| Autor: |
Anand RS; Alpert Medical School, Brown University, Providence, RI, USA.; Center for Biomedical Informatics, Brown University, Providence, RI, USA., Stey P; Alpert Medical School, Brown University, Providence, RI, USA.; Center for Biomedical Informatics, Brown University, Providence, RI, USA., Jain S; Alpert Medical School, Brown University, Providence, RI, USA.; Center for Biomedical Informatics, Brown University, Providence, RI, USA., Biron DR; Alpert Medical School, Brown University, Providence, RI, USA.; Center for Biomedical Informatics, Brown University, Providence, RI, USA., Bhatt H; Alpert Medical School, Brown University, Providence, RI, USA., Monteiro K; Alpert Medical School, Brown University, Providence, RI, USA., Feller E; Alpert Medical School, Brown University, Providence, RI, USA., Ranney ML; Alpert Medical School, Brown University, Providence, RI, USA.; Emergency Digital Health Innovation Program, Department of Emergency Medicine, Brown University, Providence, RI, USA., Sarkar IN; Alpert Medical School, Brown University, Providence, RI, USA.; Center for Biomedical Informatics, Brown University, Providence, RI, USA., Chen ES; Alpert Medical School, Brown University, Providence, RI, USA.; Center for Biomedical Informatics, Brown University, Providence, RI, USA. |
| Abstrakt: |
Diabetes constitutes a significant health problem that leads to many long term health issues including renal, cardiovascular, and neuropathic complications. Many of these problems can result in increased health care costs, as well risk of ICU stay and mortality. To date, no published study has used predictive modeling to examine the relative influence of diabetes, diabetic health maintenance, and comorbidities on outcomes in ICU patients. Using the MIMIC-III database, machine learning and binomial logistic regression modeling were applied to predict risk of mortality. The final models achieved good fit with AUC values of 0.787 and 0.785 respectively. Additionally, this study demonstrated that robust classification can be done as a combination of five variables (HbA1c, mean glucose during stay, diagnoses upon admission, age, and type of admission) to predict risk as compared with other machine learning models that require nearly 35 variables for similar risk assessment and prediction. |
