2019 HSR&D/QUERI National Conference

1122 — Thirty-Day Postoperative Infection and Increased Risk of Long-term Infections

Lead/Presenter: William O'Brien,  COIN - Bedford/Boston
All Authors: O'Brien WJ (VA Boston, CHOIR), Gupta K (VA Boston Dept of Medicine), Itani K (VA Boston Dept of Surgery)

Postoperative wound complications, mainly surgical site infections, have been shown to decrease survival in Veterans by up to 42%. The relationship between 30-day postoperative infections and long-term infections remains unknown. The goal of this study is to determine whether exposure to postoperative infection is a risk factor for subsequent long-term infection.

We assessed for inclusion the patient's first chronological surgery reviewed by the VA Surgical Quality Improvement Program non-cardiac database during 2008-2015. Those with any 30-day prior invasive procedure were excluded. Patients who died or had subsequent surgery within 30 days after the index surgery were disenrolled. We determined whether the patient had any 30-day postoperative infection (exposure group) vs. no 30-day infection (control group). The study outcome was the interval between index surgery and the occurrence of the first infection during postoperative days 31-365. Patients who died before having a long-term infection were censored. We used a propensity score with inverse probability of treatment weighting to reduce selection bias, and a weighted Cox proportional hazards model to estimate the hazard ratio of long-term infection as a function of exposure.

The final sample included 659,486 surgical patients, of whom 23,815 (3.6%) had a 30-day infection and 43,796 (6.6%) had a long-term infection. Mean age was 59.7 (SD 13.6), and 91.7% were male. The most frequent 30-day infections were surgical site infection (40.2%), urinary tract infection (27.5%), pneumonia (14.8%) and bloodstream infection (8.0%). Long-term infection types included UTI (48.7%), skin and soft tissue infection (32.6%), BSI (8.8%), and PNA (5.8%). Patients with a 30-day infection had a higher observed incidence of long-term infection (21.8%) compared with those without 30-day infection (6.1%), and the estimated hazard ratio was 1.18 [95% CI 1.15-1.22].

Controlling for patient baseline characteristics and surgical factors, we found that at any given point during the follow-up period, the incidence rate of long-term infection in the exposure group was 18% higher than that of the control group.

Cost-benefit calculations for surgical infection prevention programs should include the increased risk and costs of long-term infection. Preventive efforts in the first 30 postoperative days may improve long-term patient outcomes.