Prediction of bacteremia and bacterial meningitis among febrile infants aged 28 days or younger

Importance. Fever in the first month of life is often the only sign of life-threatening invasive bacterial infection, specifically bacteremia or bacterial meningitis. Most international guidelines recommend routine lumbar punctures for all febrile infants 28 days or younger to rule out bacterial meningitis. Clinical prediction rules may allow for select testing, but limited information exists on their performance to identify infants at low risk for invasive bacterial infections.

Objective. To evaluate the diagnostic accuracy of the updated Pediatric Emergency Care Applied Research Network (PECARN) prediction rule for identifying febrile infants 28 days or younger with bacteremia or bacterial meningitis.

Design, Setting, and Participants. This pooled analysis of 4 published prospective cohort studies from pediatric emergency departments across 6 countries within the global Pediatric Emergency Research Network included previously healthy, non–ill-appearing, full-term (≥37 weeks’ gestation) infants aged 28 days or younger with a temperature greater than or equal to 38.0 °C who underwent urine, blood, and serum testing.

Exposure. Infants were classified as low risk if they had a negative urinalysis/dipstick test result, serum procalcitonin less than or equal to 0.5 ng/mL, and blood absolute neutrophil count less than or equal to 4000/mm3.

Main Outcomes and Measures. Meta-analytic methods were applied to assess diagnostic accuracy (sensitivity, specificity, and positive and negative predictive values) of the PECARN rule for detection of infants with invasive bacterial infections (bacteremia or bacterial meningitis).

Results. Among 1537 infants 28 days or younger (905 male, 1324 hospitalized, 1080 with lumbar punctures), 69 (4.5%) had invasive bacterial infections, including 11 (0.7%) with bacterial meningitis. Overall, 632 (41.1%) met low-risk criteria. The prediction rule had a sensitivity of 94.2% (95% CI, 85.6%-97.8%), specificity of 41.6% (95% CI, 36.7%-46.7%), positive predictive value of 6.9% (95% CI, 4.8%-9.9%), and negative predictive value of 99.4% (95% CI, 98.1%-99.8%) for invasive bacterial infections. In a secondary analysis of 2531 infants from the 2 US-based cohorts from which the rule was originally derived and the 4 validation cohorts, 96 (3.8%) had invasive bacterial infections, 22 (0.9%) had bacterial meningitis, and 1079 (42.6%) were classified as low risk; rule performance was similar. No infants with bacterial meningitis were misclassified in the primary or secondary analyses.

Conclusions and Relevance. The updated PECARN rule had high sensitivity but lower specificity for identifying febrile infants 28 days or younger with invasive bacterial infections in this study, with no missed cases of bacterial meningitis. These results may support shared decision-making regarding select vs routine use of lumbar puncture among infants classified as being at low risk of invasive bacterial infections.