Table 3. Study Outcomes by Study Arm With Multiple Imputation of Missing Values.
| Outcome | SMBP, mean (SD) | Comparing enhanced with standard, difference (95% CI) | P valuea | |
|---|---|---|---|---|
| Standard | Enhanced with smartphone application | |||
| No. | 1050 | 1051 | NA | |
| Blood pressure control outcomes b | ||||
| SBP change at 6 mo (primary blood pressure control outcome), mean (SD)c | −10.6 (18) | −10.8 (18) | −.19 (−1.83 to 1.44) | .81 |
| DBP change at 6 mo, mean (SD)c | −4.2 (10) | −4.3 (10) | −.13 (−1.10 to 0.84) | .79 |
| SBP reduction >10 mm Hg at 6 mo, %c | 41 | 40 | OR, 0.98 (0.88 to 1.09) | .72 |
| BP control to <140/<90 mm Hg at 6 mo, % | 29 | 32 | OR, 1.17 (1.01 to 1.34) | .03 |
| BP control to <130/<80 mm Hg at 6 mo, % | 12 | 13 | OR, 1.06 (0.76 to 1.48) | .74 |
| Patient-reported outcomes d | ||||
| How likely are you to recommend (device) to a friend (0-10), % | ||||
| 0-6 (detractor) | 12 | 11 | NA | NA |
| 7-8 (passive) | 19 | 19 | NA | NA |
| 9-10 (promoter) | 69 | 70 | NA | NA |
| Net Promoter Score (primary patient satisfaction outcome)e | .57 | .59 | .02 (−0.05 to 0.08) | .63 |
| Use of device during last month, % | ||||
| Never | 5 | 5 | NA | NA |
| Less than once a week | 14 | 12 | NA | NA |
| About once a week | 26 | 20 | NA | NA |
| 2-3 Times a week | 24 | 21 | NA | NA |
| 4 Or more times a week | 31 | 42 | OR of being in a higher category, 1.44 (1.10 to 1.90) | .01 |
| Shared measurements with your physician during last month, shared, % | 48 | 44 | OR, 0.85 (0.67 to 1.10) | .22 |
| How satisfied are you with (1-5 scale), mean (SD)f | ||||
| Your overall treatment | 4.4 (1.0) | 4.3 (1.0) | −.10 (−0.23 to 0.02) | .10 |
| Your health care clinician | 4.3 (1.1) | 4.3 (1.1) | −0.04 (−0.16 to 0.09) | .51 |
| Your blood pressure medication(s) | 4.1 (1.2) | 4.1 (1.1) | −0.03 (−0.14 to 0.09) | .60 |
| How much do you agree with the following statements about your home BP monitoring device (1-7 scale), mean (SD) | ||||
| My device is easy to use | 6.4 (1.4) | 6.4 (1.4) | −0.03 (−0.20 to 0.14) | .70 |
| Using my device improves my ability to manage my BP | 6.1 (1.5) | 6.0 (1.5) | −0.12 (−0.27 to 0.03) | .11 |
| I find my device useful for managing my blood pressure | 6.0 (1.5) | 5.9 (1.6) | −0.15 (−0.35 to 0.06) | .13 |
| My health care clinician thinks I should regularly use my deviceg | 5.7 (1.9) | 5.7 (1.8) | 0 (−0.24 to 0.24) | .99 |
| Overall, I am satisfied with my experience using my device | 6.3 (1.4) | 6.3 (1.5) | −0.04 (−0.20 to 0.12) | .59 |
| Quality of shared decision-making score (CollaboRATE-5)h | ||||
| Mean (SD) | 3.5 (0.9) | 3.6 (0.9) | 0.08 (−0.05 to 0.22) | .19 |
| No. with the top score | 5.5 | 6.6 | OR, 1.20 (0.81 to 1.78) | .36 |
Abbreviations: BP, blood pressure; DBP, diastolic blood pressure; EHR, electronic health record; NA, not applicable; OR, odds ratio; SBP, systolic blood pressure; SMBP, self-measured blood pressure.
Summary measures of association and P values were calculated using regression models that included a fixed effect for a clinic’s participation and randomization assignment in a concurrent clinic-level quality improvement intervention,18 and accounting for clustering by clinical site with robust standard errors (see Methods). For continuous measurements, a linear model was used and differences calculated; for dichotomous outcomes, a logistic model was used and ORs calculated; for ordinal variables, ordinal logistic regression was used, and ORs for being in the next higher category were calculated.
Multiple imputation was used to impute a follow-up BP measurement for 39 participants with missing EHR data (see Methods) at the 6-month point. For patients with available EHR data (n = 2066), the last clinic measurement was carried forward. For patients with no clinic measurements during follow-up, the baseline measurement was carried forward (ie, the reduction was 0; see Methods).
Difference in BP between self-reported most recent BP measurements at baseline (at the time of randomization) and the most recent office BP measurement from EHR data within 6 months (183 days) of randomization. If no clinic BP measurements were made during follow-up, the documented reduction was 0.
Multiple imputation was used to impute survey responses at the 6-month point (see Methods). A total of 865 participants required imputation for the Net Promoter Score, 1059 for use of device during the last month, 1057 for shared measurements with a physician during the last month, 1078 for “how satisfied were you with” survey questions, 1065 for “how much do you agree” survey about home BP monitoring device, and 1594 for the quality of shared decision-making score.
The Net Promoter Score is calculated as the proportion of promoters minus the proportion of detractors.26,27
Satisfaction scores additionally excluded persons answering NA (7% for “overall treatment,” 9% for “your health care clinician,” and 20% for “your blood pressure medications.”
Additionally excludes persons answering “don’t know” (23%).
Quality of shared decision-making uses responses to 3 questions (1-5 each) adapted from the CollaboRATE-5.23 We calculated the mean score and the proportion with the top score (answering the top of the scale on all 3 items) according to published methods.