OBJECTIVES : Connected insulin pens may reduce treatment burden for people with diabetes (PwD) by automatically recording and integrating insulin dosing information with relevant data from other devices (e.g. glucose monitor). This study elicited preferences of PwD for connected pens and associated mobile applications.

METHODS : A literature review and patient interviews informed an online survey in the US and UK, which included a video-introduction to connected insulin pens, a discrete choice experiment (DCE), and background questions. Within the DCE interim analyses, PwD repeatedly chose between a standard pen and two connected pens described by five attributes: 1) device type (i.e. attached vs integrated Bluetooth® transmitter); 2) dosing support services; 3) glucose monitoring support; 4) additional features (i.e. physical activity and dietary trackers); and 5) data sharing functions. A logit model was used to estimate relative attribute importance (RAI) scores. Predicted choice probabilities (PCPs) were obtained for 15 specifications of insulin pens and corresponding mobile applications.

RESULTS : 143 T1DM PwD (US: 64; UK: 79) and 361 T2DM PwD (US: 248; UK: 113) participated in the study. A connected pen was preferred in 79.6% of DCE choices. Preferences were most affected by glucose monitoring options (total: RAI = 41.4%; US: RAI = 43.6%; UK: RAI = 38.4%), followed by additional features (total: RAI = 21.1%; US: RAI = 19.0%; UK: RAI = 24.2%), data sharing functions (total: RAI = 19.6%; US: RAI = 20.4%; UK: RAI = 18.4%), dosing support services (total: RAI = 13.5%; US: RAI = 14.0%; UK: RAI = 12.8%), and device type (total: RAI = 4.3%; US: RAI = 3.0%; UK: RAI = 6.2%). A connected pen was preferred over a standard pen in 4812 out of 6048 scenarios (US:2982/3744; UK:1830/2304).

CONCLUSIONS : Findings suggest that PwD expect connected insulin pens will be a valuable tool for insulin therapy.