Bacteriemia enterocócicaanálisis de una cohorte histórica. Diseño y validación de un modelo predictivo de mortalidad

Resumen

The aim of our study was to analyze the characteristics of the infection in our study population and design a prediction rule to describe the risk of death due to Enterococcus faecalis or E. faecium bloodstream infection. To this end, we developed a prediction rule by analyzing data collected from 122 patients diagnosed with enterococcal BSI admitted to the Clínica Universidad de Navarra (University Clinic of Navarra. Pamplona, Spain); and validated this by confirming its accuracy with the data of a population with enterococcal BSI in the Hospital del Mar, Barcelona. The test results showed that our population, as described by other studies, was characterized by severe underlying diseases, had a high number of comorbidities, had been subjected to instrumental or surgical manipulation and had received prior antibiotic treatment. According to this model, independent significant predictors for the risk of death were: being diabetic, having received appropriate treatment, severe prognosis of the underlying diseases, having renal failure, having received solid organ transplant, malignancy, the particular source of the bloodstream infection, and be immune compromise. The prediction rule showed a very good calibration (Hosmer-Lemshow statistic, P=0.93) and discrimination for both training and testing sets (area under ROC curve= 0.848 and 0.835 respectively). We concluded that our predictive rule was able to predict risk of death due to enterococcal bloodstream infection as well as to identify patients, who were below the threshold value and would have a low risk of death, with a negative predictive value of 96%.