Table 2.
Indicators of program adoption and use among participants in the intervention arm (My Diabetes Coach app).
| Indicators of program adoption and use | Intervention arm (n=93) | ||
| Uptake of the MDCaapp, n (%) | |||
|
|
Participants who had at least one “appointment” with “Laura” over 12 months | 92 (99) | |
|
|
Participants who had uploaded glucose data into the MDC app | 83 (89) | |
|
|
Number of completed chats with “Laura” over 12 months, n (%) | ||
|
|
|
0-6 | 26 (28) |
|
|
|
7-24 | 37 (40) |
|
|
|
25 or more | 30 (32) |
|
|
Total number of chats completed over 12 months | 1942 | |
|
|
Total number of valid chats completed over 12 monthsb | 1641 | |
| Individual-level MDC app usagec, mean (SD); range | |||
|
|
Number of chats completed per person | 21.8 (16.7); 1-65 | |
|
|
Number of valid chats completed per personb | 18.4 (15.0); 1-53 | |
|
|
Duration of valid chats per person (in minutes), mean (SD); ranged | ||
|
|
|
Total duration of chats | 242.7 (212.3); 0-1050 |
|
|
|
Mean duration of each valid chat | 13.4 (4.8); 3-26.8 |
|
|
Glucose data uploaded, mean (SD); range | ||
|
|
|
Number of glucose level uploads per person | 181.8 (192.1); 1-966 |
| Program delivery, total; mean per month (SD) | |||
|
|
Clinical alerts (eg, abnormal glucose level) | 297; 13.7 (8.8) | |
|
|
Technical alerts (eg, glucose uploading failed) | 179; 8.3 (6.5) | |
|
|
Posts on the web-based discussion forum | 19; 1.1 | |
aMDC: My Diabetes Coach.
bInvalid chats were defined as chats for which participants exited the app before the coach modules were fully completed with the closing remark.
cFor individual-level information, the estimation is based on 92 participants who had records of chat with Laura through the app and 83 participants who had uploaded their glucose levels into the app. Mean (SD) and range of number and duration of chats and glucose data uploads were reported.
dOnly completed chats have been included in the calculation of the total duration of chats. If the users did not exit the app after completing the chats, the duration would be continuously counted. Thus, we truncated the values if the duration of the chats were more than two interquartile ranges above the third quartile of the distribution.