Table 2.
Comparison of home blood pressure recording fidelity in all participants
| Outcome measures and parameter | Smartphone app (n=38) | Logbook (n=41) | Difference (95% CI) | P value | |
| Fidelitya |
|
|
|
|
|
|
|
Median (IQRb) | 66.7 (32.1 to 83.3) | 52.4 (29.8 to 72.6) | 8.33c (−4.76 to 22.6) | .21d |
|
|
Mean (SD) | 58.6 (29.2) | 50.7 (27.2) | 7.95 (−4.67 to 20.6) | .21e |
|
|
LS meanf (RMSEg) | 59.2 (28.3) | 49.8 (28.3) | 9.43 (−3.15 to 22.0) | .14h |
| Time-independent fidelity |
|
|
|
|
|
|
|
Median (IQR) | 88.7 (57.1 to 97.6) | 84.5 (56.0 to 96.4) | 2.38c (−3.57 to 11.9) | .30d |
|
|
Mean (SD) | 78.4 (22.4) | 73.2 (27.4) | 5.21 (−6.07 to 16.5) | .36e |
|
|
LS mean (RMSE) | 78.8 (26.2) | 72.8 (26.2) | 6.01 (−5.36 to 17.4) | .30h |
aFidelity is defined as the percentage of specified home blood pressure readings over the 3-week home blood pressure monitoring regimen which was recorded, regimen compliant, and successfully reported at the final clinic visit.
bIQR: interquartile range.
cHodges-Lehmann shift parameter estimate (median difference).
dWilcoxon rank-sum test.
e2-sample t test.
fLS mean: least-squares mean.
gRMSE: root mean square error.
hMixed-model longitudinal analysis of variance (ANOVA) adjusting for age, study arm, age by study arm interaction, and baseline systolic and diastolic blood pressures.