Bridging the Gap in Carbohydrate Counting With a Mobile App: Needs Assessment Survey

Abstract

Background: Carbohydrate counting (CC) can be burdensome and difficulty with adherence has been reported. Automated CC through mobile apps offers innovative solutions to ease this burden. Objective: This cross-sectional web-based survey aims to identify (1) perceived barriers to CC by Canadians living with type 1 diabetes (T1D) and (2) app features that would help reduce these barriers. The secondary objective aims to compare apps used by participants with the suggested app features. Methods: People with T1D aged 14 years and older, living in Canada, were recruited through the BETTER Canadian registry, diabetes organizations, and social media. Participants completed a 39-question web-based survey (closed- and open-ended) to identify barriers in CC, preferred CC app features, and current app use. Respondents rated barriers and app features using a 5-point Likert scale. The features were cross-referenced in each app reported being used by participants. Descriptive statistics summarized barriers and app feature preferences, and statistical analyses identified differences by age, app use, and insulin modality. Mean scores (out of 5) were compared using 2-tailed t tests or nonparametric tests. Open-ended questions were analyzed using inductive thematic analysis. Results: Participants (N=196; woman: n=145, 74%; mean age 40 [SD 17] years; mean diabetes duration 22 (14) years; relied on CC to determine insulin doses at mealtimes: n=178, 90.8%) reported barriers related to carbohydrate identification, nutrient interaction, and insulin dose calculation, as well as psychosocial factors. Preferred app features included nutrient analysis (165/196, 84.2%), personalization (151/196, 77.1%), insulin bolus calculation (145/196, 74%), and health care professional support (135/196, 68.8%). Among the 16 apps used by participants, most (12/16, 75%) supported nutrient analysis but only one offered bolus calculations or health care professional support, and none offered personalization. Users on injections reported greater barriers to blood glucose monitoring for insulin adjustments compared to exclusive pump users (mean score of 3.87, SD 1.22 vs mean 3.30, SD 1.28; P=.001). They also expressed higher needs for meal logs in an electronic food journal (mean 4.06, SD 1.18 vs mean 3.69, SD 1.17; P=.01), bolus dose suggestions (mean 4.37, SD 0.98 vs mean 3.84, SD 1.26; P=.001), and app personalization (mean 4.47, SD 0.86 vs mean 3.93, SD 1.21; P