Research Article
Length of Hospital Stay Prediction at the Admission Stage for Cardiology Patients Using Artificial Neural Network
Table 3
Input variables in preadmission and predischarge stages with associated values for ANN models.
| Stage | Variables | Value (Boolean value) |
| Preadmission | Gender | | Male (0) | Female (1) | Age | | 21~99 | Location | | Taipei branch (0) | Tamshui branch (1) | Main diagnosis (for non-CAS patients) | | AMI (0) | HF (1) | Comorbidity | Myocardial infarction (ICD410/412) | Absence (0) | Presence (1) | Diabetes (ICD250) | Absence (0) | Presence (1) | Cerebrovascular disease (ICD433/434/437/438) | Absence (0) | Presence (1) | Cardiac dysrhythmias (ICD427) | Absence (0) | Presence (1) | Heart failure (ICD428) | Absence (0) | Presence (1) | Chronic airway obstruction (ICD496) | Absence (0) | Presence (1) | Hypertensive disease (ICD401/402/403/404) | Absence (0) | Presence (1) | Coronary atherosclerosis (ICD414) | Absence (0) | Presence (1) |
| Predischarge | Intervention | Percutaneous transluminal coronary angioplasty (PTCA) | No (0) | Yes (1) | Percutaneous coronary intervention (PCI) | No (0) | Yes (1) | Coronary angiography | No (0) | Yes (1) | Coronary stenting | No (0) | Yes (1) | Cardiac catheterization | No (0) | Yes (1) | Left ventricular X-ray | No (0) | Yes (1) | Use TW-DRG as payment method | | No (0) | Yes (1) |
|
|