Increasing Use of Diabetes Devices: What Do Health Care Professionals Need?

This was a cross-sectional, online survey study of HCPs in primary care, adult endocrinology, and pediatric endocrinology practices in the United States. Surveys were completed in January and February 2022. A priori, we set out to collect most survey data from primary care professionals, reflecting the fact that the majority of people with diabetes are treated in primary care settings. Primary care HCPs included family practice, pediatric, and internal medicine HCPs.

Survey participants were invited to complete an online survey if they were HCPs, defined as physician, advanced practice professional, nurse, dietitian, or diabetes educator, practicing in a primary care, adult endocrinology, or pediatric endocrinology setting, including private practices, hospital-based practices, and academic centers. Participants had to actively work with patients diagnosed with either type 1 diabetes or type 2 diabetes. There were no exclusions for practice size, number of patients with diabetes, or practice setting.

HCPs were recruited by advertisements on websites, social media outlets for diabetes professionals, and university listservs; through physician interest groups and diabetes interest groups; and by virtue of being colleagues of the investigators. This approach provided a convenience sample of responses from individuals in a variety of practice settings that might not have been reached through professional organization distribution lists.

This study was not powered to test formal hypotheses. We aimed to collect the most responses from primary care HCPs, followed by adult endocrinology HCPs, and then pediatric endocrinology HCPs, to reflect in ascending order the likely degree of experience using diabetes devices. The survey response window was shortened for endocrinology HCPs to allow more time to gather responses from primary care HCPs.

The online survey used in this study was created by the investigators and field-tested with 18 HCPs for clarity, missing concepts, and comprehension. As indicated below, domains of interest included attitudes toward diabetes devices, comfort level with diabetes devices, and perceived usefulness of hypothetical clinic tools to assist in their practice. Three questions were added based on feedback from field-testers. The average time to complete the survey was 10.8 minutes. The refined final survey was then administered to HCPs across the United States (Supplementary Material).

The first part of the survey ascertained demographic information and respondents’ practice types. It included questions about personal demographics, a description of practice type, years in practice, practice setting, and the proportion of patients with type 1 or type 2 diabetes. HCPs were asked about additional personnel to whom they had access such as diabetes care and education specialists (DCESs), clinical pharmacists, and case managers.

HCPs were asked to categorize themselves into one of three “clinical personas” based on work by Tanenbaum et al. (18) describing clinicians’ attitudes about, biases toward, and acceptance of diabetes technology. The options included whether they considered themselves to have “positive attitudes toward diabetes technologies [and] perceive low patient barriers to diabetes devices” (the Ready persona), “positive attitudes toward diabetes technologies [and] perceive high patient barriers to diabetes devices” (the Cautious persona), or a “more cautious attitude toward diabetes technologies [and] having little time to review data in clinic, difficulty keeping up with advances, and concern about potential patient barriers” (the Not Yet Ready persona).

HCPs were asked about their comfort level with six aspects of engaging with diabetes devices in an outpatient setting, including identifying good candidates, answering questions about devices, checking health insurance coverage for devices, writing prescriptions for devices, training patients to use devices, and reviewing/interpreting data from devices. These workflow questions were assessed for CGM, smart insulin pens, insulin pumps, and HCL systems. Items were scored on a 10-point Likert scale, with 1 indicating “not at all” comfortable and 10 indicating “very” comfortable.

Based on previous literature (16,17), we described a series of hypothetical tools that could be used to help HCPs engage with diabetes devices (Table 1). After describing the function of each tool, we asked participants to indicate how useful each tool would be in clinical practice using a visual analog scale (VAS) with scoring from 0 (“not useful”) to 100 (“most useful”). Additionally, participants were asked how important it was that these resources be accessible directly from an EHR interface; however, their current use of an EHR system was not assessed.

