Serum Procalcitonin as a Diagnostic and Prognostic Marker in Children With Severe Bacterial Infection: A Prospective Observational Study

TL;DR

This study shows that measuring procalcitonin in blood can better predict severe bacterial infections and outcomes in children compared to traditional markers.

Contribution

The study demonstrates that procalcitonin is more accurate than C-reactive protein for diagnosing and predicting outcomes in pediatric bacterial infections.

Findings

Procalcitonin levels were significantly higher in children who did not survive compared to survivors.

Procalcitonin had a higher accuracy in predicting mortality than C-reactive protein.

Higher procalcitonin levels correlated with longer hospital stays.

Abstract

Background: Early recognition of severe bacterial infection (SBI) in children is critical, yet traditional markers such as C‑reactive protein (CRP) lack adequate accuracy. Serum procalcitonin (PCT) rises rapidly after bacterial insult and may offer both diagnostic and prognostic value.

Objectives: To compare the diagnostic and prognostic performance of day 1 PCT with CRP in children with suspected SBI at a tertiary center in eastern India.

Methods: In this prospective observational study (July 2021 - June 2022), we enrolled 212 children (one month-18 years) admitted to the Department of Pediatrics, Rajendra Institute of Medical Sciences, Ranchi. Serum PCT and CRP were measured within six hours of admission. Outcomes included PICU admission, hospital length of stay (LOS), and in‑hospital mortality. Diagnostic accuracy was assessed with receiver‑operating‑characteristic (ROC) curves; sensitivity and specificity at the optimal PCT cut‑off were calculated with Wilson 95% confidence intervals (CI).

Results: Median day 1 PCT was significantly higher in non‑survivors than survivors (3.96 ng mL⁻¹ vs. 1.22 ng mL⁻¹, p<0.001). For predicting mortality, PCT showed an area under the ROC curve (AUC) of 0.954 (95% CI: 0.90-0.99), markedly superior to CRP (AUC: 0.770, 95% CI: 0.62-0.89). A PCT threshold of 2.0 ng mL⁻¹ yielded 87.5% sensitivity (95% CI: 64.0-96.5) and 81.1% specificity (95% CI: 75.1-86.0) for mortality prediction. Mean LOS increased stepwise across PCT quartiles (Q1: 6.3 days → Q4: 8.8 days, p<0.001).

Conclusions: Day 1 PCT outperforms CRP for early risk‑stratification in pediatric bacterial infection, accurately identifying children who require intensive care and prolonged therapy. Incorporating PCT into admission protocols could enhance antibiotic stewardship and optimize PICU resource allocation, particularly in resource‑limited settings.

Figures