Point-of-care C-reactive protein test results in acute infections in children in primary care: an observational study

Background Acute infections are a common reason for children to consult primary care. Serious infections are rare but differentiating them from self-limiting illnesses remains challenging. This can lead to inappropriate antibiotic prescribing. Point-of-care C-reactive protein testing is used to guide antibiotic prescribing in adults. However, in children its use remains unclear. The purpose of this study was to assess point-of-care CRP test levels with respect to patients' characteristics, care setting, preliminary diagnosis, and management. Methods A prospective observational study was performed in children with an acute infection presenting to ambulatory care in Belgium. Results In this study 8280 cases were analysed, of which 6552 had a point-of-care CRP value available. A total of 276 physicians participated. The median patient age was 1.98 years (IQR 0.97 to 4.17), 37% of children presented to a general practitioner, 33% to a paediatric out-patient clinic, and 30% to the emergency department. A total of 131 different preliminary diagnoses were found, with acute upper airway infection as the most frequent. In 6% (n = 513) patients were diagnosed with a serious infection. The most common serious infection was pneumonia. Antibiotics were prescribed in 28% (n = 2030) of all episodes. The median CRP over all infectious episodes was 10 mg/L (IQR < 5-29). Children below 5 years of age and those presenting to a paediatrician had a higher median CRP. Median CRP in patients with serious infections was 21 mg/L (IQR 6 to 63.5). Pneumonia had a median CRP of 48 mg/L (IQR 13-113). In the episodes with antibiotics prescription, median CRP level was 29 mg/L (IQR 10-58) compared to 7 mg/L (IQR < 5-19) when they were not prescribed. Conclusion A low POC CRP as a standalone tool did not seem to be sufficient to rule out serious infections, but its potential in assessing serious infections could increase when integrated in a clinical decision rule.