Early versus delayed switching from controlled to assisted ventilation: a validation study

To the Editor: Using data from the prospective multicenter WEAN SAFE ­cohort,1 we recently reported that early switching from controlled to assisted ventilation (within 1 day of first morning eligibility) was associated with earlier successful extubation among patients with PaO2/FiO2 ratio >150 mm Hg receiving invasive mechanical ventilation for at least 2 days.2 To assess the robustness and validate these findings, we conducted analyses using the Medical Information Mart for Intensive Care (MIMIC)-IV database (v3.1),3 an independent single-center electronic health record dataset. First, we replicated the original analysis to evaluate whether the findings hold in a single-center setting, thereby addressing concerns that site-level differences could have influenced the results. Second, we refined the analysis to address concerns about misclassification related to the accuracy of identifying eligibility moments. While the once-daily morning data collection in WEAN SAFE was practical and aligned with common weaning practices, switching to assisted mode can be implemented more flexibly, with decisions often occurring outside scheduled morning assessments. Restricting eligibility assessments to mornings may therefore miss important eligibility moments, potentially leading to patients being assigned to the wrong treatment group, as previously suggested by Zhu et al.4 To address this concern, we used the high temporal resolution of MIMIC-IV to more precisely identify the first moment of eligibility. By moving beyond once-daily assessments, we redefined switching strategies using narrower time windows (12 and 6 hours, in addition to 24 hours), ensuring that early switches more accurately align with the moment of eligibility.