Individual-level factors.

Unstructured treatment holidays were noted amongst adolescents and were expressed as a desire to experience a drug-free life or a quest to understand the effects of being off drugs. However this was also as a result of underlying factors that included drug fatigue, pill burden, depression and stigma. Addressing these underlying factors could lead to resumption of treatment although in some cases the patients resumed treatment due to deteriorating health.

“There's a girl who…has been on treatment for some time but her viral load was high, so when I tried talking to her she was like "ah aah [No!]! Don’t tell me anything, I know everything, I’ve decided I’m not taking [the drugs]”.By the way she came today I was surprised!…when they asked her about her drugs she was like “I felt like I was tired of the drugs, so I needed a break”. Sometimes adolescents they give themselves breaks…for like 3 months but when he becomes sick, very sick, he comes back and says, “Now I want medicine”. It’s like they want to first see what will happen when [on] a break”.

(Expert client-1)

Social-level factors.

Delays in disclosing HIV status to perinatally infected children prior to adolescence were common and could lead to non-adherence. In accordance with government regulation [30],it is the duty of the caretakers to disclose to their children, but they were reluctant and cited barriers such as fear of being blamed for the infection, fear of unintended disclosure of the parent HIV status by the child, fear of anticipated negative reactions and concern for lack of cognitive ability of the child to comprehend the implications of the disease on their health. Due to this, caretakers would sometimes lie to their children about their condition indicating that they were suffering from other chronic diseases like asthma or cancer. The concealment of status by implicating other diseases could however impact on the adherence of the adolescents as it foils the understanding of the importance of adherence and subsequent consequences, which are specific to HIV treatment. Moreover this could result in anger and depression when the adolescents become aware of their condition. Health-care providers could at times be obligated to assist with disclosure mainly when it was imperative for the child to know their status for example during adherence support at treatment failure or treatment switch which was done as a means of averting further failures.

“This time we had a very strict doctor whom they found [during clinic visit]…she said “You must disclose…because the boy is now 15 years and he is already on second-line [treatment]…So, the doctor took her [guardian] to the counselor by force, and they made her to disclose”

(Pharmacist-1)

Participants also noted that delays in disclosure were partly due to lack of supportive mechanisms to assist caretakers with disclosure.

“Caretakers can manage when they are prepared on time about disclosure. Usually they are stuck when you tell them [abruptly] that they need to tell the child. They are not prepared on how [to go about it]?”

(Counselor-3)

It was also noted that the lack of support for disclosure could result with poor or partial disclosures where children knew about their HIV status but did not fully comprehend the implications of the disease. Late or poorly disclosed adolescents were likely to stop treatment, react negatively to their caretakers and sometimes deny their HIV status.

Diminishing or lack of family support: Adolescents living with biological parents were perceived to have better adherence compared to those living with caretakers. The latter were reported to lack support for food, transport, medication reminders, and accompaniment by caretakers during clinic visits. Some adolescents also reported that the caretakers denied them education support on the misconception that they would die soon. However, as the adolescents grow older the need to be independent as well as other structural barriers could result with diminishing influence of the parents on the adolescent’s adherence behaviour or support.

“I’m asking [the mother] about so and so, she says "his drugs are at home" So the boy would take [a few] tin[s], if he is dispensed with 3-months supply on 2nd line, he would take 2 tins, leave the 4 at home. Then the mother would not bother, how comes, its long since he left his medicine. You would call her and she says "so and so took medicine, some is still at home". Till one time I told the woman that you carry all whatever you have at home [and bring to the clinic], she carried tins and tins and returned them”

(Counselor- 1)

Perceived and experienced stigma in boarding schools: Adherence among students in boarding schools was reported to be poor with the majority already being on second-line treatment. Poor adherence was linked to stigma, and this was influenced by where and how the students take their medication. Two groups emerged in the discussion: students who keep their own medication in the dormitories and those who keep them with the school nurse. Keeping drugs with the school nurse was reported to be beneficial due to likely reminders, close monitoring and easiness in getting permission for clinic visits. The preference to remain with the drugs was however due to having not disclosed to the school administration, the desire to be independent and fear of involuntary disclosure by the teachers or nurse to fellow students. Moreover, prevention messages such as ‘AIDS Kills’ placed in most schools also inadvertently stigmatized the infected students making it difficult for them to disclose.

