Abstract

Background: It is very crucial to recognize infection in immunocompromised patients. The purpose of this study was to explore the diagnostic accuracy of C-reactive protein (CRP) in critically ill immunocompromised patients with sepsis.

Aims and Objective: To find out the diagnostic utility of CRP in immunocompromised patients with sepsis.

Materials and Methods: This was a cross-sectional study, which included immunocompromised patients with suspected sepsis. Patients were classified into two diagnostic groups: those with nonbacterial sepsis and those with bacterial sepsis, and the values of CRP were estimated.

Results: Of 94 patients (63 men and 31 women) with a median age of 56 years (95% CI 53.9–59.3), 74 (78.5%) had immunosuppression with nonbacterial sepsis and 20 (21.4%) had immunosuppression with bacterial sepsis. CRP concentrations were higher in the group with bacterial sepsis [30.94 ng/ml (95% CI 25.13–36.74)] than those with nonbacterial sepsis [7.46 ng/ml (95% CI 7.05–7.87), P < 0.0001]. CRP concentrations that were >6 mg/L had 93.33% sensitivity but only 63.20% specificity for diagnosing sepsis. The accuracy of diagnosis was 87.23%. The area under the receiver-operating characteristic curve was 0.82 (0.72–0.92).

Conclusion: Despite limited specificity in critically ill immunocompromised patients, CRP concentrations may help to rule out bacterial infection.

Key words: C-reactive protein; Immune deficiency; Bacterial sepsis; Non-bacterial sepsis