Background
Systematic screening for active pulmonary tuberculosis (TB) is recommended for high-risk populations however the lack of an accurate, simple, and low-cost screening test that can be used in high burden areas is a major obstacle to its implementation. We evaluated whether C-reactive protein (CRP) possesses the necessary test characteristics to screen individuals for active pulmonary TB.
Methods
We performed a systematic review and meta-analysis of studies evaluating the diagnostic accuracy of CRP (cut-point of 10 mg/L) for pulmonary TB. We searched four databases for eligible articles published before January 31, 2015 and extracted data for individual studies. We synthesized data separately for outpatient and inpatient studies and generated pooled summary estimates (95% CI) for sensitivity and specificity using random-effects meta-analysis. We performed pre-specified subgroup analyses to determine pooled summary estimates of CRP for diagnosis-seeking vs. screening populations and for patients with and without HIV infection.
Findings
We identified nine unique studies enrolling 1723 patients from the outpatient and inpatient setting. In the outpatient setting, CRP had high sensitivity (93%, 95% CI: 85–97) and moderate specificity (62%, 95% CI: 42–79) for active pulmonary TB. CRP was just as sensitive and specific for active pulmonary TB among patients with confirmed HIV-infection. Among hospitalized patients, specificity of CRP was poor (21%, 95% CI: 6–52).
Interpretation
CRP shows considerable promise as a tool to facilitate systematic screening for active TB, even among PLHIV. CRP-based TB screening should now be studied in other high-risk groups to determine the full impact of this simple and low-cost test.