C-reactive protein and procalcitonin in the evaluation of the efficacy of a pneumococcal conjugate vaccine in Gambian children

OBJECTIVE To determine the value of C-reactive protein (CRP) and procalcitonin in the evaluation of the efficacy of a pneumococcal conjugate vaccine in Gambian children. METHODS The investigation was done as a substudy of a phase III pneumococcal conjugate vaccine trial. A pilot study (n = 120) to compare CRP and procalcitonin concentrations in children with pneumonia was undertaken, followed by a larger study of CRP concentrations (n = 1868) obtained from a subsample of children with clinical pneumonia seen during the trial. RESULTS In the pilot study, CRP had a higher sensitivity and specificity for the detection of primary endpoint (radiographic) pneumonia than procalcitonin. In the subsequent study of 1868 episodes of clinical pneumonia, CRP showed moderate sensitivity but poor positive predictive value in identifying primary endpoint pneumonia or pneumococcal bacteraemia. Addition of CRP thresholds of 40, 60 or 120 mg/l to the diagnosis of clinical pneumonia did not give higher estimates of vaccine efficacy or vaccine attributable reduction in incidence than primary endpoint pneumonia. CONCLUSION A combination of a raised CRP concentration with a clinical diagnosis of pneumonia is a more specific endpoint than clinical pneumonia alone but less appropriate than primary endpoint pneumonia as a measure of the impact of a pneumococcal vaccine in a Gambian setting.