Diagnostic accuracy of procalcitonin for bacterial pneumonia in children – a systematic review and meta-analysis

<b>Objective:</b> The predictive role of procalcitonin for childhood bacterial pneumonia, a leading cause of death, is unclear. We aimed to evaluate the diagnostic accuracy of procalcitonin for childhood bacterial pneumonia. <b>Methods:</b> Major bibliographic databases were searched from inception through September 2019 using pre-defined index terms, including ‘procalcitonin,’ ‘pneumonia’ and ‘children’. The study is reported according to <i>Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies</i>. Meta-analyses of the diagnostic accuracy and odds ratio of procalcitonin for bacterial pneumonia were conducted along with subgroup analyses for different cut-offs of procalcitonin. <i>The Quality Assessment of Diagnostic Accuracy Studies 2</i> instrument was used to assess the methodologic quality of eligible studies. <b>Results:</b> Twenty-five studies (with 2,864 patients) showed that procalcitonin for bacterial pneumonia had an overall sensitivity of 0.64 (95% confidence interval: 0.53–0.74), specificity of 0.72 (95% confidence interval: 0.64–0.79), positive likelihood ratio of 2.3 (95% confidence interval: 1.8–3.0) and negative likelihood ratio of 0.50 (95% confidence interval: 0.38–0.66), and Area Under the Receiver Operating Characteristics of 0.74 (95% confidence interval: 0.70–0.78). Using a cut-off of 0.5 ng/ml, Procalcitonin had a sensitivity of 0.68 (95% confidence interval: 0.50–0.82), specificity of 0.60 (95% confidence interval: 0.47–0.72), and Area Under the Receiver Operating Characteristics of 0.68 (95% confidence interval: 0.64–0.72). Using a cut-off of 2 ng/ml, procalcitonin had a sensitivity of 0.59 (95% confidence interval: 0.40–0.76), specificity of 0.71 (95% confidence interval: 0.58–0.81), and AUROC curve of 0.71 (95% confidence interval: 0.67–0.75). Elevated procalcitonin was not associated with increased odds of bacterial pneumonia (odds ratio: 1.36, 95% confidence interval: 0.81–1.92, <i>p</i> = .18). Quality assessment found minimal concerns for bias or applicability. <b>Conclusions:</b> Given the moderate diagnostic accuracy of procalcitonin for bacterial pneumonia, we recommend that procalcitonin be used in conjunction with other findings for management and disposition of children with pneumonia.