Presentation Type
Event
Student
No
Track
Health Care Application
Abstract
Diabetes poses a variety of medical complications to patients, resulting in a high rate of unplanned medical visits, which are costly to patients and healthcare providers alike. However, unplanned medical visits by their nature are very difficult to predict. The current project draws upon electronic health records (EMR’s) of adult patients with diabetes who received care at Sanford Health between 2014 and 2017. Various machine learning methods were used to predict which patients have had an unplanned medical visit based on a variety of EMR variables (age, BMI, blood pressure, # of prescriptions, # of diagnoses on problem list, A1C, HDL, LDL, and a ranked variable for tobacco use severity). A radial-basis support vector machine (SVM) was the most accurate method, achieving a hit rate of 68.5% and a correct rejection rate of 62.9% during cross-validation testing. Follow-up testing of the trained SVM indicated that, of the modifiable prediction variables, high blood pressure and low levels of high-density lipoprotein (HDL) were most strongly predictive of unplanned medical visits. Future directions include refining and validating the predictive model, towards the ultimate goal of developing and implementing clinical recommendations for preventing unplanned medical visits among adult patients with diabetes.
Start Date
2-5-2019 11:00 AM
End Date
2-5-2019 12:00 PM
Included in
Applied Statistics Commons, Emergency Medicine Commons, Endocrinology, Diabetes, and Metabolism Commons, Multivariate Analysis Commons, Preventive Medicine Commons, Statistical Models Commons
Predicting Unplanned Medical Visits Among Patients with Diabetes Using Machine Learning
Pheasant Room 253 A/B
Diabetes poses a variety of medical complications to patients, resulting in a high rate of unplanned medical visits, which are costly to patients and healthcare providers alike. However, unplanned medical visits by their nature are very difficult to predict. The current project draws upon electronic health records (EMR’s) of adult patients with diabetes who received care at Sanford Health between 2014 and 2017. Various machine learning methods were used to predict which patients have had an unplanned medical visit based on a variety of EMR variables (age, BMI, blood pressure, # of prescriptions, # of diagnoses on problem list, A1C, HDL, LDL, and a ranked variable for tobacco use severity). A radial-basis support vector machine (SVM) was the most accurate method, achieving a hit rate of 68.5% and a correct rejection rate of 62.9% during cross-validation testing. Follow-up testing of the trained SVM indicated that, of the modifiable prediction variables, high blood pressure and low levels of high-density lipoprotein (HDL) were most strongly predictive of unplanned medical visits. Future directions include refining and validating the predictive model, towards the ultimate goal of developing and implementing clinical recommendations for preventing unplanned medical visits among adult patients with diabetes.