The Incremental Risk of Coronary Stents on Postoperative Adverse Events: A Matched Cohort Study.

Holcomb CN1, Graham LA, Richman JS, Itani KM, Maddox TM, Hawn MT.

Author information

Abstract

OBJECTIVE:

To determine the incremental risk of coronary stents on adverse events in surgical patients and whether it varies over time from stent placement.

BACKGROUND:

Postoperative adverse cardiac events decrease as the time from stent placement increases, but the risk attributable to the stent versus the underlying cardiac disease is uncertain, as prior studies lack a control surgical population.

METHODS:

Data for patients with coronary stents implanted in a VA hospital from 2000 to 2010 were matched with VA Surgical Quality Improvement Program data to identify noncardiac surgery within 24 months of stent placement. Each patient with stent was matched with 2 surgical patients without stent on surgical characteristics and cardiac risk factors. Outcomes of myocardial infarction (MI), revascularization, and death within 30 days after surgery were modeled using logistic regression. Adjusted risk differences between stented and nonstented populations were compared across time after stent placement.

RESULTS:

Adverse cardiac events followed surgery in 531 (5.7%) of the 9391 patients with stent and 680 (3.6%) of the 18,782 patients without stent (P < 0.001). In adjusted models, 30-day postoperative MI (odds ratio = 1.90; 95% confidence interval, 1.57-2.30) and revascularization (odds ratio = 2.03; 95% confidence interval, 1.65-2.50) but not mortality (odds ratio = 0.84; 95% confidence interval, 0.69-1.02) were higher in the stented cohort. Assessing trends over the 2 years after stent placement, the incremental risk for MI decreased from 5% immediately after stent placement to 2% at 1 year and then was no longer significantly elevated. The incremental risk did not vary by stent type.

CONCLUSIONS:

Surgery after coronary stent placement is associated with an approximate 2% absolute risk for postoperative MI but no difference in mortality compared with nonstented matched controls.