Introduction

The advent of antiretroviral treatment (ART) has dramatically slowed down the progression of HIV, reduced the death rate from AIDS and transformed the infection from a fatal illness to a more manageable chronic illness [1], [2]. Since 2004, Nepal has been providing free-of-cost ART and by the end of 2009, over 2,524 adults received free ART at 23 sites across the country out of an estimated over 63,528 People Living with HIV (PLHIV) nationally [3]. The success of a national scale-up of ART depends on bolstering the capacity of the health care system and shifting its orientation from acute care to a chronic-care model [4], [5].

However, simply making ART medicine available to PLHIV is not enough, as strict adherence is required for treatment success [2], [6]. Poor adherence can lead to the virological failure of cheap first-line treatment regimens and the spread of multi-drug resistant forms of the virus, resulting in a public health calamity [2], [7], [8]. Unlike many other diseases, it is vital that PLHIV consume all doses of the drug to prevent resistance and to improve their chances of survival. Understanding the level of non-adherence and the factors that lead to it are important clinical and public health goals. This information is essential to inform ART programmes and maximise the success of treatment.

Paterson and colleagues found that adherence greater than 95% is needed to achieve virological success; however 22% of patients with an adherence level of over 95% experienced virological failure (i.e. a sharp increase in viral load) compared to 61% of patients with adherence between 80–94.9%, and 80% of patients with an adherence level of below 80% [2]. A meta-analysis of studies of ART adherence found that a pooled estimate of 77% of patients in Africa achieved adequate adherence (>95% of prescribed pills) compared to just 55% of patients in North America [9]. However, the relationship between adherence and the development of resistance differs by regimens; for example resistance to non-nucleoside reverse transcriptase inhibitors is significantly higher at low levels of adherence than that for protease inhibitors [10].

Prescribers hope that every patient completely follows their ART instructions, but the literature shows that a proportion of PLHIV do not take medications as prescribed for various reasons. A patient’s ability to adhere to medication is greatly influenced by both individual and environmental factors. Several studies have shed light on the factors affecting adherence, highlighting socio-demographic, cultural, economic, health-systems and treatment-related factors [9], [11], [12]. Many barriers to adherence are common to both developed and developing countries such as fear of disclosure [9]. However, some are more common in the Asian developing countries such as ART-associated costs (transport fares, diagnostic costs) and problems with travel to access treatment [13]. Hence, to benefit from ART, it is important to identify adherence behaviour, understand the conditions that lead to non-adherence and develop strategies and social policies to maximise long-term adherence. This study was designed to identify the current levels of adherence and the factors influencing adherence to ART in Nepal, as to our knowledge, there is no prior study of this kind in Nepal.

Methods

A cross-sectional mixed-methods study was conducted in late 2009. It included a quantitative survey with 330 ART prescribed patients from ten ART sites across Nepal and qualitative in-depth interviews with 34 purposively selected participants, namely 17 ART-prescribed patients, 14 ART service providers (including doctors, nurses, paramedics), and three policy-makers.

For the survey eligible respondents were taken from a daily hospital visit schedule using a systematic random sampling technique [14]. Adherence was assessed by using a self-reported Adult AIDS Clinical Trial Groups (AACTG) adherence instrument [15], originally developed in the United States and now widely adopted in many countries [16], [17]. In addition to the AACTG questionnaire, a further structured questionnaire was developed from the literature. The dependent variable was ‘adherence to ART’, independent variables were socio-demographic and cultural, health condition, and service facility, knowledge, and perception of ART treatment related variables. Patients were considered adherent when they took 100% (not missing a single dose of ART) based on self-report in a four-week recall prior to the study; otherwise, they were categorized as non-adherent. If they reported having missed or skipped a dose during that time, the questionnaire asked a range of multiple-choice questions about why they had missed their pills.

