ABSTRACT 57

Abstract

Background and aims: The effects of peri-operative blood stream infection (PBSI) on outcome in young infants having congenital heart disease (CHD) surgery is unknown. Aims: Determine neurocognitive outcomes after PBSI. Methods: After IRB approval, all consecutive infants <2mo old having cardiac surgery with cardiopulmonary bypass from Sept/1996-Feb/2009 were included. Peri-operative variables were collected prospectively by the Complex Pediatric Therapies Follow-up Program. Sepsis was defined as a positive blood culture requiring >5d of antibiotics; potential contaminants required two positive cultures or hemodynamic changes. Outcomes at 4.5yr follow-up were: verbal intelligence quotient (VIQ), performance IQ (PIQ), and full-scale IQ (FSIQ) on the WPPSI. Associations of sepsis with outcomes were determined by univariate analyses (t-test), and stepwise multiple regressions adjusting for single ventricle (SV), chromosomal abnormality, and extracorporeal membrane oxygenation (ECMO). Results: Of 502 patients, 315 (63%) were male, 160 (32%) had SV anatomy, 51 (10%) had ECMO, and 45 (9%) had chromosomal abnormality. PBSI occurred in 97/502 (19%) overall, and 76/396 (19%) of 4yr survivors. PBSI was not associated with 30d (31/502, 6%), 2yr (86/502, 17%), or 4yr (92/502, 18%) mortality. At 4.5yr, 396 (97% of survivors) had follow-up completed. PBSI was associated with PIQ [84.8 (17.5) vs 92.8 (19.3), p=0.001], VIQ [91.9 (18.6) vs. 85.8 (17.7), p=0.010], and FSIQ [91.9 (18.7) vs. 84.8 (17.5), p=0.003] scores. On stepwise multiple regressions, sepsis was independently associated with PIQ [effect size -6.4; 95%CI -10.9,-1.9; p=0.005], VIQ [effect size -5.8; 95%CI -10.2,-1.4; p=0.01], and FSIQ [effect size -5.8; 95%CI -10.2,-1.5; p=0.009]. Conclusions: There is a significant independent association between the potentially modifiable variable PBSI and neurocognitive outcomes in young infants having surgery for CHD.