The Colorado Multiple Institutional Review Board determined that this study and its procedures were considered exempt research. All survey respondents provided electronic informed consent. HCPs accessed and completed the consent form and survey via REDCap, an electronic, Health Insurance Portability and Accountability Act–compliant database used for research (23,24). Those who completed the survey were offered a $50 gift card as compensation for their time.

Data were checked by three investigators for missingness, outliers, accuracy, and human (as opposed to “bot”) completion. All continuous variables were assessed for normality using the Shapiro-Wilk test, and none of the variables reported below were normally distributed. Descriptive statistics are reported as median (interquartile range [IQR]) for continuous data and percentages for categorical data. Continuous variables were compared using Kruskal-Wallis rank sum tests, and χ² or Fisher exact tests were used for categorical variables.

From January through February 2022, a total of 240 HCPs completed the survey. Because the recruitment method included advertisements on social media and listservs, a response rate could not be calculated. Of the 240 records completed, 115 (48%) were from primary care HCPs, 89 (37%) were from adult endocrinology HCPs, and 36 (15%) were from pediatric endocrinology HCPs in line with the study’s asymmetric recruitment goals (Table 2). Of the primary care HCPs, 71 (61.7%) specialized in family medicine, 39 (33.9%) were in internal medicine, 2 (1.7%) were from primary care pediatrics, and 3 (2.6%) were from medicine/pediatrics.

The HCPs in this study resided in 41 states in the United States and represented 222 unique practices (Table 2). Most HCPs were prescribers, with an average of 13.8 years in practice. Females made up the majority of the pediatric and adult endocrinology participants and 48.7% of the primary care HCPs. The highest proportion of primary care HCPs worked in clinician-owned solo or group practices (37.4%), whereas most pediatric endocrinology HCPs worked at an academic center (69.4%). Adult endocrinologists were concentrated in academic centers (40.4%) and hospital/health system– owned practices (34.8%).

The majority of HCPs across all disciplines cared for both people with type 1 diabetes and people with type 2 diabetes (Table 3). More than 55% of pediatric endocrinologists cared for adults with diabetes, whereas only 4.5% of adult endocrinologists cared for children with diabetes. Primary care HCPs cared for a median of 25 (IQR 10–50) people with diabetes who used insulin each month, with pediatric endocrinology HCPs reporting a median of 42.5 (IQR 23.75–85), and adult endocrinology HCPs reporting a median of 96 (IQR 50–122.5). Pediatric endocrinology HCPs reported that 72.2% of the insulin-requiring people with diabetes in their practice used CGM, 42.7% used insulin pumps, and 45.4% used HCL systems. This was more than CGM, insulin pump, and HCL use by insulin-requiring patients reported by adult endocrinologists (53.0, 22.1, and 29.2%, respectively) or by primary care HCPs (25.4, 11.9, and 10.1%, respectively). Smart insulin pens were used by the fewest number of people with diabetes across all HCP settings. CGM was the most frequently used diabetes technology by people with diabetes across all disciplines.

Insulin regimens used were different by practice type. More insulin-requiring patients in pediatric endocrinology practices used carbohydrate counting and correction factors, with 91.5% of pediatric endocrinologists, 15% adult endocrinologists, and 10% of primary care HCPs estimating that ≥50% of the insulin-requiring people with diabetes in their practice used this method. Fixed insulin dosing (± sliding-scale correction) was commonly used in primary care and adult endocrinology settings, with 25.5% of primary care and 29.1% of adult endocrinology HCPs estimating that ≥50% of the insulin-requiring people with diabetes in their care are prescribed this regimen. All pediatric endocrinologists estimated that <50% of the insulin-requiring people with diabetes in their care were using this regimen.

Respondents were asked to choose one of three described personas with which they most identified (Figure 1). Among primary care HCPs, 57% self-identified with the Cautious persona, 29.8% with the Ready persona, and 13.2% with the Not Yet Ready persona. Most pediatric endocrinology HCPs (73.5%) and adult endocrinology HCPs (62.8%) identified with the Ready persona, with <10% of each discipline identifying with the Not Yet Ready persona.

