Response to: defaulting from antiretroviral treatment programmes in sub-Saharan Africa: a problem of definition

By  Dr. Matthew Fox  Professor Sydney Rosen  |  | 


The recent letter to the editors by Grimsrud et al. (2011) in response to our review of rates of retention in antiretroviral treatment programmes in sub-Saharan Africa (Fox & Rosen 2010) raises two concerns that we would like to comment on. First, the authors make the point that there was variability in the definitions of loss to follow-up (LTFU) used by the studies we summarized. This is correct: as is clearly noted in the paper, we did not have access to the primary data used for each study and so relied on the definition of LTFU used by each programme for the primary analysis of their data as reported in their publication. Our methodology, while allowing more variability in definitions than would be desirable in a prospective study, did allow us to take advantage of the wealth of published information to achieve our stated goal of documenting changes in the reported retention rates over time. Fortunately, this strategy is not likely to have had a strong influence on our results. Work by Chi et al. (2010) in Zambia has shown that while the definition of LTFU applied to a cohort can change the estimate of the proportion of subjects in a cohort that are retained (as some subjects will return to care), the specificity of the definition of LTFU was found to be ‡95% for any definition of LTFU of one month or greater, meaning that few subjects listed as being LTFU would actually return to care after being out of care for more than one month.

Publication details

Tropical Medicine & International Health