METHOD: In this study, the MNSI data were collected from the Epidemiology of Diabetes Interventions and Complications (EDIC) clinical trials. Two different datasets with different MNSI variable combinations based on the results from the eXtreme Gradient Boosting feature ranking technique were used to analyze the performance of eight different conventional ML algorithms.
RESULTS: The random forest (RF) classifier outperformed other ML models for both datasets. However, all ML models showed almost perfect reliability based on Kappa statistics and a high correlation between the predicted output and actual class of the EDIC patients when all six MNSI variables were considered as inputs.
CONCLUSIONS: This study suggests that the RF algorithm-based classifier using all MNSI variables can help to predict the DSPN severity which will help to enhance the medical facilities for diabetic patients.
METHODS: Demographic data, underlying diseases, procedures and details on polymyxin B therapy were retrospectively analyzed in a cohort of 84 patients who received intravenous polymyxin B in an intensive care unit from 2010 to 2014.
RESULTS: Polymyxin B was used to treat bacteremia (46.4% of cases) and pneumonia (53.6%). Majority of the pathogens isolated were Acinetobacter spp. (96.4%). The mortality rate was 48.8%, of which 82.9% was attributed to polymyxin B treatment failure. The independent predictors of treatment failure were low doses of polymyxin B (p = 0.002), shorter duration of therapy (p = 0.009), not combining with cefoperazone/sulbactam (p = 0.030), female gender (p = 0.004), administered for treatment of bacteremia (p = 0.023) and renal impairment (p = 0.021). Low polymyxin B doses (p = 0.007), not combining with cefoperazone/sulbactam (p = 0.024), female gender (p = 0.048) and renal impairment (p = 0.022) were also significant predictors for in-hospital mortality.
CONCLUSIONS: To the best of our knowledge, this is the first report on the association of inadequate dose of polymyxin B (<15,000 units/kg/day) with poor outcome in critically ill patients. Besides that, further clinical studies are warranted to evaluate the use of cefoperazone/sulbactam as second antibiotic in the combination therapy.