Broad- Versus Narrow-Spectrum Perioperative Antibiotics and Outcomes in Pediatric Congenital Heart Disease Surgery: Analysis of the Vizient Clinical Data Base

Background: Despite guidelines recommending narrow-spectrum perioperative antibiotics (NSPA) as prophylaxis for most children undergoing congenital heart disease (CHD) surgery, broad-spectrum perioperative antibiotics (BSPA) are variably used, and their impact on postoperative outcomes is poorly understood.

Methods: We used administrative data from U.S. hospitals participating in the Vizient Clinical Data Base. Admissions from 2011 to 2018 containing a qualifying CHD surgery in children 0-17 years old were evaluated for exposure to BSPA versus NSPA. Propensity score-adjusted models were used to compare postoperative length of hospital stay (PLOS) by exposure group, while adjusting for confounders. Secondary outcomes included subsequent antimicrobial treatment and in-hospital mortality.

Results: Among 18 088 eligible encounters from 24 U.S. hospitals, BSPA were given in 21.4% of CHD surgeries, with mean BSPA use varying from 1.7% to 96.1% between centers. PLOS was longer for BSPA-exposed cases (adjusted hazard ratio 0.79; 95% confidence interval [CI]: 0.71-0.89, P < .0001). BSPA was associated with higher adjusted odds of subsequent antimicrobial treatment (odds ratio [OR] 1.24; 95% CI: 1.06-1.48), and there was no significant difference in adjusted mortality between exposure groups (OR 2.06; 95% CI: 1.0-4.31; P = .05). Analyses of subgroups with the most BSPA exposure, including high-complexity procedures and delayed sternal closure, also did not find (but could not exclude) a measurable benefit from BSPA on PLOS.

Conclusions: BSPA use was common in high-risk populations, and varied substantially between centers. Standardizing perioperative antibiotic practices between centers may reduce unnecessary broad-spectrum antibiotic exposure and improve clinical outcomes.