As shown below, the CAG dynamic was affected by life events and changing circumstances, such as a loss of geographical proximity to other members or the health facility, or change in social relationships. Family CAGs facilitated reporting and pill distribution, but the group could cease to function if there was conflict within a relationship. In some CAGs, the dynamic changed as pill counts were not carried out, members met less frequently or stopped meeting entirely before sending a representative to the clinic. Some members did not collect ART at the facility when it was their turn, and others stopped taking ART altogether. CAG members reported that health-care providers made membership of a CAG obligatory, instead of promoting voluntary participation. Some CAGs responded to adherence challenges by strengthening peer support through counselling and observed pill intake. Providers agreed that strengthening CAG rules and membership criteria could overcome these identified problems.
Events or circumstances affecting CAG dynamics
All participants agreed that a well-functioning CAG with a strong bond between members is built on social and geographical proximity: “each neighbourhood organizes its own groups, this is how we organize” (IDI29, female CAG member).
Loss of geographical affinity
CAG functioning can become challenging if members migrate to larger, more urban areas such as the city of Tete: “There, in our neighbourhood, some groups existed, but they dissolved ( …) There was an exodus of people, many went to the city.” (IDI 8, male CAG member). Working in the mines in particular was reported to interfere with being an active and engaged CAG member as this involved longer periods away from home.
Some CAG members travelled for employment, but were still classified as belonging to CAGs by health facility staff, as this health care provider explained:
“The economic situation faced by the population makes things difficult. A CAG member wants to go to the mines, is there, but is [still] counted as being a CAG member ( …) sometimes he stays there for a long time without picking up his drugs ( …) I don’t know when he’ll be back. And when he’s there, he’s not taking his medication.” (FGD 1, clinical officer).
On the other hand, CAG members assisted each other in collecting and distributing ART refills for short and unplanned trips, taking advantage of the flexibility that CAG membership allows: “I can meet with my colleague [fellow CAG member] even on a Saturday.” (IDI 16, female CAG member).
CAG members also reported being able to borrow pills from each other if they had to travel unexpectedly, returning them to the member upon their return: “If someone is preparing to travel before their next date to collect drugs, then he can ask another person for them ( …) this is amicable. Then, when he receives his [ART refill], he returns the medication.” (FGD 6, CAG members).
Some members decided to leave their CAG as their frequent travels did not allow them to fulfill responsibilities, such as representing the CAG at the health facility when it was their turn:
“I have a date to pick up the drugs but I’m in the city ( …) I move around a lot, I’m a person who doesn’t stay in a fixed place ( …) they called me to say “we’re always the ones who pick up the drugs”. Leave me; I will stay by myself [leave the CAG].” (IDI 39, female CAG member).
Loss of social affinity
Some members returned to individual clinic care because they had chosen to, or because it was suggested to them by the other members of their group. Participants reported how the composition of CAGs could change over time as members no longer got along with each other.
Sometimes even the CAG focal point was excluded from their group by the other members. “We had a focal point, but he didn’t join our group anymore, he didn’t come, thus we put him aside ( …) the group functionned without him.” (IDI 13, female CAG member).
Another cause of loss of social coherence within CAGs was alcohol use. Members known to consume large amounts of alcohol could be asked to leave a CAG, as reported by this interviewee, who believed that members who drank alcohol may then disclose the HIV status of others: “When a person drinks, he doesn’t control what he says ( …) he starts saying things he should not mention, he starts talking about the group so we exclude people who drink.” (IDI 29, female CAG member).
Alcohol use was also perceived as having serious consequences, with one person reportedly becoming ‘addicted when he was in the mines’ and subsequently dying.
Pregnancy
Pregnancy was reported as another life event causing female members to leave CAGs due to the need for increased medical follow-up, which in turn altered the group dynamic. Pregnant members returned to individual care because “the [pregnant] person has to go monthly to the health facility, whereas in the CAG this is different, a person goes once, depending on the number of group members” (IDI 33, female CAG member).
