Comparison of predictive modeling approaches for 30-day all-cause non-elective readmission risk

Abstract

BackgroundThis paper explores the importance of electronic medical records (EMR) for predicting 30-day all-cause non-elective readmission risk of patients and presents a comparison of prediction performance of commonly used methods.MethodsThe data are extracted from eight Advocate Health Care hospitals. Index admissions are excluded from the cohort if they are observation, inpatient admissions for psychiatry, skilled nursing, hospice, rehabilitation, maternal and newborn visits, or if the patient expires during the index admission. Data are randomly and repeatedly divided into fitting and validating sets for cross validations. Approaches including LACE, STEPWISE logistic, LASSO logistic, and AdaBoost, are compared with sample sizes varying from 2,500 to 80,000.ResultsOur results confirm that LACE has moderate discrimination power with the area under receiver operating characteristic curve (AUC) around 0.65-0.66, which can be improved to 0.73-0.74 when additional variables from EMR are considered. These variables include Inpatient in the last six months, Number of emergency room visits or inpatients in the last year, Braden score, Polypharmacy, Employment status, Discharge disposition, Albumin level, and medical condition variables such as Leukemia, Malignancy, Renal failure with hemodialysis, History of alcohol substance abuse, Dementia and Trauma. When sample size is small (≤5000), LASSO is the best; when sample size is large (≥20,000), the predictive performance is similar. The STEPWISE method has a slightly lower AUC (0.734) comparing to LASSO (0.737) and AdaBoost (0.737). More than one half of the selected predictors can be false positives when using a single method and a single division of fitting/validating data.ConclusionsTrue predictors can be identified by repeatedly dividing data into fitting/validating subsets and referring the final model based on summarizing results. LASSO is a better alternative to the STEPWISE logistic regression, especially when sample size is not large. The evidence for adequate sample size can be explored by fitting models on gradually reduced samples. Our model comparison strategy is not only good for 30-day all-cause non-elective readmission risk predictions, but also applicable to other types of predictive models in clinical studies.

The HOSPITAL score and LACE index as predictors of 30 day readmission in a retrospective study at a university-affiliated community hospital

“…The LACE index has variable results in the literature outside Ontario. The LACE index has been shown to have moderate discrimination in studies conducted in North America with over 26,000 Medicare admissions (Garrison et al, 2016), 110,000 discharges in the Chicago, Illinois area (Tong et al, 2016) and 600 patients in a community hospital (Spiva et al, 2016). The LACE index had fair discrimination in a study of 5,800 patients in Singapore (Low et al, 2015) and poor discrimination in a study done on about 500 patients in UK with an average age of 85 years old (Cotter et al, 2012).…”

Vital Sign Abnormalities on Discharge Do Not Predict 30-Day Readmission

“…17 More variable results have been reported elsewhere. The LACE index has been useful and accurate in a study of 26,000 hospital admissions of Medicare patients, 13 110,000 discharges from multiple hospitals in the Chicago, Illinois area, 18 and 600 patients from a community hospital. 19 The LACE index had fair discrimination in a study of 5,800 patients in Singapore 20 and poor discrimination for 500 hospital discharges with an average age of 85 years in the UK.…”

Smart City Analytics

“…We argued in section 1 that the problem of predicting large increases in home care is similar to that of predicting hospital readmissions, where recent model comparisons shows the highest AUC scores of 0.68 to 0.73 [3,9] depending on the data. e purpose of those studies has been to obtain nancial savings and to put extra focus on patients likely to be re-admi ed.…”