C-reactive protein evaluation in communityacquired pneumonia with comorbid chronic heart failure as criterion of antibiotic prescription

Abstract

To prove that diagnostic algorithm based on additional measurement of serum C-reactive protein (CRP) for administration of systemic antibacterial therapy (ABT) to patients with suspected community-acquired pneumonia (CAP) and concomitant chronic heart failure (CHF) does not influence outcomes of disease. This open, single-center, randomized, prospective, noninferiority study included 160 adult patients with documented functional class II-IV CHF who had been admitted with a preliminary diagnosis of non-severe CAP. Patients were randomized at 1:1 to two groups; group 1 - with additional measurement of CRP (n=80) and group 2 - with the use of routine diagnostic methods (n=80). In group 1, systemic ABT was administered only when serum CRP was >28.5 mg / l (threshold level of the biomarker calculated at the previous stage of the study); group 2 received a standard treatment. Noninferiority test result for both algorithms was evaluated by the number of patients with clinical success on days 12-14 (primary endpoint). Non-inferiority margin was δ=-13.5 %. In addition secondary endpoints (early clinical response on days 3-5; early in-hospital adverse events (development of complications; admission to intensive care unit (ICU); death), death, recurrent CAP or CHF worsening with readmission at 28 day; mortality at 90 and 180 days) were estimated. Standard statistical tools were used for all intergroup comparisons. 76 patients of each group reached the primary endpoint. Systemic ABT was administered to 51 (67.1 %) patients in group 1 and 76 (100 %) patients in group 2 (p<0.05). Both groups were comparable (p>0.05) regarding all endpoints: clinical success, 70 (92.1 %) vs. 69 (90.8 %), Δ=1.3 % (one-sided 97.5 % CI: - 8.25 % for non-inferiority margin δ=-13.5 %); early clinical response, 66 (86.8 %) vs. 68 (89.5 %); admission to ICU, 1 (1.3 %) vs. 1 (1.3 %); development of complications, 20 (26.3 %) vs. 22 (28.9 %); readmission, 5 (6.6 %) vs. 6 (7.9 %); in-hospital mortality, 2 (2.6 %) vs. 1 (1.3 %), mortality at 28 day, 3 (3.9 %) vs. 2 (2.6 %), at 90 day, 5 (6.6 %) vs. 4 (5.3 %), at 180 day, 8 (10.5 %) vs. 9 (11.8 %) cases, respectively. additional measurement of serum CRP in patients with CHF and suspected non-severe CAP was able to reduce rate of systemic ABT administration without outcomes and prognosis worsening.