Studies of relative risks of AIDS and/or death associated with higher RNA viral load or lower CD4 cell count, in populations of HIV-1 infected adults.

Median durations of follow-up across studies and datapoints were 4.8 years for AIDS depending on RNA; 6.3 years for death depending on RNA; 4.9 years for AIDS depending on CD4, and 4.6 years for death depending on CD4.
ARV = antiretroviral; MSM = men having sex with men; IDU = injection drug users; F = women; M = men; RR = relative risk; SC = seroconverter or seroconverting.
*Unadjusted, unless indicated.
#For studies reporting relative risks per quartile or tertile, the value stated is the average relative risk over the quartiles (Q2/Q1, Q3/Q2, Q4/Q3) or tertiles (T2/T1, T3/T2).
$3 studies that reported a relative risk as ‘not significant’ without specifying the risk value were entered at a value of 1.0. Omission of these 3 datapoints did not significantly alter the qualitative or quantitative results or conclusions (not shown).
£HIV infection/disease stage at baseline RNA or CD4 measurement, as the median year after SC in the study population.
££Baseline year not reported, but imputed from baseline CD4 level by extrapolation of the linear relationship between baseline CD4 and baseline year from the 14 studies that reported both (Pearson's R2 = 0.72, p<0.001).
+Median duration of patient follow-up, in years.
&For studies reporting only the proportion of participants who received any antiretroviral mono- or bi-therapy during follow-up but not the proportion of follow-up duration affected by therapy, we estimated the latter as half the proportion of participants who received any therapy.
♣RRs similar when including only women who had not received ART, by censoring follow-up at ART initiation, or by including ART use as time-varying covariate.
♦Baseline RNA levels similar in subgroups that did and did not receive ART during f-up; RNA predicted time to AIDS and death independent of subsequent ART.