Peer support after prison improves linkage to HIV care, South African study shows

Providing structured peer support to people with HIV after release from prison greatly improves enrolment in HIV treatment services, a randomised study carried out in South Africa has found.

The study, conducted by the Aurum Institute and designed by researchers at the University of Witwatersrand, is published in Lancet HIV.

Approximately one in seven incarcerated people in Africa are living with HIV with especially high HIV prevalence among women in prison in southern Africa. One study found that half of all women in prison in Malawi are living with HIV.

In South Africa, an observational study found that only 34% of people with HIV had visited an HIV clinic within 90 days of release, pointing to serious gaps in continuity of care. At any time, approximately 37,000 people with HIV are estimated to be incarcerated in South Africa and of these, around half return to the community each year.

A study of people with HIV in Uganda who had disengaged from HIV care and who had detectable viral load found that incarceration was a frequent cause of extended treatment interruptions among men with HIV.

Linkage to care after release may prove challenging for people with HIV due to stigma, lack of social support, economic marginalisation and substance use disorders. In South Africa, the usual form of linkage to care offered is a referral letter to a primary health care facility on release from prison.

Several studies in high-income settings have shown that peer support after incarceration can improve linkage to care. Researchers in South Africa wanted to assess the impact of a model of linkage to care that offered peer support for people with HIV leaving prison. They designed a randomised trial that recruited people on release from five correctional facilities in Gauteng province.

Participants were assigned either to standard care (a referral letter to primary care on release) or an intervention group, who joined a community adherence support group that met twice a month. The group sessions had a curriculum designed to support people as they adjusted to life after prison and were facilitated by a social worker and a person with HIV who had experience of incarceration. Although participants should have received HIV medication through the groups, this aspect of the intervention was implemented unevenly due to COVID-19 restrictions.

The study recruited 176 participants, 59 assigned to care as usual and 117 to the community adherence support group. Almost all the study population was male (95%), most were under 40 (86%) and 71% had been incarcerated at least once before. The median time of their current incarceration was 0.82 years.

Sixty percent had been diagnosed with HIV in prison and two-thirds had started antiretroviral treatment in prison too. Forty-four percent had a viral load below 1000 before release, 8% had unsuppressed viral load (above 1000). Pre-release viral load measurements were missing for 48% of participants.

Of the 116 people assigned to the study intervention arm, 44 actually attended a community adherence support group session, 80% of those judged to be linked to a group and physically able to attend. Thirty-two people either did not respond to attempts to contact them after release or no longer wished to take part, while 19 people were unable to attend meetings due to COVID-19-related restrictions on gatherings in 2020. Those assigned to a group attended a median of five sessions.

The primary outcome of the study was enrolment in an HIV treatment service. It was possible to ascertain enrolment status for 161 participants. Of those randomised, 61% of those assigned to the support group intervention and 36% of those assigned to care as usual enroled in an HIV treatment service. Participants assigned to the study intervention were 70% more likely to be enrolled in care after six months than people assigned to care as usual.

Of those who were not enrolled in care (n=83), nearly half (41) were living on the streets, 19 had returned to prison, 14 were uncontactable, three had died and six had left South Africa.

The study also assessed the influence of stigma, HIV disclosure and substance use on engagement in care. At release, 70% of participants said they had previously disclosed their HIV status to family or friends, 32% had experienced stigmatisation upon disclosure and 33% reported internalised stigma.

Moderate to severe depression was present at baseline in 14% of participants, 22% were daily opioid users in the past three months, in or out of prison, 11% were daily users of alcohol and 10% were daily users of stimulants.

Subgroup analyses showed that enrolment with a treatment centre was associated with HIV disclosure, no or mild depression and not reporting daily use of alcohol or opioids.

A qualitative study in which 16 people with HIV were interviewed about their experiences of participating in the groups found that participants highly valued the peer support element of the group, not only regarding HIV treatment but also adjustment to life after prison. Participants felt supported and also motivated to support others through attending the group meetings.

The researchers say that more research on this model is needed before it can be implemented by treatment programmes. The impact of group participation on viral suppression rates should be investigated, together with the costs of the intervention and the feasibility of scale-up.

In an accompanying editorial comment, Dr Anne Spaulding of Emory University says that the study adds to the “small toolbox of what works” to address the challenge of attrition from HIV care after release from prison. She says that future research, including her own, needs to investigate whether the same approach can be adopted to support people at high risk of HIV acquisition after release to obtain PrEP and keep taking it while at risk.