We report a recently discovered laboratory error in the busulfan assay described in the above paper. This error resulted in the reporting of busulfan concentrations up to four times higher than actual in some cases. The calculated areas under the busulfan concentration vs time curve (AUCs) were therefore falsely elevated and the dose reductions made to achieve the target AUC were more profound than they should have been. The proportion of results affected by this error is unknown, and therefore the busulfan doses we recommended may be lower than actually required to achieve the intended busulfan dose intensity.