Thirty-day mortality risk associated with the postoperative nonresumption of angiotensin-converting enzyme inhibitors: A retrospective study of the veterans affairs healthcare system

BACKGROUND Angiotensin‐converting enzyme inhibitors (ACE‐Is) are a widely used class of cardiovascular medication. However, limited data exist on the risks of postoperative nonresumption of an ACE‐I. OBJECTIVE To evaluate the factors and 30‐day mortality risks associated with the postoperative nonresumption of an ACE‐I. DESIGN A retrospective cohort study. SETTING Veterans Affairs (VA) Healthcare System. PATIENTS A total of 294,505 admissions in 240,978 patients with multiple preoperative prescription refills (>3) for an ACE‐I who underwent inpatient surgery from calendar years 1999 to 2012. INTERVENTION None. MEASUREMENTS We classified surgical admissions based upon the timing of postoperative resumption of an ACE‐I prescription from the day of surgery through postoperative days 0 to 14 and 15 to 30, and collected 30‐day mortality data. We evaluated the relationship between 30‐day mortality and the nonresumption of an ACE‐I from postoperative day 0 to 14 using proportional hazard regression models, adjusting for patient‐ and hospital‐level risk factors. Sensitivity analyses were conducted using more homogeneous subpopulations and propensity score models. RESULTS Twenty‐five percent of our cohort did not resume an ACE‐I during the 14 days following surgery. Nonresumption of an ACE‐I within postoperative day 0 to 14 was independently associated with increased 30‐day mortality (hazard ratio: 3.44; 95% confidence interval: 3.30‐3.60; P