A Qualitative Analysis of Provider-Level Barriers to Prescribing Diabetes Technology.

Background: Patient utilization of diabetes technology differs based on sociodemographic and other factors. Underserved patients have reported that providers decline to prescribe continuous glucose monitors (CGMs) and insulin pumps. This qualitative study elucidated provider perspectives regarding facilitators and barriers to the prescription and patient use of diabetes technology. Methods: Sixteen diabetes care providers (75.0% MD, 18.8% APRN, 6.3% PharmD, and 50.0% adult endocrinology) at academic health centers, a Veterans Affairs Medical Center, and a safety net hospital in the southeastern United States were surveyed and interviewed from January to May 2024. Survey data were analyzed using descriptive statistics; a thematic analysis was used for interview transcripts with an adapted version of the social-ecological model (SEM) as the coding framework. Results: On the survey, providers estimated that 80% (IQR 58.0%-86.0%) of their patients who met American Diabetes Association criteria for CGMs and 50% (IQR 48.0%-63.0%) for insulin pumps regularly used them. System factors (e.g., lack of insurance, high device costs, and insurance bureaucracy) were perceived by providers (62.5%-93.8%) as patients' top barriers to use. Across the interviews (n = 362 codes), providers' top prescribing barriers were also system-level (65.7%), including working with insurance, durable medical equipment (DME) companies, and pharmacies (15.7%), checking eligibility requirements (11.0%), and electronic health record (EHR) limitations (9.7%). Interpersonal prescribing barriers (7.5%) were tied to patients having low health literacy (3.6%) and communication with non-English-speaking patients (1.7%). Individual prescribing barriers (26.0%) included patients expressing concerns about device adhesives/appearance (5.5%) and patients having limited knowledge/interest (3.0%). Facilitators across SEM levels included simplified eligibility criteria, EHR order sets, shared decision-making, and proactive insurance/DME companies. Conclusions: These findings indicate the need for multilevel solutions to improve the prescription and use of diabetes technology. Future research and clinical practice should aim to enhance EHR functionality and system integration, improve patient-provider communication, and streamline insurance criteria and processes.