Pediatric and endocrinology HCPs reported higher comfort overall with diabetes devices and device tasks than primary care HCPs (P <0.001 for all comparisons) (Table 4). HCPs across all disciplines rated their comfort with CGM highest (median 7 out of 10 for primary care and 10 out of 10 for both pediatric and adult endocrinology). Checking insurance coverage and training patients to use the device were the least comfortable aspects of working with CGM across all disciplines. Overall comfort levels with smart insulin pens were low in primary care (3.5 out of 10) and moderate in pediatric and adult endocrinology (7 out of 10 for both). Overall comfort levels for insulin pumps were slightly higher in primary care (4 out of 10) and much higher in pediatric and adult endocrinology (10 out of 10 for both), with the lowest comfort scores for checking insurance coverage and training patients on the devices. The largest difference in comfort levels across disciplines were for HCL systems, with primary care HCPs rating their comfort level as 2 out of 10, pediatric endocrinology HCPs rating their comfort level as 10 out of 10, and adult endocrinology HCPs rating their comfort level as 9.5 out of 10.

For all three HCP disciplines, the perceived usefulness of a device insurance coverage tool was ranked with the highest median score (VAS score: 90 out of 100 for primary care, 95 for pediatric endocrinology, and 99 for adult endocrinology), followed by an online data platform to access device information (85, 91.5, and 88 out of 100, respectively; Figure 2). Primary care HCPs ranked insulin dosing support and device troubleshooting modules as the next most desirable tools (83 out of 100 for each). Both pediatric and adult endocrinologists ranked troubleshooting modules (80.5 and 88 out of 100, respectively) and patient education modules (75 and 80 out of 100, respectively) as the next most useful tools. Primary care HCPs ranked all potential tools with average usefulness scores ≥79 out of 100, whereas pediatric endocrinology HCPs ranked five tools lower than this, and adult endocrinologists ranked three tools lower.

HCPs assigned a high level of importance to having insulin dosing and glycemic information from diabetes devices accessible directly from the EHR (83 out of 100 for primary care HCPs, 81 out of 100 for pediatric endocrinology HCPs, and 78.7 out of 100 for adult endocrinology HCPs). They additionally agreed that it was important for support tools to be accessible from EHR systems (80, 75, and 71 out of 100, respectively) and that it was important for device-related information to be communicated to patients directly via the EHR (78, 73, and 76 out of 100, respectively).

As innovation and supportive evidence for diabetes technology devices continues to mount (1–3,6,7), access to technology for all people with diabetes becomes a primary concern. To ensure that all people with diabetes have access to these devices, innovative care models are needed to empower HCPs to confidently promote their use. This is especially important because most people with diabetes receive their diabetes care outside of diabetes specialty centers (14,15).

We sought to understand the comfort levels of pediatric and adult endocrinology and primary care HCPs, as well as the types of tools that could facilitate use of diabetes technology. Our data indicate important similarities and differences related to diabetes devices among HCPs working in the three disciplines. Although HCPs care for different proportions of adult and pediatric patients with type 1 or type 2 diabetes, the respective disciplines all endorsed the need for tools to facilitate increased use of diabetes devices in their clinical practices. The highly endorsed tools across disciplines were related to insurance assistance and an online diabetes device platform. Furthermore, device data integration with EHRs was considered highly important.

It is notable that 29.8% of primary care HCPs self- identified with the Ready technology persona and that 57.0% identified as Cautious, with only the remaining 13.2% identifying as Not Yet Ready for diabetes technology. In the 2018 original study on clinician personas by Tanenbaum et al. (18), only 20.1% endorsed being ready for technology, and 40.7% endorsed being ready but cautious. The results cannot be directly compared because the study personas were computed from survey items in the original study, whereas they were self- identified by participants after reading persona descriptions in this study. Nonetheless, our study indicates a generally favorable perception of working with diabetes devices in primary care. These are encouraging data that may indicate that, with additional tools and workflow support, primary care clinicians will increase their use of diabetes devices. This expansion of diabetes technology in primary care should take place in parallel with efforts to expand the endocrinology workforce to increase specialty access for people with diabetes.

