TABLE 3.
Daily adherence to maintenance therapy during months 4–6 and the difference in arms 1–4 versus arm 5
Arm 1 (n=87) | Arm 2 (n=88) | Arm 3 (n=88) | Arm 4 (n=88) | Arm 5 (n=86) | |
Baseline | |||||
n (observed) | 63 | 69 | 66 | 65 | 72 |
Mean±sd % | 76.5±24.4 | 73.7±28.6 | 69.7±33.8 | 73.1±27.4 | 73.2±30.2 |
Months 4–6# | |||||
n (observed with baseline) | 59 | 65 | 62 | 59 | 72 |
n (observed and imputed) | 83 | 84 | 84 | 82 | 85 |
Least squares mean±se¶ % | 80.9±3.19 | 77.2±3.04 | 78.3±3.11 | 77.1±3.25 | 69.0±3.19 |
CIS arm versus arm 5 | |||||
Difference % | 12.0 | 8.2 | 9.3 | 8.1 | |
95% CI % | 5.2–18.8 | 1.6–14.9 | 2.7–16.0 | 1.4–14.8 | |
p-value | <0.001 | 0.016 | 0.006 | 0.018 |
Arm 1: maintenance data to participants and HCPs; arm 2: maintenance data to participants; arm 3: maintenance and rescue data to participants and HCPs; arm 4: maintenance and rescue data to participants; arm 5: no feedback (control). CIS: connected inhaler system; CI: confidence interval; HCP: healthcare professional. #: “observed with baseline” is the number of participants between the beginning of month 4 and the end of month 6 who have completely observed adherence, or partially observed adherence with intermittent missing data being imputed, and with no missing baseline adherence. “Observed and imputed” additionally includes participants who have missing baseline adherence imputed due to a device transmission failure or a human error, or who have no observed adherence for this time period and have all of their adherence data imputed. ¶: adjusted for effects due to randomised treatment (study arm), baseline adherence, number of run-in visits, country, sex and age (in years).