Virtual Diabetes Optimization Program (VDOP) for Veterans With Uncontrolled Diabetes

Abstract

OBJECTIVE

More than 25% of U.S. veterans have diabetes, and 78% have a BMI ≥25 kg/m². Despite access to lifestyle interventions and specialty endocrine care through the Department of Veterans Affairs (VA), many veterans with uncontrolled diabetes and obesity do not achieve sustained glycemic control or weight loss. We developed the Virtual Diabetes Optimization Program (VDOP), a quality improvement initiative targeting veterans with uncontrolled diabetes (glycated hemoglobin A_1c [HbA_1c] ≥8% [≥64 mmol/mol]) with or without obesity.

RESEARCH DESIGN AND METHODS

VDOP used a multidisciplinary team—including a registered dietitian, occupational therapist, and clinical pharmacist—delivering frequent virtual care under physician diabetologist supervision. Outcomes were compared with veterans who declined VDOP and received standard endocrine care during the same period. Primary outcomes were changes in HbA_1c and body weight at 6 and 12 months.

RESULTS

A total of 39 veterans completed VDOP and 62 received standard care. Baseline characteristics were similar, though VDOP participants had higher BMI. After 6 months, VDOP participants had significantly greater mean reductions in HbA_1c compared with those receiving standard care (mean HbA_1c decreased from 9.8% [84 mmol/mol] to 7.1% [54 mmol/mol] vs. 10.6% [92 mmol/mol] to 8.2% [66 mmol/mol]) and greater percent weight loss (−6.8% vs. −0.7%), with these differences remaining significant at 12 months. Differences remained significant after adjustment for baseline values and diabetes medications, including glucagon-like peptide 1 receptor agonists.

CONCLUSIONS

VDOP was associated with greater improvements in glycemic control and weight loss compared with standard endocrine care. Prospective studies are needed to assess long-term outcomes and cost-effectiveness.

Introduction

U.S. veterans are disproportionately affected by type 2 diabetes, with estimated rates of 25%, compared with the national prevalence of 11% (1,2). Among veterans with diabetes, 68% of women and 56% of men are living with obesity (BMI >30 kg/m²) (3). Diabetes costs the U.S. health care system $306.6 billion annually (4,5). Cost of care increases as glycated hemoglobin A_1c (HbA_1c) value rises (6). Adults with an HbA_1c above 8% (64 mmol/mol) are at a significantly increased risk for hospitalization due to diabetes complications and cardiovascular disease (7).

Lifestyle interventions, including a healthy diet and regular physical activity, are considered pillars of diabetes and obesity management. Since 2005, the Department of Veterans Affairs (VA) has worked to address obesity and diabetes prevention through multidisciplinary weight management programs such as the MOVE! Program (8). However, reports evaluating the success of these approaches in terms of weight and HbA_1c reduction in veterans with uncontrolled diabetes have been inconsistent, ranging from no significant improvement to modest benefits (8).

Outside the VA, several multidisciplinary virtual programs have reported significant benefits in terms of weight loss and improvement in diabetes control (9,10). In response, professional diabetes organizations now recommend personalized diabetes technology based on a person’s needs, desires, skill level, and availability of devices as part of the standard of care (11,12).

The VA has a robust telehealth infrastructure, initially developed to increase health care access for veterans facing transportation or geographic barriers. This infrastructure enabled a 3,000% increase in telehealth appointments in response to the COVID-19 pandemic, with more than 2.4 million veterans (40% of the VA population) receiving at least some telehealth care in fiscal year 2023 (13,14).

Our group previously designed and implemented MOVE! Enhanced program, an entirely remote pilot program for veterans with obesity and uncontrolled diabetes (15). After an average duration of 9.9 months, participants significantly reduced their HbA_1c by 1.2% and weight by 10.0%. Building on our previous experience, we designed and developed a quality improvement project called the Virtual Diabetes Optimization Program (VDOP) to leverage these capabilities. In this article, we describe the VDOP model and compare outcomes with conventional endocrine care.

