Differentiation between post-operative inflammation and bacterial infection remains an important issue in infants following congenital heart surgery. We primarily assessed kinetics and predictive value of C-reactive protein for bacterial infection in the early (days 0–4) and late (days 5–28) period after cardiopulmonary bypass surgery. Secondary objectives were frequency, type, and timing of post-operative infection related to the risk adjustment for congenital heart surgery score.

This 3-year single-centre retrospective cohort study in a paediatric cardiac ICU analysed 191 infants accounting for 235 episodes of CPBP surgery. Primary outcome was kinetics of CRP in the first 28 days after CPBP surgery in infected and non-infected patients.

We observed 22 infectious episodes in the early and 34 in the late post-operative period. CRP kinetics in the early post-operative period did not accurately differentiate between infected and non-infected patients. In the late post-operative period, infected infants displayed significantly higher CRP values with a median of 7.91 (1.64–22.02) and 6.92 mg/dl (1.92–19.65) on days 2 and 3 compared to 4.02 (1.99–15.9) and 3.72 mg/dl (1.08–9.72) in the non-infection group. Combining CRP on days 2 and 3 after suspicion of infection revealed a cut-off of 9.47 mg/L with an acceptable predictive accuracy of 76%.

In neonates and infants, CRP kinetics is not useful to predict infection in the first 72 hours after CPBP surgery due to the inflammatory response. However, in the late post-operative period, CRP is a valuable adjunctive diagnostic test in conjunction with clinical presentation and microbiological diagnostics.