Summary Points
Labhardt et al have authored a timely article that synthesizes evidence on the effectiveness of same-day antiretroviral therapy (ART) and emphasizes the importance of evaluating whether valid comparisons between studies are possible in light of the statistical analysis approaches employed. Furthermore, the article sheds light on the critical balance between using suitable methodology and reflecting on clinical practice and relevance, as recently discussed elsewhere.
We believe that the following considerations are important to contextualize their study:
The authors identified several potential issues that arose in observational studies comparing patients offered ART on the same day of human immunodeficiency virus (HIV) diagnosis (or first healthcare contact) versus those who initiated treatment later (“rapid ART,” “early ART,” “late ART”). First, selection bias may occur if the study sample does not include all patients testing positive for HIV, but rather only those linked to HIV care or starting treatment. Second, the two comparison groups need to refer to the same patient population to avoid invalid comparisons. Third, immortal time bias can arise as those in the delayed treatment group have, by definition, to stay in care until treatment initiation.