Research Design and Methods

Veteran Selection

Since April 2022, VDOP has been available on a referral basis for veterans with uncontrolled diabetes (HbA_1c ≥8% [≥64 mmol/mol]) who are followed at the Milwaukee VA and are interested in participating in an intensive program involving dietary, physical activity, and medication changes. Although direct referral was an option, most primary care providers referred veterans with inadequate glycemic control for endocrinology consultation. These consults were reviewed by a triage nurse to identify veterans with a qualifying HbA_1c. The nurse then contacted the veterans, provided an overview of VDOP, and invited them to enroll in VDOP. Those who agreed were referred to VDOP. The rest were referred for endocrinology consultation.

The VDOP Model

An overview of the program is shown in Fig. 1.

Figure 1.

VDOP flow chart. VDOP participants begin with technology orientation, followed by visits with a dietitian, OT, and pharmacist in weeks 1, 2, and 3, respectively. At the end of week 3, these clinicians huddle with a diabetologist to discuss the case. This level of appointments and huddles continues for at least three cycles (approximately 3 months). If goals in terms of glycemic control and weight loss have been achieved after the initial three cycles, appointment frequency and medication changes are continued on a less intensive schedule. If goals are not met, this cycle continues for up to six cycles (approximately 6 months).

Technology Orientation

First, veterans received instructions and virtual assistance from a telehealth technology manager to set up VA Video Connect (VVC), the platform used for virtual appointments at the VA. Next, veterans were mailed an Inspire 2 Fitbit (Google, Mountain View, CA) and guided on setting up a Fitbit account and syncing their account to Fitabase (Fitabase, San Diego, CA). Veterans who did not already have a glucose monitoring device received either a continuous glucose monitor (CGM) or a glucometer according to VA guidelines (16). Approximately 33% of VDOP participants used a CGM and the remainder used a glucometer. Available CGMs included the Dexcom G7, FreeStyle Libre 3, and FreeStyle Libre 3+, whereas the Accu-Chek Guide Me was the glucometer used. Patients were then assisted in linking their glucose monitor with one of the cloud platforms used in the program: Glooko (Glooko, Palo Alto, CA), Libre View (Abbott Diabetes Care, Alameda, CA), Dexcom Clarity (Dexcom, San Diego, CA), or VA Share my Health Data (Department of Veterans Affairs, Washington, DC). This allowed providers to access participants’ blood glucose levels remotely. The telehealth technology manager notified the care team once the veteran completed these steps, indicating that weekly appointments with the providers could be scheduled.

Dietitian Visits

Each veteran started VDOP with a 60-min VVC visit with the dietitian—in our case, a registered dietitian who was also a certified diabetes care and education specialist. The veteran and dietitian jointly identified specific long-term goals regarding weight and blood glucose levels and then agreed on a diet to achieve these goals. The dietitian encouraged veterans to log dietary intake so veterans could learn how foods affected their glycemic control. The share-screen feature of VVC allowed the dietitian to use online food labels to review various macro and micronutrients, calories, and portion sizes.

Occupational Therapist Visits

A week after the dietitian visit, the veterans met with the occupational therapist (OT) for a 60-min baseline visit. During the initial visit, the OT used Fitbit data to discuss baseline activity levels. Other points of discussion included barriers to activity, hypoglycemia considerations, Fitbit use, physical activity recommendations per VA guidelines, and goal setting. The Fitbit allowed the veterans and OT to track progress. The OT asked the veterans to achieve 150 min or more of aerobic exercise and two to three sessions of resistance training each week (17). Fitabase collected data on steps, heart rate, activity intensity, calories burned, sleep, and other physical activity metrics, allowing veterans to set more detailed goals than a pedometer would allow. These data helped veterans and the OT provider to review objective metrics rather than relying on participant reports. Fitbits were also used to provide reminders for planned activities.

Pharmacist Visits

After the dietitian and OT visits, the veteran met with the team pharmacist. Before the initial visit, the pharmacist reviewed current and past medications, past medical history, and laboratory data to identify comorbidities that might affect drug therapy. Medication adjustments were permitted throughout the intervention and were made by the clinical pharmacist under the supervision of a physician diabetologist. Medication changes included initiation, discontinuation, dose adjustment, and regimen simplification based on glycemic control, comorbidities, medication tolerability, and participant goals.

