LAUSR

Postoperative morbidity and mortality account for billions of dollars in health care expenditure and millions of additional patient days in the hospital.1 Use of hydroxymethyl glutaryl coenzyme A reductase inhibitors (statins) in the perioperative period offers beneficial effects beyond those afforded by lowering of lipid levels. It is hypothesized that the pleiotropic effects of statins, which can alter endothelial function and decrease inflammation, may be the driver for the beneficial effects of statins in patients who have undergone surgery.2,3 For a variety of reasons not limited to controversy over the academic integrity of randomized clinical trials of perioperative statins (DECRASE [Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography] III and IV) as well as the reliance on retrospective data for the beneficial effects of perioperative statins, the routine use of perioperative statins to decrease complications remains an area of clinical equipoise.4,5 In addition, the optimal timing of statin therapy initiation and the duration of statin exposure before surgery is not well characterized. In this issue of JAMA Internal Medicine, London et al6 present excellent epidemiologic data suggesting that statins, even when administered early in the perioperative period, have protective effects for patients undergoing noncardiac surgery. Perioperative risk, outcome data, and type of surgery were derived from the Veterans Affairs Surgical Quality Improvement Program (VASQIP) database spanning from October 1, 2005, through September 30, 2010. Statins were classified according to American College of Cardiology/American Heart Association (ACC/AHA) guidelines with respect to type and intensity of medication. The percentage of days for which the prescription was available, the maximum dosage, and whether the prescription was active before surgery were obtained from prescription data. Statin exposure on the day of and/or the day after surgery was related to the study’s primary outcome of death from any cause within 30 days of surgery. Secondary outcomes analyzed were cardiac events, central nervous system events, thrombotic events, infection, respiratory events, renal failure or insufficiency, and a composite of these outcomes. A 1:1 propensity score–matched analysis was used to control for treatment-selection bias. Patients were categorized by patterns of statin use as continuous statin users, withdrawn users (active prescription on admission but no use during hospitalization), nonusers, and new statin users (no prescription on admission but prescription during inpatient admission). The final cohort consisted of 180 478 patients at 104 Veteran Affairs hospitals. Overall, 31.5% of the cohort was exposed to a statin on the day of or the day after surgery. Exposure rates to statins differed significantly among surgical subgroups, including 50.1% for vascular surgery, 18.6% for general surgery, 38.8% for orthopedic surgery, and 29.7% in the aggregated remainder (P < .001). Of patients with an active outpatient prescription for a statin, 62.0% were exposed to a statin on the day of or the day after surgery. The most commonly prescribed statin was simvastatin. Of patients without an active statin prescription on admission, 13.1% were exposed to a statin on the day of or the day after surgery. The primary end point, 30-day allcause mortality, was 2.2% for the entire cohort, which varied significantly by type of surgery: 2.6% for vascular surgery, 3.2% for general surgery, 1.5% for orthopedic surgery, and 1.8% for the aggregated remaining specialties. Propensity score– matched analyses for the primary outcome of all-cause 30day mortality showed an association of statin exposure with reduced 30-day all-cause mortality (relative risk, 0.82; 95% CI, 0.75-0.89; number needed to treat, 244; 95% CI, 170-432). Logistic regression analyzing statin dose intensity showed that mediumor high-intensity statin dosing was associated with improved 30-day all-cause mortality relative to low-intensity statin dosing (odds ratios [ORs], 1.03 for low-intensity, 0.84 for medium-intensity, and 0.79 for high-intensity doses). The authors found no evidence of significant interactions of surgery type, revised cardiac risk index score, age, presence of ischemic heart disease, heart failure, and renal insufficiency with statin exposure on the risk for 30-day all-cause mortality. Analysis of secondary outcomes showed significant reductions in cardiac complications, infection, renal complications, and respiratory complications (number needed to treat for the aggregate of anycomplication, 67).Comparedwithpatientswithnostatinuse, there was a significant association of high-intensity statin dosing with more renal injury in the new statin user group (OR, 1.18; 95% CI, 1.02-1.37; P = .03), and a borderline association (OR, 1.15; 95% CI, 1.00-1.33; P = .054) for any statin use in the 90 days before surgery. Finally, 38.0% of patients with an active outpatient prescription were not being exposed to a statin on the day of and/ or the day after surgery. London et al6 are to be commended on a thorough and important contribution to the existing body of data investigating the role of perioperative statin use on mortality and complications after major noncardiac surgery. The data presented in this large epidemiologic analysis suggest that statin use on the day of and/or the day after noncardiac surgery is associated with lower 30-day all-cause mortality and a decrease in a number of clinically significant postoperative complications compared with nonuse. The work gives important insights to real-world statin use in patients undergoing highrisk noncardiac surgery. The rate of statin use continues to increase, with an increase in higher-intensity statin dosing, which likely will only continue given newly available generic Related article Perioperative Statin Use in Noncardiac Surgery Invited Commentary