Patient-, Clinician-, and Institution-level Variation in Intraoperative Antihypertensive Use: A Multicenter Observational Analysis

OBJECTIVE(S) Intraoperative use of antihypertensive medications in cardiac and major noncardiac surgery to achieve hemodynamic stability is important for patient outcomes, yet limited multicenter data exist on the degree of practice variation. DESIGN Retrospective, multicenter observational study. SETTING Sixty-five U.S. hospitals in the Multicenter Perioperative Outcomes Group national registry between 2014 and 2022. PARTICIPANTS Adult patients undergoing cardiac surgery (n = 89,530; 30 hospitals) and major noncardiac surgery (n = 2,644,777; 65 hospitals). INTERVENTIONS None. MEASUREMENTS AND MAIN RESULTS Generalized linear mixed models estimated patient-, clinician-, and institutional-level variation and identified factors associated with antihypertensive infusion administration, defined as >60 cumulative minutes or continued during transport from the operating room. Among cardiac cases, 12,415 (13.9%) patients received an antihypertensive infusion for >60 minutes, most commonly nitroglycerin or nicardipine. Among noncardiac patients, 23,535 (0.89%) received an antihypertensive infusion of >60 minutes, most commonly nicardipine. In the cardiac cohort, median odds ratios for receiving antihypertensive infusions were 2.03 (95% confidence interval (CI) 1.95-2.11) and 2.77 (95% CI 2.13-3.40) between randomly selected clinicians and institutions, respectively (Table 2). Factors associated with antihypertensive infusions in cardiac patients were valve surgery (adjusted odds ratio 2.59, 95% CI 2.29-2.92), emergent procedure (2.46, 2.22-2.72), and maximum intraoperative systolic blood pressure >200 mmHg (2.28, 1.99-2.62). In the noncardiac cohort, median odds ratios for receiving antihypertensive infusions were 1.90 (95% CI 1.86-1.93) and 2.02 (95% CI 1.80-2.24) between randomly selected clinicians and institutions, respectively. Factors most strongly associated with antihypertensive use in major noncardiac surgery were elevated intraoperative systolic blood pressure (>140 mmHg) and hypertension comorbidity (adjusted odds ratio 1.82, 95% CI 1.74-1.89). CONCLUSIONS While patient clinical characteristics were major contributors to antihypertensive infusion use across U.S. institutions, clinician- and institution-level factors also appear to influence practice variation. Such differences highlight the need for evidence-based guidelines and comparative effectiveness research to ground practice, drive quality improvement efforts, and reduce potentially unexplained variation.