Association of Daily Body Temperature, White Blood Cell Count, and C-reactive Protein With Mortality and Persistent Bacteremia in Patients With Staphylococcus Aureus Bacteremia: A Post Hoc Analysis of the CAMERA2 Randomized Clinical Trial

Le résumé fourni par la source

Abstract Introduction Classification of patients with Staphylococcus aureus bacteremia as complicated versus uncomplicated is based on a combination of clinical and microbiologic variables. Whether daily body temperature and common laboratory tests such as C-reactive protein (CRP) and white blood cell (WBC) can improve risk stratification algorithms is unclear. Methods We conducted a post hoc secondary analysis of the CAMERA2 trial, which enrolled hospitalized adult patients with methicillin-resistant S aureus bacteremia and prospectively collected daily body temperature and peripheral blood WBC and CRP. We evaluated the prognostic relevance of each parameter by calculating crude and adjusted odds ratios for 90-day all-cause mortality comparing patients with the abnormal parameter of interest versus those with normal parameters on each day of illness. Results A total of 345 patients were included in this analysis, of whom 63 (18.3%) died within 90 days. Fever (body temperature ≥38.0 °C) was associated with increased odds of 90-day mortality from day 4 and onwards. Fever later in the illness course was associated with higher adjusted odds of mortality (8.78; 95% confidence interval, 2.78–27.7 on day 7 vs adjusted odds ratio 3.70; 95% CI, 1.58–8.67 on day 4). In contrast, CRP and abnormal WBC count did not demonstrate a consistent or temporal association with mortality. Conclusions Persistent fever after 72 hours is associated with increased mortality in patients with methicillin-resistant S aureus bacteremia, supporting recommendations that this should be kept as a criterion for classifying patients as either “high-risk” or “complicated.” Within this dataset, there was limited additional predictive value in WBC or CRP.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.