Glucose monitoring is essential for the management of type 1 diabetes and has evolved from urine glucose monitoring in the early 1900s to home blood glucose monitoring in the 1980s to continuous glucose monitoring (CGM) today. Youth with type 1 diabetes struggle to meet A1C goals; however, CGM is associated with improved A1C in these youth and is recommended as a standard of care by diabetes professional organizations. Despite their utility, expanding uptake of CGM systems has been challenging, especially in minoritized communities. The 4T (Teamwork, Targets, Technology, and Tight Control) program was developed using a team-based approach to set consistent glycemic targets and equitably initiate CGM and remote patient monitoring in all youth with new-onset type 1 diabetes. In the pilot 4T study, youth in the 4T cohort had a 0.5% improvement in A1C 12 months after diabetes diagnosis compared with those in the historical cohort. The 4T program can serve as a roadmap for other multidisciplinary pediatric type 1 diabetes clinics to increase CGM adoption and improve glycemic outcomes.
Glucose monitoring is a key component of diabetes care (Figure 1). The first attempt at glucose monitoring was the introduction of urine glucose monitoring in 1908 (1). Urine testing remained the standard of care until the introduction of home blood glucose monitoring (BGM) in the 1980s. The ability to closely monitor glycemia using BGM helped to revolutionize the care of diabetes to more precisely dose insulin. We are now undergoing another paradigm shift in glucose monitoring. The U.S. Food and Drug Administration (FDA) approved the first continuous glucose monitoring (CGM) system in 1999. Early-generation CGM systems were blinded devices from which data were downloaded in health care providers’ clinics.
Evolution of glucose monitoring and recommendations as standards of care (SOCs).
Evolution of glucose monitoring and recommendations as standards of care (SOCs).
CGM systems now have evolved to obtain interstitial glucose values every 1–15 minutes, display these readings on a receiver, and sound alarms to alert users to hypoglycemia and hyperglycemia. Early CGM systems required calibration with glycemic levels obtained through BGM. In 2016, the first factory-calibrated CGM system was introduced. Modern CGM devices can passively share users’ glycemic data to other individuals and to their diabetes clinic through the use of smartphone apps and Cloud-based data platforms. Because CGM has been shown to improve A1C and quality-of-life measures in people with type 1 diabetes (2–9), the American Diabetes Association (ADA) now recommends CGM for all individuals with type 1 diabetes, and the International Society for Pediatric and Adolescent Diabetes (ISPAD) recommends CGM use for youth with type 1 diabetes (10–12).
As glucose monitoring technology has evolved, so too has insulin delivery technology (13). Insulin was first administered by injection in 1922. The first commercial insulin pump was introduced in 1979, and insulin pens were introduced in the 1980s. At that time, conventional insulin therapy consisted of one to two injections per day of mixed intermediate- or rapid-acting insulin, with glucose levels monitored through urine testing or BGM. In 1993, the Diabetes Control and Complications Trial (DCCT) reported its findings that intensive insulin therapy, consisting of four or more insulin injections per day or insulin pump therapy to provide both basal and bolus insulin, decreased microvascular complications of diabetes, and intensive insulin therapy became the new standard of care for type 1 diabetes (14,15). With the goal of intensive glycemic management, diabetes technology rapidly advanced to today’s state-of-the-art automated insulin delivery (AID) systems, which combine CGM with an insulin pump and a control algorithm to deliver insulin automatically based on real-time glucose levels and trends. Currently, there are several FDA-approved hybrid closed-loop AID systems (automating basal insulin but requiring user input to deliver bolus doses) on the U.S. market. Because AID has been shown to improve glycemia and quality of life (16–24), the ADA and ISPAD as of 2022 now recommend the use of AID systems (with a level A evidence grade, indicating clear evidence from well-conducted, generalizable randomized controlled trials that are adequately powered) for people with type 1 diabetes (25–27).
Despite the findings from the landmark DCCT and subsequent advances in diabetes technology, people with diabetes, especially those in the pediatric age-group, continue to struggle to meet A1C targets (28–30). A1C typically exceeds glycemic targets soon after diabetes diagnosis (31,32), and this can have long-term implications; data suggest that hyperglycemia early in the course of diabetes is strongly associated with long-term glycemic control (33). In the DCCT, tight glycemic control was achieved in people with type 1 diabetes in part as a result of frequent insulin dose adjustments made for study participants by their care team. Unfortunately, this aspect of the DCCT intervention still has not been implemented broadly even more than 30 years later.
In post-DCCT clinical practice, accessing BGM data outside of clinical visits has been limited, and insulin dose adjustments have tended to occur only at quarterly clinic visits. With the advent of modern CGM technology, glucose data can be shared passively via Cloud-based platforms. The availability of these platforms decreases the data-sharing barrier to expanding remote patient monitoring (RPM); however, having sufficient staff to perform RPM remains a challenge. Without a way to identify individuals who would most benefit from data review, RPM would be difficult to scale without expanding clinic resources.
