Higher Risk of Acute Kidney Injury in Sepsis Patients with High PEEP Trajectories: A Study Using the Latent Class Trajectory Model Based on the MIMIC-IV Database

Objective Using the MIMIC database, this study aimed to clarify the link between high positive end-expiratory pressure (PEEP) and acute kidney injury (AKI) in septic patients, identify the characteristics of those with high PEEP, and provide evidence for protecting the kidneys of septic patients on mechanical ventilation. Methods Data of septic patients on mechanical ventilation were extracted from the MIMIC-IV database. Daily mean PEEP values for the first 10 days were calculated. The latent class trajectory model (LCTM) was leveraged to identify PEEP trajectory changes, and the optimal number of trajectories was estimated with the Bayesian information criterion. Binary regression explored the link between different PEEP trajectories and AKI. Results 2142 septic patients on mechanical ventilation were included. LCTM analysis classified 1275 (59.52%) into the low PEEP, 649 (30.30%) into the medium PEEP, and 218 (10.18%) into the high PEEP groups. After confounders were adjusted, compared to the low PEEP group, the odds ratios (ORs) and 95% confidence intervals (CIs) for AKI were 1.43 (95% CI: 1.15, 1.76, P = 0.001 ) in the medium and 1.63 (95% CI: 1.18, 2.25, P = 0.003 ) in the high PEEP groups. Conclusion In septic patients on mechanical ventilation, those with high PEEP have a higher AKI risk. PEEP could be a risk assessment indicator for AKI in critically ill septic patients.