Introduction: C-reactive protein (CRP) is a real-time and low-cost biomarker to distinguish febrile bacterial infections from non-bacterial febrile illnesses. We hypothesised that measuring the velocity of the biomarker instead of its absolute serum concentration could enhance its ability to differentiate between these two conditions.
Methods: We prospectively recruited adult patients (age >or= 18 years) who presented to the emergency department with fever. We recorded their data regarding the onset of fever and accompanying symptoms. CRP measurements were obtained upon admission. CRP velocity (CRPv) was defined as the ratio between CRP on admission and the number of hours since the onset of fever. Patients were diagnosed by clinical symptoms, blood cultures and imaging studies, and the diagnoses were confirmed by an infectious disease specialist. The efficacy of CRPv as a diagnostic marker was evaluated by using receiver operator curves (ROC). Excluded were patients who did not know the time fever started with certainty, patients with malignancy, patients with HIV infection and patients who had been using antibiotics upon presentation.
Results: Of 178 eligible patients, 108 (60.7%) had febrile bacterial infections (mean CRP: 63.77 mg/L, mean CRPv: 3.61 mg/L/hour) and 70 (39.3%) had non-bacterial febrile illnesses (mean CRP: 23.2 mg/L, mean CRPv: 0.41 mg/L/hour). The area under the curve for CRP and CRPv were 0.783 (95% confidence interval (CI) = 0.717 to 0.850) and 0.871 (95% CI = 0.817 to 0.924), respectively. In a 122-patient subgroup with a CRP level of less than 100 mg/L, the area under the curve increased from 0.689 (95% CI = 0.0595 to 0.782) to 0.842 (95% CI = 0.77 to 0.914) by using the CRPv measurements.
Conclusions: CRPv improved differentiation between febrile bacterial infections and non-bacterial febrile illnesses compared with CRP alone, and could identify individuals who need prompt therapeutic intervention.