Logistic regression analysis was performed to examine the effects of explanatory ( = independent) variables on adherence. All variables, both explanatory and dependent, were coded as binary variables prior to fitting in the model; for adherence this meant that 0 = non-adherent and 1 =  adherent. Initially, explanatory variables were included in the model one at a time to examine their univariate relationship with adherence. As many variables were analysed, only those that were significant are presented. Following on from this univariate analysis, multiple logistic regressions was used to model the effects of many variables simultaneously. Variables were fitted in the model in blocks: 1) socio-demographic and cultural; 2) health condition; 3) health-service facility; 4) knowledge; and 5) perception related variables. At each stage, the least significant variable was excluded until the model contained only statistically significant factors. Statistical analyses were conducted using the Statistical Package for the Social Science (SPSS) version 17.0 and a p-value of less than 0.05 was used to define statistical significance.

The in-depth interviews were conducted with the aid of an interview checklist and probes for further questioning [14], covering questions on how ART fitted into their daily routines and what key factors were impeding their adherence to their medication. The interviews were conducted and audiotaped in Nepali by the first author who is a native speaker; sessions lasted 1–1½ hours. Audio recordings were transcribed verbatim in Nepali and then translated into English. Data were organized using NVivo version 8 [18] and coded using a thematic analysis [19]. The quotations included below best represented the range of ideas voiced around key themes; to maintain anonymity these quotes are identified by gender and region only. For the same reason, nurses and paramedics are all coded as ‘counsellor’. To assure quality of the data, the main survey questionnaire was piloted [20] with 15 selected ART prescribed patients and the semi-structured questionnaire (checklist) was piloted with two patients on ART and a care provider; subsequently some modifications were incorporated in the research instruments. Ethical permission for the study was obtained from the Nepal Health Research Council. Individual consent was sought before interviews, often in the form of oral consent as about 42% of the population in Nepal is illiterate [21].

Results

Quantitative Results

In total 330 ART prescribed patients were approached and 100% responded to the main questionnaire; a total of 57.0% were male and the mean age was 35.8 (SD 8.30) years (range 18 to 62). The majority (70%) of the respondents were married, 38.5% had never attended school, 83.0% were unemployed, and 25.2% of the respondents travelled more than three hours to the hospital. The majority (68.5%) had disclosed their status to someone other than the health care worker. More than a quarter (30.9%) faced some form of stigma and discrimination because of their HIV positive status. Patients had been taking ART for a median of 24 months (range 4 to 48). The first line treatment regimens based on 2NRTIs+1NNRTI (Non/Nucleoside Reverse Transcriptase Inhibitor) were used by almost all patients (99.4%). Interestingly, none of the respondents were taking any traditional medicine with ART for HIV treatment. All respondents identified the trade names, dosage, and frequency of pills and all understood the notion of adherence. The majority of the respondents had positive perceptions of ART treatment and maintaining adherence; for example, 95.2% perceived that even after feeling better, patients should not stop taking ART medication.

Adherence in this study is encouraging with 282 (85.5%) respondents reporting they had not missed a single dose in the previous four weeks. When asked why they had missed their pills, the most commonly reported reasons were transport costs (64%); pills running out (62%); not wanting other people to notice (62%); wanting to avoid side-effects (58%); and busy with other things (48%) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Socio-demographic characteristics of the sample.

https://doi.org/10.1371/journal.pone.0035547.t001

Factors Associated with Adherence to ART

The univariate analysis found that 24 variables were significantly associated with adherence (Table 2), although of these only ten remained significant in the final multivariate logistic regression model: patients who had not disclosed HIV status to anyone other than health workers (OR = 17.99, p =  0.014), who drank alcohol (OR = 12.89, p = <0.001), were female (OR =  6.91, p =  0.001), illiterate (OR =  4.58, p =  0.015), had a regimen side-effect (OR = 6.04, p = 0.025), duration of ART started ≤ 24 months (OR = 3.18, p =  0.009), travelled more than 1 hour to hospital (OR =  2.84, p =  0.035) were significantly more likely to be non-adherent. Similarly, patients who thought that HIV would disappear after ART (OR = 6.82, p = 0.008), that distance affects ART adherence (OR  = 15.33, p =  0.008), that you could stop taking the medicine once you felt better (OR =  6.43, p = 0.011), or that HIV was preventable by taking regular ART (OR = 4.31, p = 0.003) were significantly more likely to be non-adherent.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Potential factors influencing adherence to ART.

