| Outcomes and Predictor Variables* | Adjusted Odds Ratio | Wald χ2 | P (χ2 test) |
|---|---|---|---|
| A. Order brain image as part screening or diagnostic evaluation for dementia (yes/no) | |||
| Family/caregiver problems (n = 0–7) regarding dementia endorsed by respondent | 1.09 | 2.47 | .12 |
| Professional time spent in direct patient care | 1.01 | 3.06 | .08 |
| Time since medical school graduation (years) | 1.01 | 1.08 | .30 |
| B. Assess positively screened patients ≥65 years old with more sensitive tests (yes vs. no) | |||
| <5% vs. >40% | 0.60 | 1.62 | .20 |
| 5–20% vs. >40% | 0.47 | 7.33 | .007 |
| 21–40% vs. >40% | 0.55 | 4.60 | .03 |
| Comorbidities (n = 0–6) associated with dementia endorsed by respondent | 0.94 | 1.09 | .30 |
| Tools (n = 0–6) needed to diagnose/treat dementia endorsed by respondent | 0.92 | 1.58 | .21 |
| C. Refer patients ≥65 years old with suspected dementia to other clinicians (yes vs. no) | |||
| <5% vs. >40% | 4.19 | 16.89 | <.001 |
| 5–20% vs. >40% | 2.94 | 19.05 | <.001 |
| 21–40% vs. >40% | 1.57 | 3.29 | .07 |
| Rural geographic location (vs. urban) | 0.56 | 6.77 | .009 |
| Suburban geographic location (vs. urban) | 0.84 | 0.72 | .40 |
| Female sex (vs. male) | 1.31 | 2.56 | .11 |
| Dementia behaviors/complications (n = 0–7) endorsed by respondent | 0.10 | 0.002 | .10 |
| Comorbidities (n = 0–6) associated with dementia endorsed by respondent | 1.11 | 3.19 | .07 |
| Tools (n = 0–5) needed to diagnose/treat dementia endorsed by respondent | 1.13 | 4.15 | .04 |
↵* Independent variables included within each model above are those found to be statistically significant (P ≤ .20) at the bivariate level with the given outcome. There were 11 independent variables for each outcome at the bivariate level.