Abstract
Background: Telemedicine can improve access between physicians and patients and improve outcomes when deployed strategically in patients with chronic diseases. Telemedicine not only showed success in the care of chronic diseases, but its application also expanded exponentially during the COVID-19 pandemic. At our institution, a 12-week telemedicine diabetes “boot camp” was launched for patients with uncontrolled diabetes as an innovative means of providing accessible and high-quality patient care in primary care settings.
Methods: Patients at primary care and endocrinology clinics with diabetes mellitus (DM) and glycohemoglobin (A1C) > 8.0% were voluntarily enrolled from September 2020 to November 2021. Dietitians and diabetes care and education specialists conducted biweekly visits via telemedicine for twelve weeks. Patient demographics, A1C, body mass index (BMI), and blood pressure were measured before and after the intervention.
Results: A total of 134 patients were included, and 94 patients (70.2%) completed 6 visits for the full 12-week program. The mean A1C reduction was –2.09% ± 2.4%, and the A1C change was uniform across age groups, gender, ethnicity, BMI, and referral clinic type. A greater A1C reduction in patients who completed all 6 visits was noted although not statistically significant. We found a negative correlation between the initial A1C and the change of A1C. No significant BMI or mean arterial pressure change was observed.
Conclusion: This single arm study demonstrated an improvement in A1C for all patients with a history of poorly controlled diabetes, regardless of patient characteristics. Higher initial A1C was associated with a greater A1C reduction.
- Diabetes Mellitus
- Electronic Health Records
- Glycemic Control
- Patient Education
- Primary Health Care
- Telemedicine
Introduction
Diabetes mellitus is a leading cause of chronic kidney disease, blindness, heart attack, stroke, and lower limb amputation, and is associated with high health care costs.1 Diabetes self-management education and support (DSMES) programs, which include psychological and behavioral components, are fundamental to optimizing glycemic control. Finding the best setting for effective diabetes education has always been difficult but when the COVID-19 pandemic caused the cessation of most in-person visits, we needed to quickly develop a new method to support our patients with diabetes. Our solution was the creation of a twelve-week telehealth diabetes “boot camp” featuring interactive, personalized counseling by a dietician and a diabetes education specialist. The goal was to provide education, support, and regular feedback via telemedicine to patients with poorly controlled diabetes, and to evaluate the effect of our telemedicine intervention on glycemic control in this population.
Methods
Design
This is a prospective, single arm study of patients enrolled in the boot camp between September 1, 2020, and November 30, 2021. Patients with a diagnosis of chronic type 1 or 2 diabetes mellitus and a recent A1C > 8.0% were eligible. Patients with active unstable illness were excluded. Eligible patients who gave verbal consent were enrolled in the program via an electronic referral built into the electronic health record (EHR). Patients were evaluated by a physician before and after the 12-week program. The study protocol was approved by the institutional review board.
Intervention
Patients logged in through the MyChart patient portal (Epic Systems Corporation, Verona, WI, USA) to access the EHR-integrated telemedicine platform Twilio (Twilio Inc, San Francisco, CA, USA) for one 30-minute visit every other week, with visits alternating between a dietician and a diabetes care education specialist. Before each visit, patients uploaded glucose data from their glucose monitoring devices. During the visit, the dietician or diabetes education specialists reviewed the patient’s uploaded glucose trends and discussed diet and exercise plans and lifestyle modifications. The diabetes care and education specialists used a predesigned EHR template to guide decisions regarding antihyperglycemic regimens.
The patients met with their physician after the twelve-week intervention to assess changes. A postintervention satisfaction survey with blinded response envelopes was sent via postal mail to patients who had attended at least one visit.
Metrics
We collected demographic data including age, gender, and race. The number of visits attended, with 6 visits considered full participation, was also noted. The values for A1C, BMI, and mean arterial pressure (MAP) were recorded at enrollment and again after twelve weeks. The postintervention survey assessed the patient’s compliance with scheduled appointments, their difficulty using the software, their satisfaction with the program, and the time saved by using telemedicine.
Statistical Analysis
We reported continuous variables as mean and standard deviation. Student’s t test was used to analyze the differences for continuous variables and paired t test for the pre and post comparative analysis within the group. The χ2 test or the Fisher exact test were employed to analyze differences for the categorical variables depending on the expected frequencies. Simple linear regression was used to assess the relationship between the predictors and outcomes. A P-value less than 0.05 was considered statistically significant. All analyses were done by SAS 9.4 (SAS Institute, Cary NC).
Results
Of the 711 patients electronically referred for the program, 168 (23.6%) initiated at least one telemedicine visit. Our final analysis included 134 patients. The remaining 34 were excluded due to lack of postintervention data. Reasons for nonparticipation included the lack of internet access, insurance coverage, or interest. Ninety-four patients completed 6 or more visits. Many voluntarily continued to participate after the initial 12-week program. The highest number of visits recorded was seventeen. Participants were mostly over 40-years-old (116/134, 86.6%), female (78/134, 58.2%), and white (77.6%, 18.7% African American, 3.7% other). Seventy-three patients (54.4%) were referred to the program by an endocrinologist and 61 (45%) were referred by a primary care physician.
