Non cardiac surgery after stroke and time course of increased risk

From a recently published Danish nationwide cohort study:

Main Outcomes and Measures Risk of major adverse cardiovascular events (MACE; including ischemic stroke, acute myocardial infarction, and cardiovascular mortality) and all-cause mortality up to 30 days after surgery. Odds ratios (ORs) were calculated by multivariable logistic regression models.

Results Crude incidence rates of MACE among patients with (n = 7137) and without (n = 474 046) prior stroke were 54.4 (95% CI, 49.1-59.9) vs 4.1 (95% CI, 3.9-4.2) per 1000 patients. Compared with patients without stroke, ORs for MACE were 14.23 (95% CI, 11.61-17.45) for stroke less than 3 months prior to surgery, 4.85 (95% CI, 3.32-7.08) for stroke 3 to less than 6 months prior, 3.04 (95% CI, 2.13-4.34) for stroke 6 to less than 12 months prior, and 2.47 (95% CI, 2.07-2.95) for stroke 12 months or more prior. MACE risks were at least as high for low-risk (OR, 9.96; 95% CI, 5.49-18.07 for stroke less than 3 months) and intermediate-risk (OR, 17.12; 95% CI, 13.68-21.42 for stroke less than 3 months) surgery compared with high-risk surgery (OR, 2.97; 95% CI, 0.98-9.01 for stroke less than 3 months) (P = .003 for interaction). Similar patterns were found for 30-day mortality: ORs were 3.07 (95% CI, 2.30-4.09) for stroke less than 3 months prior, 1.97 (95% CI, 1.22-3.19) for stroke 3 to less than 6 months prior, 1.45 (95% CI, 0.95-2.20) for stroke 6 to less than 12 months prior, and 1.46 (95% CI, 1.21-1.77) for stroke 12 months or more prior to surgery compared with patients without stroke. Cubic regression splines performed on the stroke subgroup supported that risk leveled off after 9 months.

Conclusions and Relevance A history of stroke was associated with adverse outcomes following surgery, in particular if time between stroke and surgery was less than 9 months. After 9 months, the associated risk appeared stable yet still increased compared with patients with no stroke. The time dependency of risk may warrant attention in future guidelines.