Table 2 Multiple logistic regression analysis of PIM use in each drug/drug group between 2012 and 2017
Drug groups | 2012 | 2017 | ORa (95% CI) | p-value |
|---|---|---|---|---|
All people, n | 20,889 | 35,212 | Â | Â |
Benzodiazepines, n (%) | 434 (2.1) | 511 (1.5) | 0.731 (0.637–0.837) |  < 0.001 |
Anticholinergics, n (%) | 1898 (9.1) | 2098 (6.0) | 0.657 (0.614–0.702) |  < 0.001 |
Tramadol, n (%) | 367 (1.8) | 114 (0.3) | 0.195 (0.157–0.242) |  < 0.001 |
Propiomazine, n (%) | 608 (2.9) | 352 (1.0) | 0.365 (0.318–0.419) |  < 0.001 |
Codeine, n (%) | 546 (2.6) | 408 (1.2) | 0.440 (0.385–0.502) |  < 0.001 |
Glibenclamide, n (%) | 197 (0.9) | 87 (0.2) | 0.269 (0.208–0.349) |  < 0.001 |
NSAIDs, n (%) | 1024 (4.9) | 941 (2.7) | 0.562 (0.512–0.617) |  < 0.001 |
Antipsychotics, n (%) | 2431 (11.6) | 4323 (12.3) | 1.148 (1.085–1.215) |  < 0.001 |
PIM (total), n (%) | 5995 (28.7) | 7629 (21.7) | 0.724 (0.695–0.754) |  < 0.001 |