In the 4T (Teamwork, Targets, Technology, and Tight Control) study, we adapted established methods to better manage glycemia in the first year after type 1 diabetes diagnosis to improve long-term outcomes (34–43; J.C. Leverenz, B. Leverenz, P.P., et al., unpublished observations). Initiating CGM in the first month after diagnosis was a cornerstone of this program. The 4T study used findings from the Hviodere study and others, which showed that teamwork and consistent setting of tight glycemic targets can improve clinical outcomes (44). We intensified education in the new-onset period based on the rationale that this period is when people are most open to and in need of education and that this early time investment would result in better long-term outcomes. We also aimed to implement the DCCT intervention of making frequent dose adjustments by developing a sustainable, asynchronous RPM program (41–43). This effort encompassed the principle of equitable access independent of insurance or language status (35) because, historically, the introduction of new technology has increased disparities in glycemia as a result of inequitable access (45).
Roadmap to Early CGM Initiation: The 4T Study
Before the start of the 4T study, youth with newly diagnosed diabetes had a 4- to 6-hour outpatient new-onset education visit with the clinical team to learn about diabetes care (Figure 2A). That visit was followed up by a 1-month recent-onset visit and routine quarterly diabetes visits during which dose adjustments occurred. There was no standardized approach to discussing or initiating diabetes technology (e.g., CGM or an insulin pump).
New-onset type 1 diabetes management program with standard care (A) and in the 4T study protocol (B). T1D, type 1 diabetes.
New-onset type 1 diabetes management program with standard care (A) and in the 4T study protocol (B). T1D, type 1 diabetes.
The 4T study introduced a team-based approach to diabetes care to intensify new-onset education, standardize early technology access and glucose targets, and increase touchpoints via RPM. The protocol for the 4T study has been previously described (11,37). Briefly, after routine new-onset education, youth are offered the opportunity to start on CGM within the first 30 days of diabetes diagnosis (Figure 2B). An initial month of supplies is provided by the 4T study, and subsequent CGM supplies are submitted for coverage through the youth’s insurance plan. Many diabetes programs have access to starter CGM materials, which could be used to initiate CGM early and support individuals while they await insurance approval for CGM. Those who elect to start on CGM have a follow-up visit with a certified diabetes care and education specialist (CDCES) to start CGM, with an additional CGM follow-up visit 1 week later. Youth return for a 1-month recent-onset visit, followed by quarterly clinic visits. Between clinic visits, CGM data are reviewed weekly, and insulin dose adjustments are communicated to families through electronic health record (EHR) system–based secure portal messaging (Figure 3). During the pilot 4T study, we used an A1C target of <7.5%, which was the ADA’s recommended target in 2018 (46). Youth in the pilot 4T cohort, who were diagnosed from July 2018 to June 2020 (n = 135), had an improvement of 0.5% in A1C compared with our clinic’s historical cohort, who had been diagnosed from June 2014 to December 2016 (n = 272) (36).
The 4T program has continued to evolve since the pilot phase. In 4T study 1, which has completed enrollment, the A1C target was lowered to <7%, and patient reported outcomes (PROs) and exercise education were incorporated (47). Participants in 4T study 1 were diagnosed from June 2020 to March 2022 (n = 133). In the currently enrolling 4T study 2, we are encouraging early AID adoption by standardizing a pump/AID class within the first 3 months of diabetes diagnosis and further tightening targets by lowering the A1C goal to <6.5% with associated glucose targets.
Implementation of the 4T Program
Teamwork is the foundation of the 4T program. Before rolling out the program, the clinical diabetes team, including physicians, nurse practitioners, CDCESs, registered dietitians, and social workers, participated in planning sessions to map out the program (J.C. Leverenz, B. Leverenz, P.P., et al., unpublished observations). Once the group had agreed on a rollout that would include early CGM initiation, RPM, and consistent glucose targets in the new-onset period, youth were enrolled. Since implementing the 4T program in July 2018, the diabetes team has held routine team meetings to share findings and fine-tune the program through an iterative process. The 4T program added RPM time to the CDCES team, but this was offset by hiring a pharmacy technician to whom to offload appropriate tasks (e.g., prior authorizations) to allow CDCESs to perform CGM data reviews, make insulin adjustments, and work at the top of their professional scope (48).
Development of the RPM Program
Starting in March 2019, youth enrolled in the 4T study were offered the opportunity to receive RPM. Initially, CGM tracings were reviewed from the manufacturer’s website for every patient, and a CDCES would contact families by secure portal messaging for insulin dose adjustments or additional education. This procedure quickly became unsustainable, so we collaborated with engineers in the Systems Utilization Research Force Stanford Medicine team to develop a system for prioritizing youth who would benefit from closer data review and dose adjustments. Our first iteration was an R-based tool that would run on a clinician’s laptop and use CGM consensus guidelines (5) to flag youth who would benefit from closer review (43). The 4T investigators and CDCES team made iterative changes to flags and have now developed a dashboard on Tableau, a software for data visualization. This dashboard, called Timely Interventions for Diabetes Excellence (TIDE), prioritizes youth for closer review and allows for review of CGM data within the dashboard (Figure 4). As the program has grown from one to five CDCESs, >400 youth have now participated in RPM. Compared with reviewing each CGM tracing individually, use of the TIDE dashboard decreased review time by 60% (42). The programming code for TIDE is open source and can be deployed at other clinics.
