Patient outcomes after attending pre-pump education class: disparities in initiation and glycemic outcomes

At Children's of Alabama, all patients with insulin dependent diabetes are considered for continuous subcutaneous insulin infusion therapy (CSII, also known as insulin pump). In some cases, eligibility depends on insurance requirements, including six months of MDI therapy, CSII education, and a six-week glucose log with four daily readings.

Evaluate factors influencing CSII initiation after "Prepump class" and assess glycemic changes in CSII starters.

A retrospective review of pre-pump class attendees from January 2022 to July 2023 was completed. Patients who initiated a CSII prior to January 2024 were identified as "CSII-starters," and those who remained in multiple daily injections were identified as "MDI-retainers". Demographic and medical data were compared between these groups. For CSII-starters, type of system and use of automatic insulin delivery (AID) was evaluated. Glycemic outcomes were assessed in those with type 1 diabetes (T1D) with continuous glucose monitor data. A sub-analysis was done for those outside the honeymoon period. Outcomes of AID systems and users of non-AID systems were compared.

Of 283 pediatric patients who attended pre-pump class, 187 (66%) started CSII, with a median initiation time of 108 days (interquartile range 76-154). CSII-starters and MDI-retainers differed by race (p=0.0385) and insurance (p=0.0001), but not by sex, language, or age at diagnosis. Initiators were younger (p=0.0150), had shorter diabetes duration (p=0.0001), lower HbA1c (p=0.0020), and higher CGM use (p<0.0001). Among starters, 70% chose tubeless pumps, and 62% (n=116) used AID systems. Race and insurance were not linked to AID vs non-AID choice, but were associated with CSII initiation. Insurance and race were not associated with selecting an AID over non-AID CSII systems. In glycemic analysis, 183 patients were studied. AID systems showed improved outcomes compared to non-AID systems for the full analysis and sub-analysis of patients outside the honeymoon period.

While race and insurance are not associated with the selection of an AID vs non-AID system, they are associated with CSII-starters versus MDI-retainers. As expected, AID systems outperformed non-AID systems in our cohort. Future work will aim to reduce disparities in CSII and AID access for all interested in diabetes technology.