. So it is better to advise/counsel our patients on initiation and continuation of HAART to be effective.
In our study disease stage/progression had been associated with adherence. Those participants who were in stage I were 74% less likely to adhere than those who are in the stage IV. Similar finding has been documented in other studies. In Chinese study, symptomatic disease stage had more likely to become adhere than asymptomatic disease stage . Other factors significantly associated with viral suppression were less severe disease (WHO stage II or III vs WHO stage IV) . Inconsistence to our finding in Cameroon, CDC stage B patients and specially CDC stage C patients had higher risk of pharmacy non-adherence than asymptomatic patients. When compared with asymptomatic patients, the multivariate analysis confirmed a marked risk of non-adherence for CDC stage B patients and CDC stage C patients in the worst-case scenario in Cameroon. However, HIV CDC clinical stage at the beginning of treatment significantly predicted loss to follow-up: compared with asymptomatic patients CDC stage A, CDC stage B patients and specially CDC stage C patients had greater rates of loss to follow-up .The possible reason might be those patients in stage I were not that much manifest the diseases/symptomatic and might feel that they are health looking as well not concerned about their illness as compared to those in advanced stage.
The findings of this study should be interpreted with some limitations. Because it was conducted at a single site, the findings may not be generalizable to dissimilar clinical settings. Recall bias and social desirability bias are also the possible bias which may encounter in this study. There is no gold standard for measuring adherence and our measurement of adherence is only based on patients' declarations of missed doses, scheduling instructions and dietary requirements. Despite the above limitations, the study addressed an important issue in developing country, and inclusion of several variables that predict adherence and to fully characterize the study population, we include other dimension of adherence measurement for successful treatment with ART (adhering to scheduling and to dietary instructions), reasonably large sample size (N = 319) and had a high participation rate.