https://doi.org/10.1371/journal.pone.0035547.t002

The following were significant in the univariate analyses but were not found to be so in the multivariate analysis: being unemployed, missing scheduled hospital visit, method of travelling to hospital, age ≤ 35 years, number of pills taken per day, feeling stigmatised or discriminated against, knowing that ART caused problems and side-effects, and that treatment needs to be continued even when feeling better or weight has been gained. There was also no significant difference in the multivariate analyses between those patients who did or did not know that ART prevents HIV/AIDS progression; that ART did not cure AIDS; and that forgetting ART can cause problems.

Qualitative results.

The results of the in-depth interviews complemented the survey and shed further light on its findings. The thematic analysis suggested a range of factors which negatively influenced adherence to ART.

Discussion

The rate of adherence in our sample was significantly higher (85.5%) than has been observed in other studies in Asian developing countries that also relied upon self-report [22]–[24]. A meta-analysis found that patients in Africa had better adherence than their North-American counterparts with pooled estimates of 77% of participants achieving adherence in the African studies and 55% of participants in the North-American studies [9]. The possible explanation for greater adherence in our study might be due to using (a) self-reported adherence; (b) only first-line regimens; (c) strict adherence counselling sessions pre and post ART; and (d) recruiting participants through clinics so that only active patients participated (patients who actually went to the clinic).

Many factors identified in this study are consistent with the literature from both developed and developing countries. In this study, females reported poorer adherence than males as corroborated by care providers. This finding is unique since very few studies have reported that men were more likely to adhere to ART than women [25]–[27]. This is an important finding in the Nepalese context, as women are treated differently and unequally due to socio-cultural and economic factors. Nepalese women are recognized to be ill by family members only when they are bedridden and the head of the household (usually their husband or mother-in-law) decides that they are seriously ill. Women in this study were less likely to disclose their HIV status (66.2% vs 70.2), and were more vulnerable to discrimination by others compared to male PLHIV (45.1% vs 20.2%), both of which could be used to explain this finding. Different forms and types of HIV-related stigma exist such as self-stigma (self-blame or condemnation), perceived stigma (fear of stigma associated with disclosing HIV status) and enacted stigma (gossip, rejection, discrimination) and these were barriers to adherence [28], [29]. Our qualitative findings also revealed that stigma (or fear of stigma) was identified by the majority of interviewees as impeding adherence behaviour particularly among women. This will become an increasing problem because of increasing female infection rates in Nepal, but currently the Government still lacks programmes targeted to test, treat and improve treatment adherence for women [30]. Nepali women require special and urgent consideration because of their socio-cultural and sexual subordination [31], [32]. Additionally, women are most often responsible for care of the household and children, and as a consequence these practical barriers may adversely affect adherence among women more than among men [33]. Thus, interventions that focus on women taking ART may benefit from paying special attention to the multiple life demands that can interfere with their ability to refill medication or obey specific instructions from their care provider.

Other adherence barriers raised during the interviews included local culture, especially religious activities and festivals such as “Teej” for Hindu women and “Ramadan” for Muslims. Religious beliefs are complex concepts and are part of the basic assumptions which shape people’s identities and strongly affect their decision-making such as taking medication on fasting days [34], [35]. It is difficult to see how this could be overcome but our findings reinforce the importance of considering the religious and spiritual beliefs of PLHIV as part of medical care. It has been suggested that incorporating discussions about spiritual beliefs into adherence counselling could foster adherence [36]. As a result, patients could be motivated by seeing improvements in their health condition to continue adherence all the time. It is believed that most religions give freedom to eat on fasting day especially to the sick, children, and older people. This message needs to be reinforced during counselling.