Huddle

After the three initial visits, the multidisciplinary team huddled via Microsoft Teams (Microsoft Corp., Redmond, WA) with the physician diabetologist. During the huddle, the team reviewed and discussed veteran glucose profiles, medications, physical activity, and diet. This weekly meeting addressed opportunities for intervention and health barriers VDOP veterans were experiencing. The team worked together to develop and provide personalized diet, physical activity, and medication plans.

The veterans continued 30-min alternating weekly visits with the dietitian, OT, and pharmacist. The team discussed the participants every 3 weeks during the team huddle or more frequently if needed. Between weeks 12 and 24, HbA_1c and weight were reviewed; the team discharged veterans back to their primary care team with a one-on-one discussion with the provider once individualized goals for optimization were met. In addition, participants were encouraged to join a monthly peer-support virtual session during which the same VDOP clinicians review different educational topics and hold interactive discussions with past VDOP graduates.

Evaluation of VDOP

The VDOP was deemed a quality improvement project by the Milwaukee VA Institutional Review Board; therefore, participants did not sign informed consent documents. To evaluate the effectiveness of VDOP, we compared veterans who started enrollment on 1 April 2022 and completed VDOP by 31 December 2023 with veterans who were referred to an endocrinologist and had a qualifying HbA_1c but who had opted out of VDOP during the same period (hereafter, the endocrine group). We required that veterans in the endocrine group attend at least two visits with the endocrine or diabetes specialists.

We queried the VA Corporate Data Warehouse to obtain demographic, clinical, and laboratory data of VDOP and endocrine group participants. We identified comorbidities from encounters and problem lists using ICD-10 codes.

After participating in VDOP, participants received an e-mail with a link to a unique, program-specific anonymous survey (Supplementary Material). This eight-question multiple-choice survey asked veterans to rate their satisfaction with VDOP, as well as their perceived quality of life before their first VDOP visit and at the time of the survey. Additional questions assessed their ability to adhere to dietary, exercise, and medication goals; their preference for virtual visits; and their willingness to recommend the program to other veterans. We did not validate the survey or survey the endocrine group.

Variable Definitions

We defined baseline values for weight and HbA_1c as the values closest to and within 6 months before the first visit to a VDOP clinician or an endocrine or diabetes specialist. We defined the 6-month value as the measurement obtained between 3 and 9 months that was closest to 6 months. Finally, we considered the value closest to 12 months obtained between 9 and 15 months after their first appointment as the 12-month value. Weight change and percent weight change were calculated only for participants with available follow-up data at each time point. Percent weight change was calculated for all participants, using their corresponding baseline weight and then averaged across participants. Age at the time of the first clinical visit was reported. To control for the level of medical complexity, the Charlson comorbidity index was determined for each group.

Medications to treat diabetes were identified and categorized by when they were started and by their class: glucagon-like peptide 1 (GLP-1) receptor agonists, sodium–glucose cotransporter 2 (SGLT2) inhibitors, sulfonylureas, metformin, insulin, pioglitazone, and dipeptidyl peptidase 4 (DPP-4) inhibitors.

Statistical Analyses

We performed the following descriptive and comparative analyses of baseline characteristics across groups as appropriate: χ² or Fisher exact test for categorical variables and one-way ANOVA for continuous variables. We compared percent change in HbA_1c and weight at 6 and 12 months using simple t tests. We then used general linear models to adjust for potential confounders. Outcome variables were also observed at 6- and 12-month follow-up visits via multivariable stepwise regression models; continuous outcomes were observed using general linear models, and binary outcomes were observed using logistic regression. Variables for these latter analyses were selected a priori and included the assigned group.

Multivariable models included group designation (VDOP or endocrine care), demographics (age, sex, and race), initiating new diabetes medications (GLP-1 receptor agonists, SGLT2 inhibitors, metformin, sulfonylureas, insulin, pioglitazone, and DPP-4 inhibitors), and baseline values of the outcome variable being analyzed (weight, BMI, and HbA_1c). Logistic regression and general linear models were also formed to evaluate whether group designation, in conjunction with these other factors, predicted health care utilization as defined by emergency department (ED) visits, hospitalizations, and surgeries during the post-intervention period. Duration of diabetes was not reliably available in the data set and therefore was not included in the analyses. Statistical significance was defined as a two-sided P ≤ 0.05. All analyses were performed using SAS, version 9.4.

