The Pitt Bacteremia Score Predicts Mortality in Non-Bacteremic Infections.

CLINICAL INFECTIOUS DISEASES(2020)

引用 56|浏览50
暂无评分
摘要
Background: Predicting mortality risk in patients is important in research settings. The Pitt bacteremia score (PBS) is commonly used as a predictor of early mortality risk in patients with bloodstream infections (BSIs). We determined whether the PBS predicts 14-day inpatient mortality in nonbacteremia carbapenem-resistant Enterobacteriaceae (CRE) infections. Methods: Patients were selected from the Consortium on Resistance Against Carbapenems in Klebsiella and Other Enterobacteriaceae, a prospective, multicenter, observational study. We estimated risk ratios to analyze the predictive ability of the PBS overall and each of its components individually. We analyzed each component of the PBS in the prediction of mortality, assessed the appropriate cutoff value for the dichotomized score, and compared the predictive ability of the qPitt score to that of the PBS. Results: In a cohort of 475 patients with CRE infections, a PBS >= 4 was associated with mortality in patients with nonbacteremia infections (risk ratio [RR], 21.9; 95% confidence interval [CI], 7.0, 68.8) and with BSIs (RR, 6.0; 95% CI, 2.5, 14.4). In multivariable analysis, the hypotension, mechanical ventilation, mental status, and cardiac arrest parameters of the PBS were independent risk factors for 14-day all-cause inpatient mortality. The temperature parameter as originally calculated for the PBS was not independently associated with mortality. However, a temperature <36.0 degrees C vs >= 36 degrees C was independently associated with mortality. A qPitt score >= 2 had similar discrimination as a PBS >= 4 in nonbacteremia infections. Conclusions: Here, we validated that the PBS and qPitt score can be used as reliable predictors of mortality in nonbacteremia CRE infections.
更多
查看译文
关键词
Pitt bacteremia score,Carbapenem-resistant Enterobacteriaceae,Klebsiella pneumoniae,mortality,risk score
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要