14 August 2023: Clinical Research
A Simple and Accurate Model for Predicting Fall Injuries in Hospitalized Patients: Insights from a Retrospective Observational Study in Japan
Shizuka Yaita 1ACDEF* , Masaki Tago 1ABCDEFG** , Naoko E. Katsuki 1ABCDEF , Eiji Nakatani 2ACDE , Yoshimasa Oda 3BDE , Shun Yamashita 1DEF , Midori Tokushima 1DE , Yoshinori Tokushima 1DE , Hidetoshi Aihara 1ADE , Motoshi Fujiwara 1DE , Shu-ichi Yamashita 1DEDOI: 10.12659/MSM.941252
Med Sci Monit 2023; 29:e941252
Table 3 Results of multivariable logistic regression analysis for fall injuries.
| Variable, category (reference) | Analysis 1 | Analysis 2 | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P value* | OR | 95% CI | P value* | |
| Age, years | 1.0 | 1.0–1.1 | 0.003 | 1.0 | 1.0–1.1 | 0.001 |
| Sex, Female (Male) | 0.5 | 0.4–0.8 | 0.001 | 0.5 | 0.4–0.8 | 0.001 |
| Department, Neurosurgery (Internal Medicine) | 1.0 | 0.4–2.4 | 0.965 | |||
| Department, Others (Internal Medicine) | 0.8 | 0.5–1.2 | 0.216 | |||
| Emergency admission, Presence (Absence) | 1.2 | 0.7–2.2 | 0.431 | |||
| Emergency admission, Missing category (Absence) | 3.2 | 0.1–121.4 | 0.533 | |||
| Emergency transport, Presence (Absence) | 0.4 | 0.2–0.7 | 0.003 | 0.4 | 0.3–0.6 | |
| Emergency transport, Missing category (Absence) | 0.5 | 0.3–0.8 | 0.010 | 0.6 | 0.3–1.0 | 0.037 |
| Medical referral letter, Presence (Absence) | 1.5 | 1.0–2.2 | 0.042 | 1.5 | 1.0–2.2 | 0.041 |
| Medical referral letter, Missing category (Absence) | 2.2 | 0.1–79.9 | 0.672 | 3.3 | 0.4–25.7 | 0.258 |
| History of falls, Presence (Absence) | 1.7 | 1.1–2.7 | 0.013 | 1.7 | 1.1–2.6 | 0.012 |
| Hypnotic medications, Using (Not using) | 0.8 | 0.5–1.4 | 0.491 | |||
| Hypnotic medications, Missing category (Not using) | 0.5 | 0.2–1.4 | 0.158 | |||
| Visual impairment, Presence (Absence) | 1.4 | 0.6–3.6 | 0.448 | |||
| Visual impairment, Missing category (Absence) | 0.0 | 0.0- | 0.998 | |||
| Parkinsonism, Presence (Absence) | 1.2 | 0.3–4.9 | 0.822 | |||
| Permanent damage by stroke, Presence (Absence) | 0.4 | 0.1–1.1 | 0.088 | |||
| Eating, requiring assistance (Independent) | 1.0 | 0.6–1.7 | 0.864 | |||
| Eating, Missing category (Independent) | 0.7 | 0.3–1.7 | 0.482 | |||
| Bedriddenness rank, J (Normal) | 4.7 | 1.9–11.7 | 0.001 | 4.9 | 2.0–12.2 | 0.001 |
| Bedriddenness rank, A (Normal) | 7.9 | 3.5–17.8 | 8.9 | 4.0–19.8 | ||
| Bedriddenness rank, B (Normal) | 6.2 | 2.6–15.2 | 7.3 | 3.1–17.1 | ||
| Bedriddenness rank, C (Normal) | 4.1 | 1.6–10.8 | 0.004 | 4.5 | 1.9–11.1 | 0.001 |
| Bedriddenness rank, Missing category (Normal) | 1.5 | 0.3–7.7 | 0.636 | 1.2 | 0.2–5.6 | 0.846 |
| OR – odds ratio; CI – confidence interval. Bedriddenness ranks: J – independence/autonomy; A – house-bound; B – chair-bound; C – bed-bound. The factors used in the multivariable logistic regression were all available on admission and had low collinearity with each other. The model was designed as a parsimonious model using 6 factors that had significance by the multivariable logistic regression of 13 factors. * value for Wald test. | ||||||