Sanchez‐Conde 2005.

Methods	Participants receiving HAART including stavudine were allocated to either reduce the dose of stavudine or continue with the standard dose
Participants	Participants receiving a stavudine containing regimen in the first semester of 2003 at the Department of Infectous Diseases and Service Pharmacy, Hospital Carlos III, Madrid, Spain HIV‐1 RNA < 50 copies/mL during at least previous 3 months Those receiving hepatitis C therapy with interferon and ribavirin or those receiving interleukin‐2 were excluded
Interventions	Half of the participants were allocated to reduce stavudine dose from 40mg to 30mg BD and the other half continued with 40mg BD Participants continued all other antiretroviral drugs
Outcomes	Plasma HIV‐1 RNA and CD4 T‐cell measurements Virologic failure defined as repeated values of plasma HIV‐1 RNA above 50 copies/mL
Notes	Baseline age not reported Funding source: In part by grants from Fundacion Investigacion y Educacion en SIDA (IES) and Red de Investigacion en SIDA (RIS project 173)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Participants were allocated to either reduce or continue the stavudine dosage. The method used to generate allocation sequence is not described.
Allocation concealment (selection bias)	High risk	The method used to conceal allocation is not described
Blinding (performance bias and detection bias) All outcomes	High risk	Participants were allocated to either reduce their stavudine dose or continue their dose. It is assumed that both participants and providers were not blinded, however, it is unclear if the outcome assessors were blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	92 Participants were included in the trial, 47 in the low dose and 45 in the high dose arm. 41 participants did not complete the 12 month follow‐up period. Loss to follow‐up=6; Initiation of Hepatitic C therapy=6; Drug switch for reasons other than virologic failure=12; Voluntary withdrawal=8. The lack of completion was relatively evenly distributed. Intention‐to‐treat analysis was used with missing=failure
Selective reporting (reporting bias)	High risk	Nine participants are reported to have experienced virologic failure (4 in the low dose group and 5 in the high dose group), however, the HIV‐1 RNA measurements are not reported. The mean CD4 count is reported as not significantly different between the two groups but the mean values are not reported.
Other bias	High risk	Age at study entry not reported