Students keeping their own medication were said to face lack of privacy, anticipated (perceived) and enacted (experienced) stigma from peers and they also sometimes failed to access medication from the dormitories on time.

“There is a girl we lost, she passed away, she was 18… she had [experienced] stigma at school because they came across her drugs in her suitcase, and they pulled them out and they put them there and put her [medical] card on her bed and she was a head-girl and that killed her [spirit]! She had to switch school. Most of them you get these calls, when they are saying they have found out, you see, so she had to switch out schools”

(Counselor-3)

However, those keeping their medication with the school nurse also reported facing stigma emerging from their peers who persistently inquire of their frequent visits to the infirmary. Moreover these students could also experience stigma arising from inadvertent disclosures by the school nurse or administration.

“In the first school I disclosed…. but our nurse discriminated me…whenever the time of [taking] drugs approached she could come in the class [and say], "you don’t know that you have to take your drugs? Come and take your drugs" while all other students are there, [listening]… So they [other students] used to inquire what drugs do you take, the nurse is [always asking]…"you know you have your problem come and take your drugs ah"…If you don’t want you will die" ah, while other people are listening so [you get such] nicknames, “Mr. drugs, that is madam drugs”

(Expert client-2)

Health-care level factors.

Declining or lack of clinic support: Most ALWHA at the study facilities were reported to have been infected through vertical transmission and had started ART early in life at the pediatric clinic. At 18 years of age they are expected to transition to the adult-based care. Transitioning was however a challenge due to difficulties in integrating with adult patients, reduced attention from health-care providers, long waiting hours, lack of medicine for treatment of other ailments and a general sense of abandonment from health care workers. Subsequently some of the adolescents would be lost-to-follow up.

“When we would transit them to adult [clinics], some of them would fail to come to pick their drugs, they were not getting the attention they were getting in the pediatric [section], because in pediatric, they are few [in numbers] and we try to know them personally”

(Pharmacist-1)

Although most participants expressed the importance of peer group support across all patients, it was noted that a decline in funding had either resulted with fewer activities or the complete phasing-out of these groups. The peer groups were reported to have served a pivotal role in providing an avenue for strengthening adherence, dealing with emerging challenges, providing motivation from peers and HIV-infected role models. Moreover, they also served as platform for social networking, where participants would get potential life partners. The subsequent decline of these groups was reported to be have impacted especially the adolescents with some of them opting out of care.

“Other challenges are the lack of peer groups and activities for adolescents. We used to have them but because of funding they stopped. And some adolescents dropped off after these activities and some even died. They had a strong belonging to these groups”

(Counselor-6)

Moreover, while it was noted that adolescents were a group in need of more counseling, this was hindered by a shortage of counselors. Frequent transfers of health-care providers due to shortage in staff also posed a problem to the adolescents who found it a challenge to confide and discuss freely about their health with new health-care providers. The use of social media coordinated by the clinic was a notably noble alternative, which was used both socially and formally to facilitate health-care providers and peer-peer counseling but was said to lack sufficient participation from health-care providers.

“And another thing is all about our social network pages, especially the Facebook ones. We need more counselors there…one can post via a friend so that others can know how to handle the problem but you fail to get even a counselor who can comment on the post or give advice on the post”

(Expert client-2)

Individual-level factors.

Temporary migrants and challenges with treatment access: Mobile persons especially those travelling temporarily to nearby countries for business comprised a substantial majority of patients and they had notably poor appointment keeping and drug pickups. Reasons given included lack of funds to cover for transport costs, perceived feeling of wellness hence prioritizing business to drug refill visits, desire to match clinic visits with trips for replenishing business stocks, and political conflicts in the host countries (mainly Southern Sudan and Congo), which affected their travel.

“I’ve had experiences of people saying that they travelled to Juba [South Sudan]…they go there for business…hoping to come back like after 3 months and the majority of them were complaining of this recent war that was in Juba, they could not make it…others talk about transport, they didn’t get enough money to come back [for drug refills] … then others talk of no proper health facility elsewhere that could give them drugs, otherwise they would have continued taking HAART…others would point-blank tell you they were not feeling ill, health-wise they were feeling ok, and since they were doing their business they decided to continue [with] their business”.

