Were Procalcitonin Levels Associated with the Diagnosis of Bacterial Infections for Hospitalized Pediatric Patients?

Background: Procalcitonin (PCT) has been suggested as a biomarker not only for diagnosis but for
risk-stratification of bacterial infection in pediatric patients. PCT has been used in adult patient populations to guide management of suspected
infections and particularly to discontinue unnecessary antibiotics. Several studies have demonstrated the utility of PCT to identify bacterial infections in neonates as well as community-acquired pneumonia in older children. The utility of a bacterial-specific biomarker would be in guiding management and limiting unnecessary testing. However, few studies demonstrate the use of PCT in diagnosing serious bacterial
infections or have utilized a procalcitonin- based risk stratification of bacterial infections in a hospitalized pediatric patient population

Methods: With IRB approval, retrospective analysis was done of PCT levels for 130 pediatric
patients admitted to a single center using hospital-based laboratory data from 2018 to
2019. The highest PCT levels during an admission were stratified with respect to the risk for
progression to severe sepsis as low ( <0.50 ng/dL), intermediate (0.50-2.0 ng/dL), and
high (>2.0 ng/dL). Data were analyzed with respect to PCT-based risk stratification. The WBC,
PLT , and CRP levels were categorized as abnormal if WBC was <4 or
>15, PL <150 or >550 and CRP >0.50. Cultures collected during hospitalization (blood [BC], cerebral spinal fluid [CSF], urine,
and wound/abscess) were analyzed. Discharge diagnoses were classified as bacterial infection (sepsis, meningitis,
pneumonia with infusion, UTI, cellulitis, osteomyelitis, and perforated appendicitis), viral syndrome, and other non-infection related conditions.

Results: Of 130 patients, 75 (Group 1) had PCT <0.50, 25 (Group 2) had PCT 0.50-2.0, and
30 had PCT >2.0. Groups were comparable for age, gender, race/ethnicity (Table
1), discharge diagnosis, and prevalence of positive BC and/or CSF. The frequency of CRP>0.50 was higher in Groups 2 and 3 than in Group 1. Controlled models revealed increased likelihood for PCT>0.50 if CRP was > 0.50 and PL was <150 or >550 (Odds ratio [OR], 95%Confidence Interval [95%CI] 2.44, 95%CI 1.35, 4.41 and 2.01, 95%CI 1.15, 2.94), respectively. Patients in Group 1 were less likely than those in Groups 2 and 3 to receive
antibiotics during hospitalization.

Conclusion: Our study revealed the direct correlation of increased PCT with antibiotic treatment
and abnormal CRP and PL but not with diagnosis of bacterial infection in
hospitalized pediatric patients. Further studies are needed to understand the relationship between procalcitonin
and bacterial infections in the hospitalized pediatric population.