With regard to comfort with devices, the HCPs indicated the highest overall comfort level with CGM across all disciplines. This finding is not surprising given the direct-to-consumer marketing that has greatly contributed to increased CGM awareness. In addition, HCP education and resources surrounding CGM have been developed by many professional organizations including the American Diabetes Association (ADA) (25), the Association of Diabetes Care and Education Specialists (ADCES) (26), and the American Academy of Family Physicians (AAFP) (27), as well as academic projects such as DiabetesWise (diabeteswise.org) and the Panther Program (pantherprogram.org). Moreover, CGM systems are approved for user self-starting and do not require special training from clinical staff or manufacturer representatives before initiation. As expected, pediatric and adult endocrinology HCPs endorsed having greater comfort using HCL systems than did primary care HCPs. Across all devices, given the time and training needed to successfully manage device use by people with diabetes, clinics should leverage team members such as DCESs and find collaborative ways to achieve core diabetes technology competencies (28). Educational resources can help to expand role competencies to increase the number of HCPs willing to promote diabetes devices and thus expand access.

The most desirable tools identified by HCPs (health insurance assistance and a data capture platform) currently exist in some forms. Device manufacturers often supply resources for determining insurance coverage, but the resources are device specific and often do not provide detailed information. Organizations such as ADCES and DiabetesWise offer more generic assistance for a multitude of devices. Each device company also offers a device-specific online platform to review data, and some additional data-aggregating platforms exist as well. It is possible that the existence of multiple platforms is a substantial barrier for HCPs who would rather learn one system that covers a broad array of diabetes devices. Furthermore, our study also confirms the importance of EHR integration for device data systems, which has been recognized as an essential need for simplifying workflows and increasing HCP use of devices (17,29). Efforts are ongoing through the iCoDE Project to develop standardized recommendations for CGM integration into EHRs (30).

The American Association of Clinical Endocrinology’s clinical practice guidelines suggest that diabetes technology should be managed by HCPs who are “trained, committed, and experienced” in using these devices in their clinical practice and that “ . . . clinicians should have the infrastructure to support the needs of persons with diabetes using the technology” (31). One way to achieve this goal could be to bundle the workflow facilitators and tools described in this article into an all- in-one infrastructure: an online system that could assist with insurance coverage and device ordering, host a single online platform for several devices, and additionally provide device selection assistance, diabetes education, troubleshooting help, and automated insulin dosing support. With fewer software systems to learn, passwords to remember, and programs to operationalize, HCPs may be able to streamline their workflow enough to meaningfully engage with diabetes devices as part of routine care for people with diabetes. Ideally, such a system either would be integrated directly into EHR systems or could be launched directly from within EHR systems, which could further reduce complexity and increase uptake. Novel innovation models such as this have yet to be created and tested in either the endocrinology or primary care settings. Because needs will vary by specialty, individual HCP, and even by device type, it will be important for any infrastructure solutions to be customizable for each practice.

A strength of this study is the large number of primary care and endocrinology HCPs representing a variety of practices and practice settings across the United States. The HCPs had a wide range of experience with people with diabetes and with diabetes technology, making them a diverse sample from which to draw conclusions.

There are limitations as well. Because the sample was obtained in part through social media posting and other forms of advertisement, it is unknown how representative this sample is of the larger population of primary care and endocrinology HCPs. It is possible that HCPs with a higher baseline interest in diabetes technology were more likely to respond to a survey about diabetes technology and that these results may not be reproducible or representative of the diabetes care field as a whole. Results are specific to HCPs practicing within the United States. Additionally, our largely White, non-Hispanic sample may not represent unique technology-related challenges for HCPs who identify as non-White and/or Hispanic. Data were largely collected using quantitative scales, and the addition of qualitative interviews could enrich understanding beyond what the scales revealed.

