Table 4 Classification results and contingency table for our logistic regression model (n = 5,176)
From: Mining geriatric assessment data for in-patient fall prediction models and high-risk subgroups
logistic regression model | |||||
|---|---|---|---|---|---|
Classification accuracy | 56.2% | ||||
sensitivity | 63.5% | Prediction | |||
specificity | 55.4% | No | yes | Sum | |
neg. predictive value | 93.5% | no fall | 2596 | 2087 | 4684 |
pos. predictive value | 13.0% | Fall | 180 | 313 | 493 |
AUC | 0.63 | Sum | 2776 | 2400 | 5176 |