Multivariable Logistic Model Results for Electronic Health Record-Documented Outcomes*
| EHR-Documented Outcomes | Time Points† | |
|---|---|---|
| Intervention Visit AOR (95% CI), P Value | Postintervention Visit AOR (95% CI), P Value | |
| Ask | 2.26 (0.94–5.46), .07 | 1.49 (0.62–3.57), .37 |
| Advise | 1.44 (0.66–3.13), .36 | 1.44 (0.54–3.79), .47 |
| Assess | 4.49 (1.62–12.49), .004 | 3.34 (1.32–8.44), .01 |
| Assist | 17.63 (4.67–66.57), < .001 | 2.02 (0.54–7.56), .30 |
| Arrange | 10.28 (2.92–36.23), < .001 | 1.80 (0.44–7.39), .42 |
AOR, adjusted odds ratio; EHR, electronic health record; 95% CI, 95% Confidence Interval.
↵* All models used generalized estimating equations (GEE) to adjust PCP clusters. Multivariable logistic regression models included the following as covariates: ethnicity (Korean vs. Vietnamese), age (<60 vs 60 or older), education (<high school vs. high school or beyond), presence vs. absence of a tobacco-related disease condition (hypertension, diabetes, chronic obstructive pulmonary disease, cardiovascular diseases, stroke, or cancer), and number of cigarettes smoked per day (<10 vs 10 or more). Bold text denotes significant AOR, 95% CI, and P-value (P < 0.05).
↵† Referent is preintervention visit. Three time points were included in this study: (1) at preintervention, one primary care (PCP) visit prior to the intervention; (2) intervention visit with participants receiving the interactive Mobile Doctor (iMD) intervention; and (3) at postintervention, one PCP visit after the intervention visit.