Similarly, being illiterate, drinking alcohol and starting ART within the past two years were all associated with increased likelihood of non-adherence to ART in our study and similar findings have been reported elsewhere [25], [37]–[41]. Education may impact on adherence in several ways including facilitating communication with health care providers, increasing retention of information provided by health workers and thereby enhancing adherence to ART medication. It is possible that patients with limited literacy might be reluctant to ask others for the kind of help they need to take their medicines correctly [42]. Better-educated people convinced of ART efficacy, perhaps as a result of educational programmes, show a propensity towards better adherence [43]. A meta-analysis of drinking alcohol showed that individuals who drank alcohol had a reduced adherence to ART of between 40% to 50% [40]. Therefore, adherence interventions among females, those new to ART, illiterate patients and patients who drink alcohol need to consider these factors when reinforcing continuing adherence practices [33], [44]. It may be necessary to schedule follow-ups that are more frequent, monitor adherence more closely and ensure that education and discussions take place in a safe environment.

Patients’ beliefs, knowledge and expectations regarding treatment strongly influence their medical decision making [45]. Our findings show that a few patients believed that ART cured HIV or were unclear about how long they should take ART. Every ART-prescribed patient needs to understand that with appropriate treatment and continuing adherence HIV/AIDS is now a manageable chronic disease [2], [46]. Therefore, every patient needs to understand the importance of adherence because inaccurate information and misconceptions towards the disease and treatment regimens are associated with poor therapeutic outcomes [47] that in turn may be an impediment to achieving optimal levels of adherence [33], [48]. PLHIV who believe in the efficacy of ART are more likely to adhere [49]. Hence, care providers should continue to educate PLHIV and their families and develop intervention strategies that address the local context in order to encourage people to adhere to ART.

Although ART drugs have been provided free of charge since 2004 in Nepal, ART associated costs are not (including transport, prescription, procuring food and diagnostic testing) and remain barriers. In addition, 62% of respondents cited pills running out as a second main reason for skipped medication. These ART-associated costs have been found to be a barrier to adherence in other Asian developing countries [13] and many other developing countries [9], [50]–[52]. A study in India found that almost all the participants discussed the cost of ART as a barrier, with many reporting drug holidays, turning to family and or friends or taking drastic measures (i.e. selling jewels, property) for financial assistance [53]. A study in Botswana showed that adherence was higher when cost was not an obstacle and poor patients were able to achieve excellent rates of adherence when they had access to free laboratory monitoring and subsidized ART [54]. This is a useful finding and it is important that policy makers use this information to introduce social policy which includes subsidies for transport, and other costs associated with ART. Addressing the ART-associated cost issue of non-adherence in Nepal may therefore require a somewhat different approach to solutions applied in developed countries where financial issues are not such a major concern.

In the in-depth interviews, participants raised the issue that physicians prescribed ART for periods that were too short thus increasing costs associated with transportation to collect medication. This has also been documented elsewhere [55], [56]. Patients were worried and reported that to refill medicines every month caused difficulties in maintaining adherence. Thus, the Government of Nepal needs to consider carefully the amount of medicine that is prescribed on each occasion as well as subsidizing travel costs. To make available ART collection points in local health institutions may also help to manage this short period between refills. Once these have been implemented, further research will need to be focused on the effectiveness of these and other possible interventions to decrease barriers related to transportation costs.

In addition, non-disclosure of HIV status to anyone except health workers was associated with poorer adherence in our study and has been documented elsewhere too [9], [13], [57], [58]. Studies have shown that HIV disclosure is necessary to facilitate adherence to ART [59] and PLHIV who did not disclose to their family or peers had poorer adherence than those who disclosed their status [9], [59]. Our findings suggest that disclosure is an important issue, which should be discussed during the counselling sessions. For those who have not disclosed, providers should explain the importance of disclosure for the success of ART.

