Box 2.

Details About the Automated Insulin Delivery Amongst Pregnant Women with Type 1 Diabetes Trial and Training Provided

AiDAPT trial

The AiDAPT trial was designed to evaluate the clinical efficacy of HCL insulin delivery throughout pregnancy (∼24 weeks duration, from 14 to 38 weeks) in real-life NHS maternity care settings (see: https://www.isrctn.com/ISRCTN56898625).

To be eligible for the trial, women had to have been diagnosed with T1D for at least 12 months, have a viable pregnancy confirmed by ultrasound (up to 13 weeks and 6 days duration), be using intensive insulin therapy (MDI or pump, including sensor-augmented pumps or other HCL system), be able to wear and use study devices, and have an HbA1c level ≥48 mmol/mol at booking (first antenatal contact) and ≤86 mmol/mol (at randomization). Women with an HbA1c <48 mmol/mol were considered to be optimally managed and achieving glucose targets as per current recommendations.20 These women were not included in the trial as it was felt that they would be unlikely to derive significant glycemic benefit from using HCL technology.

Following randomization, participants in the intervention arm received training to use the CGM, insulin pump, and CamAPS FX app. Due to the Covid-19 pandemic, participants in both study arms were given the option to attend training (and research visits) virtually via video-call or telephone, with virtual support provided by a research educator. Initiation on the HCL included starting/stopping the system, meal bolus procedure, use of alarms, and device trouble-shooting. Staff at local sites assessed women's competency to use the HCL. Training recommendations were reinforced using pregnancy ‘top tips' leaflets, mealtime and CGM advice (https://abcd.care/dtn-uk-top-tips); and webinars (https://camdiab.cdept.org.uk/). Participants had access to a 24 h telephone helpline to contact their local study team with any study-related matters. They could also contact the research educator if requiring technical support.

The primary outcome is the percentage of time spent with CGM glucose levels between 3.5 and 7.8 mmol/L between 16 weeks' gestation and delivery. Secondary outcomes include maternal glycemic, obstetric, and psychosocial outcomes, neonatal health outcomes, safety outcomes, healthcare professional experiences, and health economic outcomes. Further detail about the trial is available in the study protocol.18

Training for healthcare professionals

The trial was conducted in nine UK sites involving healthcare professionals who had varying levels of experience supporting diabetes technologies, ranging from those who had experience of both CGM and insulin pumps to those with none. Changes to NICE guidelines5 meant that all sites began using Freestyle Libre during 2020, followed by CGM in 2021. Healthcare professionals not familiar with the Dana insulin pump and Dexcom G6 CGM were encouraged to attend training with representatives from the device manufacturers before their site start.

The research educator provided training on the HCL's individual components for the first participants at each study site and as required, thereafter. The initial HCL training session at each site was used as a training opportunity for the local clinical team, delivered by the research educator either face-to-face or virtually (depending on site location and Covid-19 restrictions). If there were long gaps between HCL participants, or if staffing was difficult due to Covid-19 redeployments, the research educator delivered the training either in full or by involving the local clinical team as a refresher. Competency checklists were used for each system component to ensure all aspects of the training had been covered.

Additional resources used to support training included electronic information sheets (Top Tips for using CGM or closed-loop during pregnancy); peer support from experienced colleagues and case-based discussions to share healthcare team experiences; and publicly available webinars and CamDiab online training (https://www.camdiabtraining.com/) on utilizing the CamAPS FX HCL during pregnancy. The CamAPS FX clinic portal allowed HCL users' glucose data to be aggregated for review (daily, weekly, or monthly) by local clinical teams and/or by members of the research team. There is also an in-app link to the user manual, which includes trouble-shooting tips and Frequently Asked Questions. In addition to delivering initial HCL training, the research educator provided ongoing technical support for sites, as required.

AiDAPT, Automated insulin Delivery Amongst Pregnant women with Type 1 diabetes; MDI, multiple daily injections; T1D, type 1 diabetes.