Data and Resource Availability

The data sets generated and/or analyzed during this study are not publicly available, due to federal privacy laws, including The Privacy Act of 1974 and Health Insurance Portability and Accountability Act regulations. However, deidentified data may be available from the corresponding author upon reasonable request.

Results

Between April 2022 and December 2023, 39 veterans completed VDOP (the average duration was 4 months), and 62 veterans established care in the endocrine clinic. Table 1 lists baseline group characteristics. Veterans in the VDOP and endocrine groups had similar demographics, being mostly male (90% and 94%, respectively), of White non-Hispanic race (85% and 71%, respectively), and with an average (±SD) age of 64 ± 9 and 65 ± 11, respectively. Medical complexity, as determined using the Charlson comorbidity index, was also not different between groups. Baseline HbA_1c was not significantly different between the VDOP and endocrine groups (mean HbA_1c of 9.8% [84 mmol/mol] vs. 10.6% [92 mmol/mol], respectively; P = 0.076); however, veterans in the endocrine group had a lower baseline mean BMI (32.4 vs. 36.1 kg/m² for the VDOP group; P < 0.05). New diabetes medications initiated during the intervention period in the VDOP and endocrine groups are summarized in the Supplementary Material.

Table 1.

Baseline characteristics

Variable	Endocrine group (n = 62)	VDOP group (n = 39)	P value
Male sex, %	94	90	0.227
White race, %	71	85	0.260
Age, years	65 ± 11	64 ± 9	0.593
Charlson comorbidity index	2 ± 1.7	1.6 ± 1.0	0.152
Baseline HbA1c, % (mmol/mol)	10.6 ± 2.1 (92)	9.8 ± 1.7 (84)	0.076
Baseline weight, kg	101.3 ± 23.5	109.8 ± 17.5	0.055
BMI, kg/m2	32.4 ± 7.5	36.1 ± 6.3	0.012*

Data are mean ± SD unless otherwise indicated. *Statistically significant.

Health care utilization during the 6 months before program enrollment is presented in the Supplementary Material. For this analysis, ED visits, hospitalizations, and surgeries for any reason were included. The number of veterans who used the ED or inpatient services at least once was lower in the VDOP group than in the endocrine group (ED use: 8 [21%] vs. 25 [40%], P < 0.05; inpatient services use: 4 [10%] vs. 20 [32%]; P < 0.05). The number of veterans who underwent one or more surgical procedures was not significantly different between groups (6 [15%] in VDOP vs. 13 [21%] in the endocrine group; P = 0.166). Counts of preprogram outpatient visits per month also did not differ significantly between VDOP and endocrine groups (3.5 vs. 4.7, respectively; P = 0.392). During the intervention period, VDOP participants had significantly more notes entered into the electronic medical record than did endocrine group participants (21 vs. 5; P < 0.05). These notes included VVC setup, telehealth app education, and provider visits.

Table 2 lists changes in HbA_1c and weight at 6- and 12-month follow-up for both groups. Of the 43 veterans who enrolled in VDOP, 39 (90.7%) completed the program and were included in the outcome analysis. Because follow-up data were obtained from routine clinical care rather than a structured study protocol, not all participants had HbA_1c or weight measurements available at the 6- and 12-month time points.

Table 2.

HbA_1c and weight changes

Variable	Endocrine group (n = 62)	VDOP group (n = 39)	P value
Baseline HbA1c, % (mmol/mol)	10.6 ± 2.1 (92)	9.8 ± 1.7 (84)	0.076
6-Month HbA1c, % (mmol/mol)	8.2 ± 1.7 (66) (n = 40)	7.1 ± 0.8 (54) (n = 38)	<0.001*
12-Month HbA1c, % (mmol/mol)	8.2 ± 1.6 (66) (n = 21)	7.1 ± 0.9 (54) (n = 18)	0.014*
Baseline weight, kg	101.3 ± 23.5	109.8 ± 17.5	0.055
6-Month weight, kg	99.2 ± 23.2 (n = 56)	102.1 ± 15.8 (n = 39)	0.497
6-Month weight, % change	−0.70	−6.80	<0.001*
12-Month weight, kg	95.9 ± 22.1 (n = 29)	102.9 ± 15.4 (n = 19)	0.237
12-Month weight, % change	−1.20	−7.20	0.006*

Data are mean ± SD unless otherwise indicated. *Statistically significant.

