Selecting the Best Prediction Model for Readmission

Abstract

ObjectivesThis study aims to determine the risk factors predicting rehospitalization by comparing three models and selecting the most successful model.MethodsIn order to predict the risk of rehospitalization within 28 days after discharge, 11 951 inpatients were recruited into this study between January and December 2009. Predictive models were constructed with three methods, logistic regression analysis, a decision tree, and a neural network, and the models were compared and evaluated in light of their misclassification rate, root asymptotic standard error, lift chart, and receiver operating characteristic curve.ResultsThe decision tree was selected as the final model. The risk of rehospitalization was higher when the length of stay (LOS) was less than 2 days, route of admission was through the out-patient department (OPD), medical department was in internal medicine, 10th revision of the International Classification of Diseases code was neoplasm, LOS was relatively shorter, and the frequency of OPD visit was greater.ConclusionsWhen a patient is to be discharged within 2 days, the appropriateness of discharge should be considered, with special concern of undiscovered complications and co-morbidities. In particular, if the patient is admitted through the OPD, any suspected disease should be appropriately examined and prompt outcomes of tests should be secured. Moreover, for patients of internal medicine practitioners, co-morbidity and complications caused by chronic illness should be given greater attention.

Design of a Clinical Decision Support Model for Predicting Pneumonia Readmission

“…In contrast, in our study, the predictive accuracy achieved 83.85% for the model designed with RBF-SVM technique using only 6 selected variables. Although the model proposed in this study outperformed the model reported in Lee [10], different dataset and diseases might be the reasons causing such difference. Future study must be conducted to compare the models designed with different artificial intelligence techniques based on the same data set.…”

“…Lee [10] compared that the readmission models designed using regression analysis, decision tree, and neural network with the misclassification rates of 0.214, 0.189, and 0.214, respectively. The model designed using decision tree was demonstrated to have the best predictive rate of 81.1% accuracy.…”

Preventing hospital readmissions: the importance of considering ‘impactibility,’ not just predicted risk

“…This contrasts with the apparent direction of the research field, which sometimes seems more intent on the pursuit of increasingly complex analytical methods. Although further innovation in analytical methods is possible (eg, using neural networks, decision trees or random forests,8 or by incorporating information from electronic health records)9, it is striking that many of the most well-validated (and perhaps therefore, the most useful) models have adopted comparatively simple approaches. For example, the HOSPITAL score is a weighted summation of just seven variables,10 and produces C statistics over 0.70 when applied to international data11 12 (in this issue of BMJ Quality and Safety , Aubert and colleagues have managed to simplify that model still further, while retaining a C statistic at around the same level13).…”

Predictors of Thirty-Day Readmission Among Hospitalized Family Medicine Patients

“…[13][14][15][16][17][18][19][20][21] There are relatively few studies examining the risk factors for readmission among all primary care patients, and many of these were performed outside the United States or in hospitalist practices. 10,[22][23][24][25] It is well known that a small proportion of patients account for a disproportionate number of hospital admissions and health care costs. 26 -28 Identifying these patients is important for quality of care and economic reasons.…”