Benefits and challenges with family CAGs
CAGs were also formed within families: “
we are in a group of four people; I am with my three sisters-in-law. We were five including my brother, but my brother just passed away, now we are four.” (IDI 22, female CAG member), which was seen as beneficial because of the convenience of being in the same household. This was also supported by a health-care provider:
“ … they live in the same backyard, so they sit there to check this matter, who took it [the pills], who did not take it” (IDI 22) and “you have direct control: you don’t have to do any effort to take the bottles and to call the people for a meeting” (FGD 4, CAG focal points).
However, other health-care providers suspected that adherence challenges were under-reported in family CAGs as members wanted to protect each other from criticism or judgement from a nurse.
“A member came in the following month, and he collects pills for another [member], but he does not “tell the truth”. He does not say what is happening in the group, and the health care provider does not know whether or not he is not taking his treatment, or if he already left the CAG.” (FGD 2, nurse)
Health care providers reported problems in a CAG after a couple joined a group and their relationship ended. “One of the group members came to me with about four bottles [ART] of one the members of the group. (…) He said ‘well, I’m the husband of one the members of the group, we are divorced and unfortunately I haven’t been able to find her to give her the drugs for four months.’ (FGD 2, health care providers, nurse).
Modifications to the CAG dynamic
Most CAG members who were interviewed reported that they met in the community, typically the day before the next scheduled ART refill date. Many members reported that they carried out a pill-count, shared experiences, counselled each other and reported problems to the nurses at the health facility: “the day before picking up the refill, we meet to talk a little bit … we do the pill count ( …) we see who is taking it [ART] properly and who is not ( …) this helps us.” (IDI 10, female CAG member).
Adaptations to CAG meetings and reporting
Many CAG members, however, reported adapting the typical CAG model and ‘rules’ to better fit with their circumstances. Some did not conduct a formal pill count:” we don’t ask people to bring the bottles for verification [pill-count]: we just ask the person “how many pills are remaining?” (IDI 13, female CAG member).
Other groups did not organise meetings for all CAG members together (a key part of creating a CAG), with the focal point making home visits to assess the adherence of members instead: “We don’t come together. We are in a village, in a neighbourhood, and I, as focal point, go door-to-door. One day before the next refill, I take the CAG group card and have a look at the number of pills that are remaining.” (FGD 4, CAG focal points).
Sometimes members stopped participating in their CAG, and did not pick up their pills from the focal point in their community. Then the CAG representative can return to the health facility to return the pills to the nurse. Providers reported that those who represented the CAG at the monthly health-facility visit were not always well-informed about the whereabouts of the other members. According to providers, peer groups that were established many years ago had weakened and members rarely met in person to discuss treatment adherence. As the below examples show, health-care providers reported that some CAG members asked family members to distribute ART instead of meeting with other members themselves:
“When that member stays there … he is making charcoal, or he is selling products along the street, then I didn’t manage [to meet them], then maybe I gave [the medication] to the son.” (FGD 2, health care providers, clinical officer).
“We meet in this season as there are no agricultural activities. But, in the season of agricultural activities each person wakes up to go to the field and returns tired so that’s why we don’t talk.” (IDI 2, female CAG member).
Rotating representation of the CAG at the health facility was reported to be challenging by CAG members and health care providers, which also meant that there were some members who were not benefitting from viral load and CD4 monitoring. “One or two go regularly to the health facility, and the other four, or three, five, might not come to the health facility. Their blood is not collected to do the CD4 analysis.” (FGD 2, health care providers, nurse).
According to health-care providers, some members were still reported as being active in the group, even though they did not pick up their pills through the CAGs.
Some CAG members were reported as being ongoing CAG members, but did not take their pills. In one case, a CAG member died and medication was found in their home: “She died and when we looked in her house we encountered these unopened bottles.” (FGD 4, CAG focal points). Providers suspected that during pill counts such patients may “take an old bottle and show it [the remaining number of pills], while the new bottle is kept aside” (FGD 1, health care providers, nurses).
