# Identifying the thresholds of C-reactive protein, procalcitonin, and interleukin-6 among children ≤36 months’ old with fever without source at risk of serious bacterial infections: a systematic review and meta-analysis

## TL;DR

This study identifies optimal thresholds for CRP and PCT to detect serious bacterial infections in young children with unexplained fevers.

## Contribution

The paper provides evidence-based thresholds for CRP and PCT to guide clinical decisions in diagnosing serious bacterial infections in children under 36 months.

## Key findings

- A CRP cut-off of 10–20 mg/L showed the highest sensitivity for detecting serious bacterial infections.
- A PCT cut-off of <0.5 ng/mL had the best sensitivity but lowest specificity for identifying serious bacterial infections.
- The best diagnostic performance was observed with CRP 10–20 mg/L and PCT <0.5 ng/mL, with pooled AUCs of 0.84 and 0.816, respectively.

## Abstract

Management of children ≤36 months of age presenting with fever without source remains a challenge because the underlying aetiologies may range from self-limiting viral infections to serious bacterial infections (SBIs) including bacteraemia, urinary tract infection (UTI), pneumonia, bacterial meningitis, osteomyelitis, or septic arthritis. This systematic review was conducted to determine the thresholds at which C-reactive protein (CRP), procalcitonin (PCT), and interleukin-6 (IL-6) can predict SBIs in this population.

We systematically searched electronic databases such as MEDLINE, Cochrane, CINAHL, and Web of Science for studies that evaluated the diagnostic accuracies of CRP, PCT, and IL-6 in detecting SBIs in children ≤36 months of age presenting with fever without source, during the period between November 2013 and November 2023. Area under the summary receiver operating curve (SROC) was calculated by the Rutter and Gatsonis method. I2 was used to quantify study heterogeneity. All tests were two-sided, and a p-value <0.05 was considered statistically significant. This review was registered with PROSPERO, CRD42023439093.

Datasets from 37 studies were included. A CRP cut-off of 10–20 mg/L had the highest pooled sensitivity of 0.75 (95% CI: 0.54–0.89), while a CRP cut-off of >40 mg/L had the highest pooled specificity 0.92 (95% CI: 0.87–0.95). A PCT cut-off of <0.5 ng/mL had the highest pooled sensitivity of 0.7812 (95% CI: 0.59–0.90) but the lowest pooled specificity of 0.69 (95% CI: 0.54–0.81). Based on receiver operating curve (ROC) analysis, a CRP cut-off between 10 and 20 mg/L and a PCT cut-off of <0.5 ng/mL showed the best diagnostic performance with a pooled AUC of 0.84 (95% CI: 0.79–0.90) and 0.816 (95% CI: 0.727–0.928), respectively. Only 1 study for IL-6 reported that a threshold of 20 pg/dL had a sensitivity and specificity of 79.1% and 91.6%, respectively.

A PCT cut-off of 0.5 ng/mL and a CRP cut-off of 10–20 mg/L display the best performance in identifying SBIs in children ≤36 months of age with fever without source.

https://www.crd.york.ac.uk/PROSPERO/view/CRD42023439093, identifier CRD42023439093.