CD4 and VL at switch to second-line for first-line failure in patients receiving 12-weekly CD4 count monitoring, but no VL monitoring (LCM).

*mostly CD4 declines to just over 100 cells/mm3†65/265 (25%) also had WHO 4 events at the time of switch (plus 1 of the 43 switching for other CD4 reasons)**CD4 3 on day of switch: did not have any previous CD4 3 or 2 previous CD4s 3‡retrospectively on stored plasma††p = 0.95 in a multivariable logistic regression model for VL assayed (yes/no) adjusted for whether switched to second-line before or after 1 Jan 2007 (p = 0.12), whether or not joined second-line studies (p = 0.05), sex (p = 0.89) and age at switch (p = 0.61). More patients switched for other CD4 reasons later in the trial, reflecting wider promotion of other immunological criteria as switch criteria in WHO 2006 guidelines[9]; see Results.‡‡ROC area under the curve = 0.62 (95% CI 0.45–0.79) in 115 patients enrolled in second-line studies, vs 0.78 (95% CI 0.58–0.99) in 72 patients not enrolled in second-line studies.Note: ROC = receiver operating characteristic, see Figure 2. One switch 34 weeks after ART initiation: all others ≥48 weeks.