Pressure to join CAG
Some patients felt pressured to join a CAG by the nurse or by fellow patients, which resulted in CAGs lacking in a peer dynamic and bond: “those from the clinic called me and told me that if did not accept [joining a CAG], that maybe they will not give me their medication” (FGD 1, health care providers, nurse).
Sometimes patients in individual care are asked by the care providers to join a CAG without exploring the patient’s preferences: “Haaaaa, it was them [care providers] who told me to join these people ( …) they did not explain the CAG. I believed that if they told me to join, I have to believe that the CAG is good for me.” (IDI 9, male CAG member).
Moreover, some CAG members found it difficult to negotiate changes such as leaving one CAG to join another: “Well, I cannot order the government and say: I don’t want this, I don’t want that ( …) I have no voice. But for me it would be good if I could change group.” (IDI 5, female CAG member).
Response to challenges
Some members identified non-adherence by observing the deterioration of a member’s health as ‘the first sign’: “
The first sign is a person with continuous health problems” (IDI 16:128–129). On other occasions “
[lab] results arrive which show that a person has difficulties with taking [the medication]” (IDI 16, female CAG member). CAG responses to identified non-adherence often included home visits to explore potential barriers. Some CAGs were able to adapt the social and peer dynamic to respond to such emerging needs:
“When she was not adhering to her treatment [she said] ‘I don’t want this treatment anymore because I have many problems with my son, it’s better to stop taking medication so that I can die.’ But when I did the follow-up to improve adherence, she was able to say: “Yes, I’m adhering well with the treatment and with the pill count, also with the support of the group.” (FGD 5, health care providers, counsellors:).
In cases where CAG members did not accept their HIV status, it became challenging to offer them peer support: “as the person doesn’t feel sick, and sometimes thinks it [HIV] is an invention, it’s tough to accept, therefore it is difficult for a person who is ill to take medication as he feels healthy” (FGD 3, CAG focal points).
Others wanted to be left alone, as reported by CAG focal points: “This is my life, if I have to die, I will die alone, do not interfere.” (FGD 3, CAG focal points).
Some CAGs such as the one cited here implemented direct observed therapy (DOT) to support each other: “The group was always providing support, there is a neighbour who always takes the medication together, and now she’s fine.” (FGD 5, health care providers, counsellors).
Such an intensified adherence support strategy can also involve all the members of the group as well as monitoring of the strategy: “In the group, we take turns to support the person ( …), we stay there to observe [the pill intake], first goes this person, then goes the other ( …). After some time we go and have a look if the person returned to the normal routine, and then we leave the person taking the medication by himself.” (IDI 16, female CAG member).
Providers felt that they should apply more rigorously pre-established criteria when forming CAGs. Moreover, they proposed to review the composition of the CAGs to identify those who were active, and to verify if members were still living in the same geographical area, or if some had moved: “Now, what we should do ( …) is to revitalize the CAGs, truly check who is inside [the CAG]. They have to be from the same neighbourhood ( …) because the CAG is very important.” (FGD 1, health care providers, nurses).
To properly assess the CAG model of care, providers proposed various strategies, including: “taking turns to visit a different neighbourhood every month, to talk with the group and find out how things are going, and if people are adhering or not. Because even though some had no contact [with the clinic], there are always a few who know about the others.” (FGD 1, clinical officer).
Providers believed that relevant information about the whereabouts of patients can be adequately assessed through the social network of CAG members, even if some had lost contact with the clinic.
The role of the focal point was also discussed as something which may need to change: “Some focal points of some groups will need to be replaced ( …). We may need to meet with the group to identify another focal point, who is a bit more dynamic, and who really supports the adherence of the group.” (FGD 2, health care providers, nurse). They agreed that a focal point had to be “a person who is the pillar of that group with regards to treatment” (FGD 5, health care providers, counsellors).