Abstract
Objective: To quantify intraoperative pulmonary arterial catheter (PAC) use during cardiac surgery and identify hospital-, anesthesiologist-, and patient-level factors associated with PAC utilization.
Design: A cross-sectional, observational study using generalized logistic mixed models to examine variations in PAC use.
Setting: Fifty-three US academic hospitals participating in the Multicenter Perioperative Outcomes Group (MPOG) national registry
Participants: 145,343 adult patients undergoing cardiac surgery between January 1, 2016, and December 31, 2022.
Interventions: Receipt of intraoperative PAC, defined by ≥60 minutes of physiologically plausible pulmonary arterial pressures.
Measurements and Main Results: The primary outcome was PAC utilization. Mixed-effects logistic regression quantified fixed-effect predictors, and variation attributable to anesthesiologists and then to anesthesiologists nested within a hospital was characterized using median odds ratio (MOR). Of the 145,343 cardiac surgeries performed across 53 hospitals, 104,626 (72%) included PAC monitoring. PAC use varied widely across hospitals (0-98%) and across anesthesiologists (0-100%). PAC was used most frequently in heart transplants (94%) and lung transplants (87%) and least frequently in pulmonic valve procedures (30%). A patient’s likelihood of receiving a PAC was influenced most strongly by hospital (MOR, 15.00; 95% confidence interval [CI], 8.98-28.32), with substantially less variation attributable to an anesthesiologist within the same hospital (MOR, 1.70; 95% CI, 1.61-1.81).
Conclusions: Intraoperative PAC monitoring is used in nearly three-quarters of cardiac surgeries at US academic centers, with hospital practice pattern the factor most closely associated with PAC utilization.
Bo Zhang
Assistant Professor of Biostatistics
My research interests include design of observational studies, instrumental variables, application of causal inference in medicine and applied statistics in general.