Time-varying mortality risk after gastrointestinal surgery complicated by postoperative infections: a Danish...

Postoperative infections are associated with increased mortality risk, but it is unclear if this risk increase persists over time. This study aims to estimate mortality risk associated with postoperative infections at different time-points within the first postoperative year in a nationwide cohort of gastrointestinal surgery patients.

We included all individuals residing in Denmark who underwent gastrointestinal surgery between 1996 and 2018 and were alive at postoperative day 30. For different time-intervals during the one-year follow-up, we calculated mortality rates and cumulative incidences of death for patients with and without 30-day postoperative infections. Time-varying Cox regression analyses estimated the relative mortality risk associated with postoperative infection exposure.

Of 859,766 patients (female:49.2%; median age:51 years), 25,126 (2.9%) had at least one 30-day postoperative infection. In patients with or without infections, cumulative incidences of death between postoperative days 30-365 were 13.5% versus 4.7%. Adjusted hazard ratios (HRs) for mortality from postoperative days 30, 91, 181, and 271 until end of follow-up (until postoperative day 365) were 2.25(95% CI:2.13-2.38), 1.88(1.74-2.04), 1.44(1.29-1.62), and 1.11(1.00-1.28) for any postoperative infection compared to no infection (p_{time-interaction}<0.001). The adjusted HRs for mortality for postoperative days 30-365 in patients exposed to different postoperative infection types were: sepsis: 4.38(3.90-4.93), pneumonia: 2.60(2.37-2.85), urinary tract infection: 1.26(1.05-1.52), surgical site infection: 1.16(1.04-1.30).

Compared to patients with no infection, patients exposed to 30-day postoperative infections after gastrointestinal surgery had a 2.3-fold risk of mortality at postoperative days 30, and the relative risk of mortality attributed to infection exposure gradually diminished over time.