(Counselor-4)

Other mobile populations included commercial sex workers, truck drivers, persons relocating to the villages for farming, and temporary emigrants. As a way of supporting these patients, they were supplied with drugs for 6–12 months instead of the custom 3 months so as to minimize on the number of drug-refill visits. Relatives or friends could also pick up drugs periodically on behalf of the patients but they would still be required to visit the clinics for blood tests at least once or twice a year. Despite these interventions, some clients still had poor appointment keeping prompting the health-care providers to encourage them to transfer to near-by facilities. Some were however reluctant to transfer-out citing poor standards in other facilities. Some migrants also reported reluctance by some of the host countries to provide them with ART.

Individual-level factors.

Disclosure in intimate relationships: New challenges with disclosure were cited when engaging in romantic relationships leading to compromises in adherence and potentially fuelling new infections. In adults, non-disclosure was linked to avoiding marital conflicts, fears of losing the partner or/and financial support. Adolescents generally lacked the skills to disclose and also feared losing their partner.

“In young adults, there are those getting boyfriends and girlfriends. They don’t disclose, and that is what prevents them from taking the drugs. They feel ashamed and don’t want to lose their partners”

(Expert client-3)

While support on disclosure was offered at the clinic, some health-care providers acknowledged their own bias when counseling discordant couples because they did not approve such relationships. This in turn could lead to the clients not to seek for disclosure support in subsequent times.

“They will tell you, when we take these partners to the hospital…. they are quick to judge you, you who is positive and yet you have brought in your partner really for testing. So they don’t help them and they end up breaking up instead. They don’t support them into this relationship and guide them on what to do. Instead they say heey! Trouble! So they separate them. So [they] are stuck with their relationships, they can’t come out open, [they have questions] who is going to support [me]… where should I go for counseling anyway? Because we are already biased”

(Counselor-03)

Health-care level factors.

Treatment-related factors: Treatment side effects were also a concern for long-term treated patients in particular those arising after switch to second-line regimens. In adolescents this was said to result with re-emergence of stigma due to changes in body fat distribution from lopinavir-based treatment, and yellow eyes (jaundice) from atazanavir.

“We have many people who were changed to second-line, they were fat but now they are small. And they ask the doctor's why? Does ‘lopinavir’ takes off the kilograms…they fail to adhere whenever they see they have lost weight, they are looking bad, it seems as if ‘lopinavir’ is the cause”

(Expert client-2)

There were also challenges with drug supply logistics where separate pills were supplied for use by patients on fixed dose combinations (FDC). Moreover there were reports where facilities were supplied with pediatric formulations involving multiple low milligram tablets, for use by adults thereby increasing the pill burden.

“They send pediatric drugs for use in managing adults. So what we do, if the person has been taking like one pill, she will take 3 or 4, to make an adult dose. It is a challenge, which increases on client's pill burden”

(Counselor-8)

Moreover, adults also expressed a dislike of the sweet taste of pediatric drugs. In addition, facilities reported being supplied with drug of different brands, some that had bitter taste (uncoated drugs) and sometimes with varying colors and sizes. This was reported to confuse the patients and affect their adherence especially when they were given the bitter drugs. There were also claims of emerging side effects associated with change of drug brands.

“When we change from one company to another, some do experience some abnormal side-effects and they always say: “when I was taking the other type I didn’t have any problem but this time…”. Before we used to say that maybe they are used to this type of drug, but it does really affect them. When you change them and put them to a previous one, the complaints do go away.”

(Pharmacist-2)

Although there were no cases of drug stock-outs, there were reports of insufficient supply of drugs or those with shorter expiration period, which necessitated more frequent drug refill visits. There was also a limited supply of third-line drugs, which was mainly available to patients enrolled in medical research.

Staff shortages and missed counseling opportunities: Staff shortage leading to increased workload and long waiting times resulted in missed opportunities for adequately counseling the patients on vital information such as interpretation of laboratory test results. This could lead to misconceptions with some patients inferring that an undetectable viral load implied that they had been cured of the disease.

““So when [viral-load test] results were undetected, the person [clinician] would say that ‘akauka takalabika’ directly translating into the virus is invisible, so the patients would take it as gospel truth [or]…they say ‘akauka tekalio’, meaning the virus is not there…. The health worker should explain to those patients: last time we tested your viral load is undetectable, it doesn't mean you are healed, it simply means that you are taking your drugs well and the virus has just slept so you have to keep taking drugs so that the virus keeps sleeping. When you stop, it wakes up. But most of them don't get that explanation”.

(FGD-2)