The 'flash' adhesive study: a randomized crossover trial using an additional adhesive patch to prolong freestyle libre sensor life among youth with type 1 diabetes mellitus.

Aims: Although strategies to prevent premature sensor loss for flash glucose monitoring (FGM) systems may have substantial benefit, limited data are available. This study among youth with high-risk type 1 diabetes evaluated whether an additional adhesive patch over FGM sensors would reduce premature sensor loss frequency and not cause additional cutaneous adverse events (AEs).

Methods: This is a six-month, open-label, randomized crossover trial. Participants were recruited at completion of prior 'Managing Diabetes in a Flash' randomized controlled trial and allocated to three months of Freestyle Libre FGM sensors with either standard adhesive (control) or additional adhesive patches (RockaDex, New Zealand) (intervention), before crossing over to the opposite study arm. Participants self-reported patch use or non-use, premature sensor loss and cutaneous AEs fortnightly via an electronic questionnaire.

Results: Thirty-four participants were enrolled: mean age (± SD) 17.0 (± 2.2) years; mean HbA1c (± SD) 89 (± 16) mmol/mol (10.3% ± 1.4%). The response rate of questionnaires was 77% (314/408). Premature sensor loss was reported in 18% (58/314) of questionnaires: 20% (32/162) from intervention and 17% (26/152) from control (p = 0.56). Thirty-eight percent (118/314) of questionnaires were non-compliant to protocol allocation. However, per-protocol analysis showed similar findings. No significant difference in AEs was reported between compliant adhesive patch use and non-use (6% [5/78] and 3% [3/118], respectively, p = 0.27).

Conclusions: The adhesive patch investigated in this study does not appear to prevent premature FGM sensor loss. However, the low risk of AEs and low cost of an adhesive patch suggest an individualized approach to their use may still be warranted. Further research is needed to explore alternative strategies to prevent sensor loss.