Table 1 Items and item sets used to induce the classification models along with the percentages of missing values in our data set (n = 5,176 cases); the 19 Lachs and 22 Tinetti sub-scores are not listed separately
From: Mining geriatric assessment data for in-patient fall prediction models and high-risk subgroups
Item (set) name | Missing values in% |
|---|---|
Age on admission | 0.0 |
Sex (m/f) | 0.0 |
Social status (35 sub-items concerning social contacts, activities, living, economic situation) | 54.3 |
Barthel index sum score [12] | 2.2 |
Lachs score (16 sub-items [13]) | 0.0-19.4 |
Timed 'Up & Go' test total time | 58.8 |
Performance-Oriented Mobility Assessment (POMA) by Tinetti (22 sub-items [6]) | 31.7-68.5 |
Mini-Mental State Examination (MMSE) score on admission | 53.2 |
Number of diagnoses on admission | 0.7 |
Number of different medications on admission | 1.0 |
Fall (yes/no) | 0.0 |