Tracking down patients lost to HIV care: home visits need to be carefully targeted

Keith Alcorn
Published: 06 April 2017

Efforts to track patients lost to care are likely to produce very little return on the money and time spent unless clinics target tracing activities carefully, a study of patients lost to follow-up in Kenya, Uganda and Tanzania has found.

The findings, published in the journal Clinical Infectious Diseases, show that almost one in four of these patients turned out to have died and one in five had transferred care to another clinic.

Considering that some cohorts report that up to one-third of patients are lost from care after five years, follow-up of lost patients is vital to bring people back into care and re-establish antiretroviral treatment (ART) and viral suppression.

Tracking efforts might involve a combination of the following:

  • Cell phone text messages or calls to remind people to attend the clinic.
  • Visits by community health workers to the patient’s last known address.
  • Checks with other clinics to find out if a patient has sought care elsewhere.

Although tracking by community health care workers is a common means of looking for patients who are lost from care, it is unclear how well tracking efforts work, and whether they could be better targeted. A meta-analysis of studies of physical tracing found that retention in care was higher in treatment programmes that practised tracking, but it’s not clear whether this association is a direct consequence of tracing or just evidence that better-resourced clinics with high-quality care tend to retain more patients.

Current plans to develop the health care workforce by recruiting two million community health workers over the next two years assume that many of those workers will be employed in physical tracing in the community. But is that the best use of human resources, and if tracing does take place, how can it deliver the maximum return on investment?

To investigate whether tracing is effective, and which sub-groups of patients are most profitably traced to improve retention, researchers associated with treatment programmes in Kenya, Tanzania and Uganda looked at the effects of tracing in random samples of patients who had been lost to follow-up at 14 clinics.

Patients were eligible for tracing if they were more than 90 days late for their most recent appointment and had attended one of the study clinics in the previous two-and-a-half years. A random sample of patients missing from care was selected. Patients were traced by community health workers, commonly people living with HIV, who were experienced in seeking patients in the community.

If patients were traced, they were asked whether they were in care elsewhere, and their reasons for transferring or stopping care. Those who had dropped out of care were encouraged to return to their clinic.

A total of 5781 patients were lost to follow-up, and 991 (17%) were selected for tracing. Of these, 40% had been lost for 1-2 years and 22% for more than two years. Forty-five per cent had been on ART for more than a year, but 21% had started treatment less than a month before their last clinic visit.

Tracing showed that 23.5% of missing patients were dead, 21% were in care elsewhere, 13% were untraceable and 27% were reported to be alive but could not be contacted. Only 15% of those missing from care could be interviewed and ascertained as out of care.

One year after the sample was selected, 13% of those selected for tracing had returned to care compared to 10% of those lost to follow up not selected for tracing (adjusted hazard ratio 1.30, 95% CI 1.08-1.58). The researchers calculated that it was necessary to attempt to contact 33 patients in order to bring one patient back to care.

But, when the analysis was confined to those patients who were contacted by a community health worker, the probability of return increased from 15.1% to 37.1%, implying that it would be necessary to contact between four and five patients in order to bring one patient back into care.

Return to the clinic was most likely within 14 days of tracing and the researchers found that likelihood of returning to care declined rapidly after that point.

The performance of tracing might be improved by developing clinical prediction rules to select patients for tracing, based on analysis of deaths in a cohort. Previous research by the same research group found that death in this sample of people lost to follow-up was strongly predicted by:

  • Male sex,
  • Last CD4 cell count < 50 cells/mm3.
  • WHO stage 3 or 4 HIV disease at treatment initiation.

They also found that patients at two out of five clinics in which the research took place were more likely to have died if they were lost from care. That study found that 15.8% of all those who started ART at five clinics in Kenya, Tanzania and Uganda died within three years of starting treatment.

Tracing might also produce a greater yield if patients who had transferred care to other facilities could be identified through electronic records prior to selecting patients for face-to-face tracing. In an accompanying editorial Wendy Armstrong and Carlos del Rio of Emory University, Atlanta, highlight the barriers in the American healthcare system to sharing patient information between clinics for the purpose of reducing loss to follow-up. Better exchange of information between clinics and across state boundaries will be important for achieving the 90% viral suppression target in the United States, they argue.

References

Armstrong E, del Rio C Patient tracking as a tool to improve retention in care, is the juice worth the squeeze? Clin Infect Dis, advance online publication, March 2017.

Bershetyn A et al. The causal effect of tracing by peer health workers on return to clinic among patients who were lost to follow-up from antiretroviral therapy in eastern Africa: a randomized “natural experiment” arising from surveillance of lost patients. Clin Infect Dis, advance online publication, March 2017.

Geng E et al. Estimation of mortality among HIV-infected people on antiretroviral treatment in east Africa: a sampling based approach in an observational, multisite, cohort study. Lancet HIV 2: e107-16, 2015. (Full-text article available here).

Community Consensus Statement on Access to HIV Treatment and its Use for Prevention

Together, we can make it happen

We can end HIV soon if people have equal access to HIV drugs as treatment and as PrEP, and have free choice over whether to take them.

Launched today, the Community Consensus Statement is a basic set of principles aimed at making sure that happens.

The Community Consensus Statement is a joint initiative of AVAC, EATG, MSMGF, GNP+, HIV i-Base, the International HIV/AIDS Alliance, ITPC and NAM/aidsmap
close

This content was checked for accuracy at the time it was written. It may have been superseded by more recent developments. NAM recommends checking whether this is the most current information when making decisions that may affect your health.

NAM’s information is intended to support, rather than replace, consultation with a healthcare professional. Talk to your doctor or another member of your healthcare team for advice tailored to your situation.