Improved Diabetes Control
The overall mean A1C declined from 10.69% to 8.60% (Table 1). The reduction in mean A1C was seen regardless of age, gender, ethnicity, or initial BMI, with the greatest reduction seen in patients with higher A1C levels at the start of the program. Patients who completed 6 virtual bootcamp visits experienced a mean A1C change of −2.33%, while patients who participated in 1 to 5 visits had a mean change of −1.54% (Table 1).
Subgroup Analysis of the Changes of A1C, BMI, and MAP
Patients in our study showed a nonsignificant BMI change and a marginally significant reduction in MAP. Patients referred to the program from a primary care clinic showed larger reductions in BMI than those referred by an endocrinologist, but the overall BMI change was not statistically significant (−0.12, SD 2.10, 95% CI, P = .5172) (Table 2). Patients with BMI>=40 demonstrated the most significant MAP reduction - from 101.09 to 95.11 mmHg (mean difference −5.98, SD 14.6, 95%CI, P = .0143) - but when the population was considered as a whole, the mean change was not statistically significant (−2.42, SD 14.5, P = .0644) (Table 3).
The Factors Associated with the Changes of A1C, BMI, and MAP
Patient Satisfaction
Of the 134 surveys sent, 45 (34%) were returned. Thirty-two patients (71%) completed the boot camp, 9 patients (20%) did not. The information was missing for 4. Reasons for not completing the program included being unable to take time off work, difficulty with appointment scheduling, the repetitiveness of the visits, and the achievement of good glycemic control. Most patients were satisfied with the experience and most said they had saved time by participating in virtual – rather than in-person – visits.
Discussion
Telehealth has been reported to have a positive effect on DSMES by strengthening the self-management feedback loop and increasing communication with diabetes care teams,2 a finding confirmed by our study. The implementation of our telemedicine intervention during the pandemic significantly benefited both type 1 and type 2 diabetic patients with poor glycemic control.
Faruque et al. found that telemedicine interventions, especially those that allow medication adjustments similar to our model, are associated with greater A1C reduction,3 however the optimal amount of interaction has been difficult to assess. In our boot camp, the scheduled education and counseling enabled frequent patient interactions with the instructors. Whether scheduling more frequent visits with the poorly controlled diabetes population will lead to better glycemic control can be studied further.
Our study, where a higher preintervention A1C was associated with a significantly higher degree of A1C reduction, supports the findings of Wu et al, who conducted a meta-analysis of 19 randomized controlled trials of DSMES with telehealth implementation, and found that the A1C reduction was −1.22% in those with baseline A1C ≥9%, while patients with initial A1C < 9% only achieved an average reduction of −0.35%.4
DSMES on BMI and Blood Pressure
The effect of DSMES on BMI or blood pressure is not evident in the literature.5,6 Studies have shown inconsistent results for BMI and weight-related secondary outcomes.7⇓–9 Patients in our study showed a borderline significant reduction in MAP and a nonsignificant BMI change, which is consistent with previous studies. Patients with morbid obesity had the most significant blood pressure improvement despite no significant BMI changes. We also found the female population and those who completed the boot camp had substantial improvement in blood pressure. These subpopulations may be more compliant, or this education style may be more effective for them. Discovering a suitable educational style for different subgroups would be a valuable topic for future discussion.
Conclusion
Integrating the virtual program into existing clinical workflows was a challenge that required careful planning to overcome. The program required sufficient technological infrastructure, including electronic health records integration, remote monitoring capabilities, and patient-friendly telemedicine platforms. Identifying specific objectives for the telemedicine visits played a pivotal role in formulating the workflow. Our boot camp curriculum is available on request. Maintaining patient engagement revealed challenges related to patient motivation, education, and self-management. We provided training and readily available technical support. Still, technical issues excluded those who may have benefited but were unable or unwilling to use the technology. Insurance and reimbursement issues for telemedicine related to diabetic education remain hurdles.
Our results are promising. This telehealth DSMES model is effective in glycemic control and could be widely applied to patients with elevated levels of A1C who have poor compliance or poor access to medical resources. Both patients and the health care team require comprehensive, ongoing technical support, which must be in place before the program starts. Future study could focus on finding the suitable educational model for each subpopulation and to investigate whether the frequency of education is the key to patient interaction and success in glycemic control.
Acknowledgments
The authors express gratitude to Vanessa Snell, RN, BSN, and Erica Hornung, RD, for coordinating virtual diabetes and nutrition education teams. We would also like to thank Uba Chinyere Udeh, MD, Hassan Choudry MD, Saddam Hussain Abbasi, MD Kripa Rajak, MD, and Helen Houpt for their project help and support.
Notes
This article was externally peer reviewed.
Funding: None.
Conflict of interest: The authors have no conflicts of interest to disclose.
To see this article online, please go to: http://jabfm.org/content/38/3/556.full.
- Received for publication June 1, 2024.
- Revision received November 23, 2024.
- Accepted for publication December 9, 2024.
