Disengagement from HIV care happens for numerous reasons and efforts to promote re-engagement in care should develop strategies that address the multiple patterns of disengagement, a study of people with unsuppressed viral load in Uganda reports.
Interrupting treatment, or disengaging from care, is common, although calculating what proportion of people with HIV are not currently engaged in care is complicated by the fact that people may move from one clinic to another.
A study of unplanned care interruptions in South Africa found that interrupting treatment for more than six months was associated with an increased risk of death after resuming treatment. The large study found that approximately one in three people who started treatment from 2004 onwards interrupted treatment and of those who interrupted, 30% did so within a year of starting treatment.
Understanding why people interrupt treatment can help in devising strategies to re-engage people with HIV in care and minimise disengagement with care. The South African study could not investigate reasons for treatment interruptions because these were not recorded consistently in medical records.
To learn more about the reasons for disengagement from care and how and when people had disengaged, researchers from Bloomberg School of Public Health at Johns Hopkins University and the Rakai Health Sciences Program in Uganda carried out a qualitative study in people with high viral loads in the Rakai Community Cohort Study in Uganda. The Rakai district combines semi-urban trading settlements, farming and fishing villages and has a high prevalence of HIV, as well as high mobility.
The researchers interviewed 38 people in Rakai with viral loads above 1000 copies/ml between 2021 and 2023. The group represented approximately 10% of those with high-level viremia in the Rakai Community Cohort during the study period. The median age of interviewees was 34, just over two-thirds were men (68%), 39% were cohabiting and 28% were in a non-cohabiting partnership. The median time since HIV diagnosis was five years and 90% were on first-line treatment (mostly tenofovir disoproxil / emtricitabine / dolutegravir).
The interviews identified six types of disengagement in people with unsuppressed viral load in Uganda.
Denial and non-disclosure leading to delays in testing for HIV and subsequent engagement in care, predominantly in men. People who had delayed starting treatment were often fearful of the commitment involved in taking daily medication or had not disclosed their HIV status to intimate partners.
Early treatment discontinuation due to difficulties in accepting an HIV diagnosis, side effects or fear of the consequences of disclosing an HIV diagnosis. Men were more likely to quit treatment early due to side effects, while women discontinued treatment soon after starting due to concerns about disclosure, including fear of intimate partner violence.
Treatment cycling, consisting of numerous short interruptions in treatment due to frequent changes in residence, driven by economic circumstances. Running out of medication and not being able to attend clinic appointments while in another location tended to characterise treatment cycling. These episodes usually lasted less than three months before return to the original healthcare facility.
Prolonged treatment interruption. Mostly identified in men, this was usually a consequence of incarceration or severe mental health problems. If people returned to care after a prolonged interruption, it was often due to the return or emergence of clinical symptoms.
Transfer-related treatment disruption, often caused by COVID-19-related travel restrictions. During the COVID-19 pandemic, restrictions prohibiting movement between districts were imposed in Uganda, as in many other countries. The restrictions made it difficult to obtain the paperwork needed for a successful transfer of care, leading to interruptions in medication supply. People were sometimes left in ‘therapeutic limbo’ by processes which prioritised procedures designed to safeguard supply chains and reporting performance indicators over person-centred care.
Episodic viraemia. Not everyone with unsuppressed viral load in this study had disengaged from care. With no history of treatment interruption or disengagement from care, people with episodic viraemia were more often women, taking second-line treatment. Reasons for viral rebound were uncertain. People experiencing episodes of viraemia were often accused of non-adherence by healthcare workers, leading to distress and depression.
The patterns of disengagement suggest that measures such as extra adherence support or contacting people who missed appointments are unlikely to address the underlying reasons why people with HIV have disengaged from care.
Whereas men tended to report the drivers of care disengagement as psychosocial, women most frequently reported interpersonal drivers of disengagement, such as fear of intimate partner violence.
Psychosocial distress and mental health problems were frequently implicated in disengagement from care. Same-day ART initiation processes may overlook the mental health needs of people who have just been diagnosed with HIV, the researchers argue. Psychosocial support and mental health services are also needed by clinically stable people on ART who experience unexplained viral rebound.
The researchers also say that more attention needs to be paid to continuity of care for incarcerated people; incarceration was a frequent feature in the stories of people who had experienced frequent treatment interruptions.
Treatment programmes also need to adopt greater flexibility in responding to people moving from one area to another, whether through dispensing medication for longer periods or lowering the barriers for people wishing to transfer care from one clinic to another, the researchers conclude.
Rosen JG et al. Identifying longitudinal patterns of HIV treatment (dis)engagement and re-engagement from oral histories of virologically unsuppressed persons in Uganda: a thematic trajectory analysis. Social Science & Medicine 339: 116386, 2023.