Table 2.
Adjusted Relationships of Exploration With Patient and Physician Characteristics
| Characteristic | AOR (95% CI) | P Value |
| Patient | ||
| Major depression | 4.12 (1.95–8.71) | <.01 |
| Prompt (yes)* | 2.29 (1.00–5.24) | .05 |
| Physician | ||
| Age | 0.95 (0.90–1.02) | .14 |
| Female | 0.95 (0.30–2.95) | .92 |
| White | 3.30 (0.90–12.18) | .07 |
| Family medicine | 0.81 (0.27–2.38) | .70 |
| Site | ||
| Sacramento, Calif | 2.06 (0.31–13.44) | .45 |
| San Francisco, Calif | 1.23 (0.25–6.11) | .80 |
| Rochester, NY | 1.19 (0.21–6.88) | .84 |
| Solo | 2.90 (0.70–12.00) | .14 |
| Academic | 10.03 (2.13–47.31) | <.01 |
| Barriers score | 1.39 (0.39–4.94) | .61 |
| Confidence score | 0.71 (0.19–2.64) | .61 |
| Personal experience with depression (yes) | 3.11 (1.09–8.89) | .03 |
| MPCC score | 1.04 (0.99–1.10) | .10 |
AOR = adjusted odds ratio; CI = confidence interval; MPCC = Measure of Patient-Centred Communication.
Note: The main outcome of interest was whether the physician explored the topic of suicide with the standardized patient. This variable is referred to as exploration. The model included standardized patient dummy variables (not shown). For categorical variables, AORs were calculated by comparison with the alternate characteristic. For continuous variables (age, barriers score, confidence score, MPCC score), AORs reflect the effect of a unit change in value.
* By design, there were 3 standardized patient conditions: (1) request for Paxil (paroxetine), (2) request for a “medication that might help,” and (3) no specific request. The first 2 conditions were collapsed into 1 category, called “prompt.”