Prognostic value of C-reactive protein levels in pulmonary infections: A systematic review and meta-analysis.

BACKGROUND

C-reactive rotein (CRP) has been extensively studied as a biomarker that can predict mortality in patients with acute lung disease and our study aimed to elucidate the prognostic value of CRP levels for mortality in patients with various airway diseases, accounting for these differences and potential confounding factors accounts.

METHODS

An extensive literature search was conducted in several databases including PubMed, Embase, Web of Science, Scopus, and ProQuest to ensure the inclusion of up-to-date evidence from studies published between January 2019 and December 2024. Both fixed-effects and random-effects models were used to calculate pooled mean hazard ratios (HR) and odds ratios (OR) for mortality.

RESULTS

For mortality, the fixed effects model revealed a HR of 1.0065 (95% CI: 1.0054-1.0075, P < .0001), indicating a slightly increased risk of death associated with higher CRP levels. However, the random effects model, considering study heterogeneity, suggested an HR of 1.0488 (95% CI: 0.9978-1.1024, P = .0608), with significant heterogeneity (Q = 135.31, P < .0001). The OR analysis under the random effects model showed a more substantial increase in mortality risk with an OR of 1.2033 (95% CI: 1.0635-1.3614, P = .0033). Regarding ICU admissions and ventilation needs, substantial heterogeneity was also observed. The analysis did not find a statistically significant association between elevated CRP levels and ICU admission (OR = 1.1108, 95% CI: 0.9604-1.2847, P = .1568) or the necessity for ventilation (OR = 1.8981, 95% CI: 0.9651-3.7331, P = .0633), although both indicated trends towards increased risk.

CONCLUSION

CRP levels show a potential yet inconsistent association with mortality risk in patients with pulmonary infections. While elevated CRP levels suggest an increased risk of mortality, the results should be interpreted cautiously due to potential overestimation of the effect and the presence of publication bias.