Monitoring of body temperature, but also of C-reactive protein (CRP) level is performed in infected patients treated with antibiotics. These two parameters having low specificity for any diagnosis, our aim was to evaluate the usefulness of CRP in this context.
Patients and method: A representative sample of patients was randomly extracted from our medical computerized tables. All patients presented community-acquired infections and had at least two CRP level assessments. Kinetics of body temperature and CRP allowed to quantify clinical and biological discrepancy while the patient's chart was studied to determine the etiologies.
Results: Three hundred and ninety-two patients over 942 (42%) were admitted in our department over 2 years, including 147 cases of respiratory infections (37%), 91 of urinary infections (23%), 65 of cellulitis (17%), 70 of primary bacteremia (18%), 19 of digestive infections (5%). Ninety-four percent of the patients had been prescribed antibiotic therapy. We observed a correlation between temperature and CRP in 83% of the patients. Forty-seven percent of patients presented with normalized body temperature and persistently high levels of CRP, which was most of the time related to comorbid conditions. Twenty patients (5%) presented with unexplained persistent fever despite CRP normalization. Therapeutic modifications were mostly observed in the presence of clinicobiological discrepancy: 21% versus 6%, p<0.001.
Discussion: Body temperature and CRP are two parameters leading to comparable information in more than 80% of infected patients receiving specific antibiotic therapy. These clinical and biological discrepancies are associated to a modified antibiotherapy with inconclusive results.