OBJECTIVES: Recruitment maneuvers (RMs) during invasive mechanical ventilation (IMV) for acute hypoxemic respiratory failure are not characterized in pediatrics. We aimed 1) to describe changes in oxygenation, dynamic compliance, and outcome related to RMs; and 2) to identify which patients may benefit from RMs using machine learning. DESIGN: Single center, retrospective cohort study, 2016-2023. SETTING: Quaternary PICU in Montreal, Canada. PATIENTS: We included all IMV patients between May 2016 and May 2023 who received at least one RM. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We identified 479 patients who had a total of 8276 RMs. We studied change in Pao2 to Fio2 (Pao2/Fio2) ratio and dynamic compliance data from before and 90 minutes after the RM. In a multivariable model, baseline Pao2/Fio2 ratio (OR, 0.99 95% CI, 0.99-1.00; p = 0.004) and initial positive end-expiratory pressure (PEEP) level (OR, 1.14 95% CI, 1.00-1.31; p = 0.046) were independently associated with oxygenation response. Recruitment delta in pressure was associated with lower odds of compliance response (OR, 0.93 95% CI, 0.90-0.97; p < 0.001). Patients with PEEP lower than that described as "low" in the PEEP-Fio2 tables, vs. those not in this category, had greater odds of oxygen response (OR, 3.33 95% CI, 1.49-7.73; p = 0.004). We failed to identify an association between improved oxygenation or dynamic compliance vs. not and ventilator-free days, duration of IMV or PICU length of stay, and PICU mortality. CONCLUSIONS: In our 2016-2023 PICU cohort of IMV pediatric cases, we found various associations following RMs. Importantly, there was more than a three-fold greater odds of responsiveness associated with having a lower PEEP relative to Fio2 than that described in the recommended practice tables. These data may help decision-making during IMV when lower PEEP is used.
Beauchamp et al. (Wed,) studied this question.