Table 2.
Results binary logistic regression on recognition of cognitive dysfunction by physicians
| Predictor variable | Univariable OR [95%CI] | Multivariable OR [95%CI] |
|---|---|---|
| Patient age (in years) | 1.0 [0.98—1.04], P = 0.68 | - |
| Patient gender | ||
| Female | 0.91 [0.59—1.41], P = 0.68 | - |
| Male | Reference level | |
| Work experience physician (in years) | - | |
| < 5 | Reference level | |
| 5–10 | 1.10 [0.57—2.12], P = 0.78 | |
| 11–15 | 1.37 [0.36—5.23], P = 0.64 | |
| > 15 | 1.17 [0.55—2.48], P = 0.68 | |
| Hospital type | - | |
| General | Reference level | |
| University | 0.91 [0.41—2.03], P = 0.82 | |
| Medical department* | ||
| Cardiology | 0.27 [0.10—0.72], P < 0.01 | 0.36 [0.13—1.0], P = 0.05 |
| Gastroenterology & hepatology | 0.19 [0.06—0.62], P < 0.01 | 0.25 [0.07—0.85], P = 0.03 |
| Internal Medicine | 0.30 [0.11—0.81], P = 0.02 | 0.35 [0.12—0.96], P = 0.04 |
| Neurology | 0.60 [0.22—1.63], P = 0.32 | 0.72 [0.26—2.02], P = 0.53 |
| Orthopedics | 0.38 [0.12—1.19], P = 0.1 | 0.41 [0.13—1.34], P = 0.14 |
| Pulmonary Medicine | 0.22 [0.06—0.77], P = 0.02 | 0.26 [0.07—0.93], P = 0.04 |
| Surgery | 0.27 [0.10—0.72], P < 0.01 | 0.37 [0.13—1.02], P = 0.05 |
| Geriatrics | Reference level | Reference level |
| Mini-Cog score | ||
| 0 | Reference level | Reference level |
| 1 | 0.35 [0.19—0.64], P < 0.001 | 0.36 [0.19—0.69], P < 0.01 |
| 2 | 0.31 [0.18—0.54], P < 0.001 | 0.34 [0.19—0.59], P < 0.001 |
Abbreviations: OR Odds ratio, 95%CI = 95% Confidence interval, AIC Akaike information criterion
Multilevel regression analysis showed no difference between hospitals: model with random intercept of hospitals AIC = 451.9; model without random intercept of hospitals AIC = 450.1
* Medical departments with less than 15 included patients were excluded from the analysis. Adding this variable did not improve the multivariable regression model (AIC)