Table 2. Estimated Percentage of Patients With Correct Inhaler Technique by Intervention Groupa.
| Event | In-Person TTG Group | Virtual TTG Group | Difference (SE), % | 95% CI Lower Bound, % |
|---|---|---|---|---|
| Discharge | ||||
| Observations, No. | 59 | 59 | NA | NA |
| Observed, % | 83 | 69 | −14 (8) | −26 |
| Marginal, %b | ||||
| Model 1 | 82 | 71 | −10 (7) | −22 |
| Model 2 | 82 | 72 | −10 (7) | −22 |
| Model 3 | 81 | 70 | −11 (5) | −19 |
| 30 d After Discharge | ||||
| Observations, No. | 51 | 49 | NA | NA |
| Observed, % | 63 | 53 | −10 (10) | −26 |
| Marginal, %b | ||||
| Model 1 | 61 | 55 | −6 (9) | −21 |
| Model 2 | 60 | 56 | −4 (9) | −18 |
| Model 3 | 64 | 52 | −12 (7) | −23 |
Abbreviations: NA, not applicable; TTG, teach-to-goal.
a
Correct inhaler score equals at least 10 of 12.
b
Estimated by setting all observations to in-person TTG (or virtual TTG) and averaging estimated percentages over the observed sample. Model 1 represents logistic regression adjusting for baseline score, model 2 represents logistic regression adjusting for baseline score and health literacy (binary indicator for low literacy), and model 3 represents ordinal logistic regression of score (range, 1-12) adjusting for baseline score and health literacy.