The association between intraoperative low driving pressure ventilation and perioperative healthcare-associated costs: A retrospective multicenter cohort study

Study objective: A low dynamic driving pressure during mechanical ventilation for general anesthesia has been associated with a lower risk of postoperative respiratory complications (PRC), a key driver of healthcare costs. It is, however, unclear whether maintaining low driving pressure is clinically relevant to measure and contain costs. We hypothesized that a lower dynamic driving pressure is associated with lower costs. Design: Multicenter retrospective cohort study. Setting: Two academic healthcare networks in New York and Massachusetts, USA. Patients: 46,715 adult surgical patients undergoing general anesthesia for non-ambulatory (inpatient and same-day admission) surgery between 2016 and 2021. Interventions: The primary exposure was the median intraoperative dynamic driving pressure. Measurements: The primary outcome was direct perioperative healthcare-associated costs, which were matched with data from the Healthcare Cost and Utilization Project-National Inpatient Sample (HCUP–NIS) to report absolute differences in total costs in United States Dollars (US$). We assessed effect modification by patients' baseline risk of PRC (score for prediction of postoperative respiratory complications [SPORC] ≥ 7) and effect mediation by rates of PRC (including post-extubation saturation < 90%, re-intubation or non-invasive ventilation within 7 days) and other major complications. Main results: The median intraoperative dynamic driving pressure was 17.2cmH₂O (IQR 14.0–21.3cmH₂O). In adjusted analyses, every 5cmH₂O reduction in dynamic driving pressure was associated with a decrease of −0.7% in direct perioperative healthcare-associated costs (95%CI −1.3 to −0.1%; p = 0.020). When a dynamic driving pressure below 15cmH₂O was maintained, -US$340 lower total perioperative healthcare-associated costs were observed (95%CI −US$546 to −US$132; p = 0.001). This association was limited to patients at high baseline risk of PRC (n = 4059; −US$1755;97.5%CI −US$2495 to −US$986; p < 0.001), where lower risks of PRC and other major complications mediated 10.7% and 7.2% of this association (p < 0.001 and p = 0.015, respectively). Conclusions: Intraoperative mechanical ventilation targeting low dynamic driving pressures could be a relevant measure to reduce perioperative healthcare-associated costs in high-risk patients.