Incorporating loss to follow-up in estimates of survival among HIV-infected individuals in sub-Saharan Africa enrolled in antiretroviral therapy programs
Verguet, Séphane.; Lim, S.S.; Murray, C.J.L.; Gakidou, E.; Salomon, J.A.
Journal of Infectious Diseases 207(1): 72-79
2013
ISSN/ISBN: 0022-1899 PMID: 23100567 DOI: 10.1093/infdis/jis635Document Number: 572534
Measuring the survival of human immunodeficiency virus-infected adult patients enrolled in antiretroviral therapy (ART) programs is complicated by short observation periods and loss to follow-up. We synthesized data from treatment cohorts in sub-Saharan Africa to estimate survival over 5 years after initiation of ART. We used data on retention, mortality, and loss to follow-up from 34 cohorts, including a total of 102,306 adult patients from 18 sub-Saharan African countries. These data were augmented by data from 13 sub-Saharan African studies tracking death rates among adult patients who were lost to follow-up (LTFU). We used a Poisson regression model to estimate survival over time, incorporating predicted mortality among LTFU patients. Across studies, the median CD4(+) cell count at ART initiation was 104 cells/mm(3), 65% of patients were female, and the median age was 37 years. Survival at 1 year and 5 years were estimated to be 0.87 (95% confidence interval [CI], 0.72-0.94) and 0.70 (95% CI, 0.36-0.86), respectively, after adjustment for loss to follow-up. The life-years gained by a patient during the 5-year period after starting ART were estimated at 2.1 (95% CI, 1.6-2.3) in the adjusted model, compared with 1.7 (95% CI, 1.1-2.0) if there was 100% mortality among LTFU patients and with 2.4 (1.7-2.7) if there was 0% mortality among LTFU patients. Accounting for loss to follow-up produces substantial changes in the estimated life-years gained during the first 5 years of ART receipt.