Skipping and stopping doses were attributed to fear that other people would discover their HIV status. Stigma and discrimination were not found to be significant influencing factors in the multivariate analysis, but in the in-depth interviews the majority of participants (seven female vs. four male) raised as this as a barrier to adherence. Patients were too embarrassed to take their medication in front of others, and concerned about privacy when collecting repeat prescriptions and taking medication; these worries increased non-adherence and have also been found in other studies outside Nepal [13], [58], [60]. One study stated that PLHIV were unwilling to seek treatment at the nearest health facility because of fear of stigmatization [61] and this same issue was discussed in our study too. Hence, policy makers should encourage a supportive environment where PLHIV do not need to worry about stigma and discrimination but talk openly in order to facilitate adherence. In addition, patients should also be taught strategies on how to handle taking pills in secret to increase adherence to their medication.

Distance to treatment centre is of great concern to PLHIV and one of the key factors preventing adherence in this sample. Patients who travelled more than one hour to hospital were more likely to be non-adherent and this was also discussed in the in-depth interviews. Participants stated that although patients were willing to take ART they became non-adherent because of difficulties in reaching the treatment centres due to unexpected transport and other strikes; long travel distance; geographical difficulty including lack of transportation services in many remote areas; and the seasonal deterioration of poorer roads during the rainy season. This has also been found to be the case with respect to maternity services in Nepal [62]. Others have also found that travel time and access to treatment centres were barriers to ART adherence [52], [63] and that better access to care was significantly associated with optimal adherence [13]. Patients who are from rural areas have difficulties in travelling long distances and finding their travel costs, and have most to gain from nearby ART facilities. Although this benefit may be offset by patients who fear of disclosure and worry about others finding out, are still travelling to more distant sites. Thus, any new policy will need to address this issue and improve access to medical care services by integrating ART treatment into the mainstream of health care rather than concentrating treatment in a limited number of ART centres, which may be hard to reach for many patients.

ART regimens have toxicities and adverse side-effects that vary from mild to severe and acute to chronic, which can prevent adherence [64]. Our study revealed that patients who had side-effects were more likely to be non-adherent and this has also been reported from several studies conducted in both developed and developing countries [9], [13], [65]. Similarly, patients who have been on ART for less than two years were more likely to be non-adherent, but there is no clear indication for why this is the case. There may not be a single reason behind non-adherence of this group but side-effects may be one of the possible explanation because the literature shows that up to 45% of ART prescribed patients discontinue one or more of the drugs within a year of starting due to side-effects and the inconvenience of the regimen [66]. Therefore, there is a need for continuing follow-up, clinical support and counselling for patients on ART in the early years of treatment as well as providing instructions on how to cope with side-effects [67]. Health care providers should review possible anticipated side-effects and develop treatment plans that either prevent, or at the very least, reduce the likelihood of side-effects happening. These may be tackled by strengthening institutional support, including clinicians and other health professionals informing patients in advance what side-effects to expect and how they might be managed, as well as prescribing antiemetic or anti-diarrheal agents to alleviate these side-effects. In addition interviewees raised the issue of discrepancies between the labelling and contents of packages of pills. Therefore, counting pills in front of patients is necessary to ensure complete dose of medicines prescribed to avoid incomplete supplies.

Participants discussed the twofold nature of social support systems: lack of support as a barrier to adherence and, when actually in place, as a facilitator of adherence [68]. Some people were accepting and supportive whilst others overtly or subtly distanced themselves from people with HIV. In this study, family members reminded 57.3% respondents of their medication although this was not significantly associated with adherence. Furthermore, findings from the in-depth interviews highlighted that family support increased the likelihood of patients maintaining optimal adherence. Others have also found that social and family support, communication and access to care that anticipate patients’ individual needs play an important positive role [9], [13], [69], and family acts as a facilitating factor of adherence primarily among children and women [70]. A meta-analysis reported that adherence is 1.74 times higher in patients from cohesive families [71] whereas absence of this is a factor leading to poor adherence [72]. Therefore, health care providers should continue discussions with new starters and follow up individuals without support to find an alternative source of help to foster adherence.