Diabetes devices such as CGM systems, insulin pumps, and HCL systems have an established and growing evidence base showing that they improve clinical outcomes in people with diabetes, but their availability is largely limited to individuals seeking care at diabetes specialty centers. To decrease prescribing disparities and expand technology use more broadly beyond diabetes centers, new and innovative care models and tools are needed. This study sheds light on what may be the most desirable tools to develop, including insurance assistance and online data platforms. Additionally, our findings highlighted the importance of integrating device data into EHRs. Further work should focus on the development of these tools and their integration into EHRs to expand access to diabetes devices for all people with diabetes who may benefit from them.

This study was sponsored by the Leona M. and Harry B. Helmsley Charitable Trust (grant G-2206-05306).

L.H.M. has received research/education grants from Beta Bionics, Insulet, the Leona M. and Harry B. Helmsley Charitable Trust, the National Institutes of Health, and Tandem Diabetes Care. She has served as a consultant/speaker for Capillary Biomedical, Dexcom, and Tandem Diabetes Care and on an advisory board for Eli Lilly. As of January 2023, she is also employed by Tandom Diabetes Care. H.K.A. has received research grants from Dexcom, Eli Lilly, the Institute for the Advancement of Food and Nutrition Sciences, IM Therapeutics, Mannkind, Medtronic, REMD Biotherapeutics, and Senseonics. He has served as a consultant for Eli Lilly and on advisory boards for Ascensia and Mannkind. G.P.F. has received research grants from Abbott, Beta Bionics, Dexcom, Eli Lilly, Insulet, JDRF, Medtronic, and Tandem Diabetes Care and has served as a consultant/speaker for Abbott, Beta Bionics, Dexcom, Insulet, Eli Lilly, Medtronic, and Tandem Diabetes Care. S.M.O. has received research grants from the AAFP, the Colorado Department of Public Health & Environment, and the Leona M. and Harry B. Helmsley Charitable Trust. He has served on advisory boards for the ADA, ADCES, Cecelia Health, Dexcom, the Jaeb Center for Health Research, Medtronic, the National Committee on Quality Assurance, and Pendulum Therapeutics. S.P. has received research grants from Dexcom, Eli Lilly, JDRF, the Leona M. and Harry B. Helmsley Charitable Trust, Medtronic, the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), and Sanofi and has served on an advisory board for Medtronic. V.N.S. has received research grants from Dexcom, Eli Lilly, Insulet, JDRF, the National Institutes of Health, Novo Nordisk, and Tandem Diabetes Care. He has served as a consultant/speaker for Dexcom and Insulet and on advisory boards for DKSH Singapore, LifeScan, and Medscape. R.P.W. has received research grant support from Dexcom and Tandem Diabetes Care and has served as a consultant/speaker for Tandem Diabetes Care. T.K.O. has received research grants from Abbott, the AAFP, the Leona M. and Harry B. Helmsley Charitable Trust, the National Institute of Nursing Research, and NIDDK. She has served on advisory boards for ADA, ADCES, Cecelia Health, Dexcom, and the Jaeb Center for Health Research. No other potential conflicts of interest relevant to this article were reported.

L.H.M. and T.K.O. designed the study, researched data, and wrote the manuscript. T.V. and L.P. designed the study, analyzed the data, wrote the analysis section of the manuscript, and reviewed/edited the full manuscript. H.K.A., G.P.F., K.B.H., A.J.K., E.M., S.M.O., S.P., V.N.S., and R.P.W. designed the study, researched data, and reviewed/edited the manuscript. L.H.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Data reported in this article were presented, in part, in poster form at ADA’s 82nd Scientific Sessions in New Orleans, LA, 3–7 June 2022.