Previous studies indicate that pre‐admission glucocorticoids increase the risk of perioperative complications.

To examine whether pre‐admission use of glucocorticoids affects 30‐day mortality after colorectal cancer (CRC) surgery.

We conducted a nationwide population‐based cohort study by linking Danish medical registries. All residents in Denmark who underwent CRC surgery from 2001 to 2011 were included. We characterised subjects who filled their most recent glucocorticoid prescription ≤90, 91–365 and >365 days before their surgery date as prevalent, recent and former users, respectively. Prevalent users were subgrouped into new (first‐ever prescription ≤90 days before surgery date) and continuing users. We estimated 30‐day cumulative mortality by the Kaplan–Meier method and corresponding mortality rate ratios (MRRs) using Cox proportional hazard regression, adjusting for potential confounders.

Of the 34 641 CRC patients included, 3966 (11.5%) had filled one or more prescriptions of glucocorticoids within the year before the surgery date. Thirty‐day mortality among prevalent users of oral glucocorticoids was 15.0% vs. 7.3% among non‐users [MRR = 1.28; 95% confidence interval (CI): 1.03, 1.58]. Among new users, the 30‐day mortality was 17.8% (MRR = 1.92; 95% CI: 1.30, 2.83) while it was 14.2% among continuing users (MRR = 1.13; 95% CI: 0.88, 1.44). No associations were found for recent or former use of oral glucocorticoids nor for use of inhaled, intestinal‐acting, and mixed glucocorticoids.

Prevalent use, particulary new use, of oral glucocorticoids was associated with markedly increased 30‐day mortality after colorectal cancer surgery compared to patients not exposed to any glucocorticoids.