Hospital readmissions are expensive and can be harmful to patients. However, many times readmission are preventable and can be reduced with the help of accurate predictions. CONTENT is a deep model that captures both global and local contexts from EHR of CHF patients and predicts readmissions using interpretable representations by generating a context vector for every patient to characterize their overall condition. It uses hybrid TopicRNN model incorporating the power of both deep neural network and probabilistic generative models to learn patient representations and predicts whether or not a CHF patient currently admitted will be readmitted with 30 days of his discharge from the hospital.
Input variables : Features from EHR data like ICD Codes, Comorbidities, Clinical Texts
Output Variables : Indication of Readmission (binary classification)
| Statistical | : | Somers D | Accuracy | Precision and Recall | Confusion Matrix | F1 Score | Roc and Auc | Prevalence | Detection Rate | Balanced Accuracy | Cohen's Kappa | Concordance | Gini Coefficent | KS Statistic | Youden's J Index |
| Business | : | Readmission Rate | Avg Hospital LOS |
| Infrastructure | : | Log Bytes | Logging/User/IAMPolicy | Logging/User/VPN | CPU Utilization | Memory Usage | Error Count | Prediction Count | Prediction Latencies | Private Endpoint Prediction Latencies | Private Endpoint Response Count |
Visit Model : github.com
Additional links : journals.plos.org
| Model Category | : | Public |
| Date Published | : | April, 2018 |
| Healthcare Domain | : | Provider |
| Code | : | github.com |
| Utilization Management |
| Readmission |