Roadmap: Promoting Equity in the 4T Study
Individuals on public insurance and those from minoritized groups have higher A1C levels and higher rates of diabetes-related complications than non-Hispanic White individuals and those from higher income groups (29,49–51). Although many socioeconomic and racial factors contribute to this disparity (52), technology access is one key component if it is preferentially available to individuals of higher socioeconomic status. Insulin pumps are typically covered by public insurance, but there is variability in CGM coverage in the United States. Youth from underserved groups have lower use of diabetes technology (45,50,53,54). Although CGM access is universal in other countries (e.g., Australia [55]), the gap has widened in the United States (45). We reported that individuals from minoritized groups have persistent CGM use when CGM is accessible (56) and improvements in A1C when CGM coverage is uninterrupted (2). We observed similar findings in the pilot 4T study. We found similar A1C reductions in 4T participants by insurance status and race/ethnicity, although the program did not eliminate A1C disparities (35), likely because of additional social determinants of health that the 4T program was unable to address. Still, the 4T program offers a roadmap that can be used to achieve equitable introduction of diabetes technology to avoid increasing disparities. An article by Ebekozien (57) in this From Research to Practice section provides insights on improving CGM equity from the T1D Exchange Quality Improvement Initiative.
In 2018, when we started the 4T program, CGM coverage was unpredictable for youth on public insurance. We were able to obtain first philanthropic and then research funding to provide ongoing equal access to CGM and a safety net for individuals with insurance disruptions. We used the data from the pilot 4T study and other studies to work with California Children’s Services leadership to advocate for CGM coverage for all youth with public insurance. CGM coverage is now available for all youth with type 1 diabetes in California.
The 4T program also depends on having access to a smart device from which to share CGM data and access RPM messages through the secure patient portal. To address disparities in access to such a smart device, we provided iPod Touch devices to participants who needed them. Although we could not supply the Internet access our participants needed, we were able to collaborate with schools to allow children access to the school’s Internet service for data-sharing during the school day.
Conclusion
The DCCT demonstrated the benefits of intensive glucose management combined with frequent insulin dose adjustments. Thirty years later, technology has advanced to allow for more frequent glucose measurements via CGM, automated insulin delivery, and now RPM to provide precision medicine approaches on a population level. Clinical implementation of recommendations for diabetes technology as a standard of care should be initiated widely. The 4T program has shown that a team-based approach to early technology initiation and RPM can improve outcomes in youth with newly diagnosed type 1 diabetes (34–40, J.C. Leverenz, B. Leverenz, P.P., et al., unpublished observations).
Care should be taken to ensure that implementation of technology does not increase disparities. Health care providers and patient groups should engage in advocacy efforts to ensure equitable access to diabetes technology. In addition, there should be advocacy efforts to allow for free basic Internet service to facilitate the sharing of medical data. Until that goal is achieved, health care teams can collaborate with local schools to facilitate data-sharing via school-based wireless Internet service. Integrating wifi or cellular functionality directly into diabetes devices would eliminate the need for users to have another smart device for data-sharing. Finally, attention should be taken to ensure that copayments for RPM services are not overly burdensome, resulting in disparities in the use of RPM-based interventions.
An RPM program also needs to be flexible enough to adapt to changing technology and clinical needs. Developing these programs requires a team-based approach, leadership support, and infrastructure to enable clinical teams to provide this service (38). Currently, one of the barriers to streamlining RPM workflows is inaccessibility of data from diabetes devices. Although some device manufacturers have application programming interfaces to retrieve data, many others have made data access difficult. People with diabetes and their caregivers should have the primary voice in data accessibility and data-sharing. In addition, not all devices passively upload data, which can increase the burden on families and may introduce disparities. Clinical teams would like for CGM data to be integrated in the EHR to further streamline data-sharing and reviewing processes. However, such integration to date has only been achieved through custom work for individual institutions (58–60). Interdisciplinary groups such as iCode have developed standards to help guide these efforts (61,62), but a more sustainable solution should be developed. Finally, dashboards that facilitate RPM (e.g., Tidepool+ and Glooko) should have flags that are customizable to an institution’s capacity and to new technology such as AID systems and exercise trackers (47). While RPM can add to the workload of existing teams, developing a sustainable reimbursement model is key to gaining hospital support. Although we do not currently charge for RPM, financial modeling of patients receiving RPM at our institution shows that current RPM billing codes are potential revenue-positive solutions (41).
In summary, we have developed a roadmap to implement CGM and RPM in a pediatric population with new-onset type 1 diabetes with the use of existing technology. Growing this program will require partnerships among clinics, payers, hospital leadership, and industry to improve health equity and care for all people with type 1 diabetes.
