Outcomes of Patients Lost to Follow-up in African Antiretroviral Therapy Programs: Individual Patient Data Meta-analysis

Abstract

Background: Low retention on combination antiretroviral therapy (cART) has emerged as a threat to the Joint United Nations Programme on human immunodeficiency virus (HIV)/AIDS (UNAIDS) 90-90-90 targets. We examined outcomes of patients who started cART but were subsequently lost to follow-up (LTFU) in African treatment programs.

Methods: This was a systematic review and individual patient data meta-analysis of studies that traced patients who were LTFU. Outcomes were analyzed using cumulative incidence functions and proportional hazards models for the competing risks of (i) death, (ii) alive but stopped cART, (iii) silent transfer to other clinics, and (iv) retention on cART.

Results: Nine studies contributed data on 7377 patients who started cART and were subsequently LTFU in sub-Saharan Africa. The median CD4 count at the start of cART was 129 cells/μL. At 4 years after the last clinic visit, 21.8% (95% confidence interval [CI], 20.8%-22.7%) were known to have died, 22.6% (95% CI, 21.6%-23.6%) were alive but had stopped cART, 14.8% (95% CI, 14.0%-15.6%) had transferred to another clinic, 9.2% (95% CI, 8.5%-9.8%) were retained on cART, and 31.6% (95% CI, 30.6%-32.7%) could not been found. Mortality was associated with male sex, more advanced disease, and shorter cART duration; stopping cART with less advanced disease andlonger cART duration; and silent transfer with female sex and less advanced disease.

Conclusions: Mortality in patients LTFU must be considered for unbiased assessments of program outcomes and UNAIDS targets in sub-Saharan Africa. Immediate start of cART and early tracing of patients LTFU should be priorities.

Figure 1.

Flow of identifying eligible studies that contributed individual patient data.

Figure 2.

Cumulative incidence plot of the outcomes in patients lost to follow-up (LTFU), including death, stop of combination antiretroviral therapy (cART), transfer to another clinic, and retention on cART. Cumulative probabilities, together with 95% confidence intervals, are given for each tracing outcome at 1–4 years since the last clinic visit.

Figure 3.

Cumulative incidence plot of outcomes for 5152 patients lost to follow-up (LTFU) after starting combination antiretroviral therapy (cART) (A) and 18285 patients retained in care (B) in 2 treatment programs in Malawi and Mozambique. Cumulative probabilities, together with 95% confidence intervals, are given for each outcome at 1–4 years after cART initiation.