C-reactive protein, an early marker of community-acquired sepsis resolution: a multi-center prospective observational study

Abstract

Introduction: C-reactive protein (CRP) has been shown to be a valuable marker in the diagnosis of infection and in monitoring its response to antibiotics. Our objective was to evaluate serial CRP measurements after prescription of antibiotics to describe the clinical course of Community-Acquired Sepsis admitted to intensive care units (ICU).

Methods: During a 12-month period a multi-center, prospective, observational study was conducted, segregating adults with Community-Acquired Sepsis. Patients were followed-up during the first five ICU days, day of ICU discharge or death and hospital outcome. CRP-ratio was calculated in relation to Day 1 CRP concentration. Patients were classified according to the pattern of CRP-ratio response to antibiotics: fast response if Day 5 CRP-ratio was < 0.4, slow response if Day 5 CRP-ratio was between 0.4 and 0.8, and no response if Day 5 CRP-ratio was > 0.8. Comparison between survivors and non-survivors was performed.

Results: A total of 891 patients (age 60 ± 17 yrs, hospital mortality 38%) were studied. There were no significant differences between the CRP of survivors and non-survivors until Day 2 of antibiotic therapy. On the following three days, CRP of survivors was significantly lower (P < 0.001). After adjusting for the Simplified Acute Physiology Score II and severity of sepsis, the CRP course was significantly associated with mortality (ORCRP-ratio = 1.03, confidence interval 95%= (1.02, 1.04), P < 0.001). The hospital mortality of patients with fast response, slow response and no response patterns was 23%, 30% and 41%, respectively (P = 0.001). No responders had a significant increase on the odds of death (OR = 2.5, CI95% = (1.6, 4.0), P < 0.001) when compared with fast responders.

Conclusions: Daily CRP measurements after antibiotic prescription were useful as early as Day 3 in identification of Community-Acquired Sepsis patients with poor outcome. The rate of CRP decline during the first five ICU days was markedly associated with prognosis. The identification of the pattern of CRP-ratio response was useful in the recognition of the individual clinical course.

Figure 1

Individual CRP course for a group of patients during the first five ICU days. Dots indicate the individual observed values at the different time points, from day 1 to day 5, and the dashed lines represent the predictions obtained by a linear mixed model for CRP with random intercept and random slope for time. The intercept (α) describes the initial CRP value and the slope (β) describes the CRP rate of change per day for a specific patient (see text for further explanation). CRP, C-reactive protein; ICU, intensive care unit.

Figure 2

CRP course during the first five ICU days in survivors and nonsurvivors. Observed means of CRP during the first five days in ICU stay for survivors (dashed line) and nonsurvivors (solid line) at hospital discharge. Error bars represent point-wise 95% confidence intervals. (*P < 0.001); CRP, C-reactive protein; ICU, intensive care unit.

Figure 3

Temperature and WBC course during the first five ICU days according to CRP-ratio patterns. Mean temperature and WBC count (log transformed) during the first five days in ICU, according to the different patterns of CRP-ratio response to antibiotics. Error bars represent point-wise 95% confidence intervals. CRP, C-reactive protein; ICU, intensive care unit; WBC, white blood cell.

Figure 4

SOFA score course during the first five ICU days according to CRP-ratio patterns. Mean SOFA score for the first five days in ICU stay, according to the patterns of CRP-ratio response to antibiotics: fast response, slow response and no response. Error bars represent point-wise 95% confidence intervals. CRP, C-reactive protein; ICU; intensive care unit; SOFA, Sequential Organ Failure Assessment.