A systematic review of 14 studies and a total of 6599 patients showed CRP as the most useful biomarker of pneumonia 1. LR+ was 2.08, 3.64 and 5.89 and LR- was 0.32, 0.36 and 0.47 for the cutoff value of CRP 20, 50 and 100 mg/l, respectively.
A meta-analysis of 8 studies and 5308 patients studied the added value of CRP in the diagnosis of pneumonia 2. CRP value was added to models including different variables like age, shortness of breath, abnormal lung auscultation, smoking, fever and chronic diseases (heart failure, diabetes, COPD, asthma, immune deficiency, cancer, chronic kidney failure). CRP improved the accuracy to separate those with or without pneumonia in all models.
A systematic review including 8 studies and a total of 2194 patients assessed the value of C-reactive protein (CRP) for the diagnosis of community acquired pneumonia (CAP) 3. The diagnostic utility of CRP at different cut-points was assessed against a reference standard of chest X-ray. The median prevalence of CAP in the studies was 14.6 % (range 5%-89%). At a threshold of less than 20 mg/l, positive likelihood ratio (LR+) was 2.1 (95% CI 1.8 to 2.4) and negative likelihood ratio (LR-) was 0.33 (95% CI 0.25 to 2.4; 6 studies). There was little evidence of heterogeneity in either outcome (p>0.15). Results at the other two CRP cutoffs (HASH(0x2ed5798) 50, > 100 mg/l) were heterogeneous. The pooled LR+ was 5.01 (95% CI 2.96-8.48) and LR- 0.54 (95% CI 0.33-0.91) for those with CRP >100 mg/l.
Primary/Secondary Keywords