Diagnostic Utility of WHO Defined Sepsis Syndromic Criteria with Procalcitonin in Identification of Sepsis in Children with Suspected Infections

Date of Award


Document Type


Degree Name

Master of Medicine (MMed)

First Supervisor/Advisor

Dr Doris Kinuthia

Second Supervisor/Advisor

Prof William Macharia


Paediatrics and Child Health (East Africa)


Title: Diagnostic Utility of WHO Defined Sepsis syndromic Criteria with Procalcitonin in Identification of Sepsis in Children with Suspected Infections

Background: Bacterial infections are a major cause of morbidity and mortality in children but accurate and timely diagnosis remains a challenge to the frontline clinicians. Cultures of sterile bodily specimens are considered the gold standard but are insensitive and results are delayed. Procalcitonin (PCT) has recently emerged as a biomarker with high sensitivity and specificity making it superior to culture based techniques performed in non-ideal circumstances which prevail in resource poor settings. This study investigated the diagnostic utility World Health Organization (WHO) defined criteria for sepsis in diagnosis of bacterial sepsis using procalcitonin (PCT) as the “gold standard”.

Methods: This was a prospective survey conducted between July to October 2014 whose primary objective was to determine the sensitivity, specificity, and predictive values of WHO definitions of sepsis syndromes using procalcitonin as the ‘gold standard’ in children with suspected infections. Our secondary objective was to investigate association between PCT elevation and severity of disease and other markers of sepsis. The study was conducted at the paediatrics casualty of the Aga Khan University Hospital, Nairobi (AKUHN) and enrolled children aged >30 days and ≤15 years with suspected infection. This was defined as presence of systemic inflammatory response syndrome (SIRS) (abnormal temperature (>38.5°C or12hours. Children with obvious causes of SIRS such as mechanical/surgical trauma, severe burns, cancer, and without parental/guardian consent were excluded. A standard clinical proforma that allowed classification into various WHO clinical sepsis syndromes for enrolled patients was fulfilled by attending clinician. Blood sample of 0.5ml was collected and PCT test done. Sensitivity (Sp), specificity (Sn), negative and positive predictive values (NPV, PPV), and positive and negative likelihood ratios (LR+, LR-) were calculated for the WHO syndromic definitions. The gold standard for bacterial sepsis was defined as PCT level ≥0.5μg/l. Logistic regression was done and odds ratios calculated to test for association between procalcitonin and disease severity/clinical signs. Kaplan Meier survival analysis was done to compare duration of hospitalisation in various sub-groups.

Results: A total of 231 children were enrolled and 22% (51 children) fulfilled criteria for WHO sepsis syndrome. WHO sepsis syndromes definitions had low Sn, 56.9% (95% CI; 50.5%, 63.3%) and Sp, 66.7% (95% CI; 60.6%, 72.8%) for diagnosis of sepsis. PPV was 32.6% (95% CI; 26.5%, 38.6%) and NPV was 84.5% (95% CI; 79.8% 89.2%). Presence of WHO sepsis syndrome increased the likelihood of having sepsis (PCT ≥0.5ng/ml), LR+= 2.1 (95% CI: 1.3, 3.4) but absence of the sepsis syndromes did not, however, lower probability of exclusion of sepsis, LR-= 0.9 (95% CI: 0.7 to 1.0). Individual clinical signs were poorer at identifying children with sepsis. Children with severe pneumonia had higher mean procalcitonin levels compared to those with mild pneumonia, mean±SE, 19.4± 7.5ng/ml versus 1.9±0.5 ng/ml, respectively, p=0.007; and the levels were higher in patients with more multiple clinical signs as compared one sign (Score test for trend of odds; X2=4.4, p=0.04). There was a trend towards longer hospital stay for patients defined to have bacterial sepsis (PCT≥0.5ng/ml) in the subgroup without WHO sepsis syndrome (logrank, p=0.23) but no similar trend in group with WHO sepsis syndromes (logrank, p=0.23). Antibiotics administration were more likely to have been prescribed in those with WHO sepsis than in for those without, odds ratio=5.8 (95% CI: 2.4, 14.2; p

Conclusions: Overall, WHO syndromic definition is inaccurate for diagnosis of bacterial sepsis. There is need to investigate if diagnostic utility would be superior in those with more severe or established sepsis.

This document is available in the relevant AKU library