Table 3. Ratings of Clinician Communication Behaviors During Office Visits With Standardized Patients With Back Pain.
| Item within habits domain | Score, mean (SD)a | ASD (95% CI)c | P value | |
|---|---|---|---|---|
| Intervention (n = 22)b | Control (n = 28) | |||
| Habit 1, invest in beginning (Cronbach α = .28) | ||||
| Overall | 7.1 (1.6) | 6.7 (1.4) | 0.25 (−0.26 to 0.75) | .33 |
| Expansion of concerns | 4.7 (0.9) | 4.9 (0.5) | NA | NA |
| Eliciting full agenda | 2.4 (1.1) | 1.9 (1.1) | NA | NA |
| Habit 2, elicit patient perspective (Cronbach α = .53) | ||||
| Overall | 10.2 (2.7) | 9.3 (2.9) | 0.62 (0.05 to 1.19) | .04 |
| Eliciting patient’s understanding | 3.7 (1.2) | 3.4 (0.8) | NA | NA |
| Eliciting patient’s goals | 3.3 (1.1) | 2.9 (1.3) | NA | NA |
| Attention to psychosocial issues | 3.2 (1.5) | 2.9 (1.8) | NA | NA |
| Habit 3, convey empathy (Cronbach α = .84) | ||||
| Overall | 10.5 (3.3) | 6.9 (3.5) | 1.16 (0.55 to 1.77) | <.001 |
| Expression of emotion | 3.5 (1.5) | 2.4 (1.3) | NA | NA |
| Acceptance of feelings | 3.9 (1.5) | 2.4 (1.5) | NA | NA |
| Identification of feelings | 3.2 (1.2) | 2.0 (1.2) | NA | NA |
| Habit 4, invest in the end (Cronbach α = .40) | ||||
| Overall | 22.1 (2.6) | 22.1 (2.5) | 0.10 (−0.50 to 0.70) | .74 |
| Using patient’s concerns as frame of reference | 4.1 (1.2) | 3.6 (1.6) | NA | NA |
| Convey optimism | 4.5 (0.9) | 4.4 (0.9) | NA | NA |
| Advocate conservative treatment plan | 4.9 (0.4) | 4.9 (0.4) | NA | NA |
| Watchful waiting message | 4.4 (1.0) | 4.4 (0.9) | NA | NA |
| Convey plans for follow-up | 4.2 (1.2) | 4.7 (0.7) | NA | NA |
Abbreviations: ASD, adjusted standardized difference; NA, not applicable.
a
Item scores range from 1 to 5, with higher scores indicating clinician communication that is more aligned with targeted behavior.
b
Three of the 25 clinicians randomized to the intervention did not complete the standardized patient evaluation visit at approximately 9 months of follow-up, so were not included in the analysis.
c
Computed by dividing the regression coefficient for the standardized scale score for each domain by the root mean square error of the dependent variable in the regression models. Models were adjusted for physician age, sex, and stress from uncertainty.25