Diagnostic performance of the lab-score in predicting severe and invasive bacterial infections in well-appearing young febrile infants

Abstract

Background and objectives: The "Lab-score" combining C-reactive protein, procalcitonin and urine dipstick results has recently been derived and validated as an accurate tool for predicting severe bacterial infections (SBIs) in children with fever without source. We aimed to assess the Lab-score usefulness in predicting SBI, especially invasive bacterial infections (IBIs), in well-appearing infants <3 months with fever without source.

Methods: A multicenter retrospective study was conducted in 7 pediatric emergency departments in Spain and Italy. An SBI was defined as isolation of a bacterial pathogen from urine, blood, cerebrospinal fluid or stools, an IBI as isolation of a bacterial pathogen from blood or cerebrospinal fluid. The diagnostic characteristics of the Lab-score for detection of SBI and IBI were calculated.

Results: An SBI was diagnosed in 287 (28.3%) of 1012 patients and an IBI in 23 (2.1%) of 1098. The positive and negative likelihood ratios of a score ≥3 for SBI prediction were 10.2 (95% confidence interval [CI]: 9.5-10.9) and 0.5 (95% CI: 0.5-0.5), respectively. The area under the receiver operating characteristic curve was 0.83 (95% CI: 0.80-0.86). The same diagnostic accuracy measures for identification of IBI were 4.3 (95% CI: 4-4.6), 0.4 (95% CI: 0.3-0.5) and 0.85 (95% CI: 0.76-0.94), respectively. Use of Lab-score would have resulted in misdiagnosis of 7 (30%) infants with IBI.

Conclusions: In well-appearing infants with fever without source, the Lab-score seems a more useful tool for ruling in, rather than ruling out, SBI. Its accuracy for IBI prediction was unsatisfactory.