All VDOP participants had weight measurements available at 6 months, and 38 (97.4%) had HbA_1c measurements. Among endocrine participants, 56 (90.3%) had weight measurements available at 6 months, and 40 (64.5%) had HbA_1c measurements. Both groups demonstrated significant improvements in mean HbA_1c at 6 months; however, the HbA_1c value was significantly lower in the VDOP group than in the endocrine group (7.1% [54 mmol/mol] vs. 8.2% [66 mmol/mol]; P < 0.05). In terms of weight, VDOP participants experienced greater average weight loss (−7.7 kg vs. −2.1 kg; P < 0.001) at 6 months. This corresponded to a mean percent weight loss of −6.8% vs. −0.7% at the 6-month follow-up (P < 0.001).

At 12 months, HbA_1c values were available for 18 of the 39 VDOP participants (46.2%) and 21 of the 62 endocrine participants (33.9%). Weight values were available at 12 months for 19 VDOP participants (48.7%) and 29 endocrine participants (46.8%). The mean HbA_1c value at 12 months was 7% (54 mmol/mol) in the VDOP group and 8.2% (66 mmol/mol) in the endocrine group. Average weight loss was −6.9 kg in the VDOP group and −5.4 kg in the endocrine group at 12 months (P < 0.05). This corresponded to a mean percent weight loss of −7.2% compared with −1.2%, respectively, at 12 months (P < 0.05).

Both groups had a mean follow-up of 14 months (P = 0.706) for the post-program health care utilization analysis (Supplementary Material). The VDOP group had fewer veterans with inpatient stays (3 [7.7%] vs. 21 [34%]; P < 0.05) and ED visits (12 [31%] vs. 34 [55%]; P < 0.05) compared with the endocrine group. The number of veterans who underwent at least one surgical procedure was not significantly different between the VDOP and endocrine cohorts (6 [15%] vs. 16 [26%], respectively; P = 0.095). The number of outpatient visits per veteran per month did not differ between groups in unadjusted or adjusted analyses. After controlling for baseline values, race, sex, and medication initiation in the multivariable logistic regression analysis, VDOP participants were significantly less likely to be hospitalized (odds ratio 0.224; 95% CI 0.059, 0.856; P < 0.05), but no more likely to have an ED visit or a surgical procedure than participants in the endocrine group.

Table 3 presents the results of the general linear models evaluating the effect of group assignment on the outcomes of interest after adjustment for baseline HbA_1c, weight, age, sex, race, and diabetes medications started during the intervention. Compared with the endocrine group, VDOP participants experienced significantly greater mean HbA_1c and weight reductions at the 6- and 12-month follow-up visits. During the intervention period, GLP-1 receptor agonists were initiated more frequently in the VDOP group than in the endocrine group (53.8% vs. 29.0%; P < 0.05) (Supplementary Material). However, after adjustment in the multivariable models, GLP-1 initiation was not independently associated with HbA_1c or weight outcomes. Only initiation of sulfonylureas and pioglitazone were associated with less favorable HbA_1c changes at 6 months.

Table 3.

Adjusted results for weight and HbA_1c changes between groups

Variable	Follow-up (months)	Endocrine group	VDOP adjusted effect (β ± SE)	95% CI	P value
HbA1c change from baseline (%)	6	Ref	−1.3 ± 0.28	−1.81, −0.73	<0.001*
	12	Ref	−1.4 ± 0.41	−2.16, −0.54	0.003*
Weight change from baseline (kg)	6	Ref	−13.19 ± 1.15	−18.16, −8.22	<0.001*
	12	Ref	−14.83 ± 2.02	−23.55, 6.11	0.002*
% Change in weight from baseline	6	Ref	−5.60 ± 1.11	−7.74, 3.38	<0.001*
	12	Ref	−8.60 ± 2.0	−12.47, −4.63	<0.001*
HbA1c change from baseline (%)	6	Sulfonylurea	+1.1 ± 0.05	0.99, 1.18	0.036*
	6	Pioglitazone	+1.6 ± 0.70	0.23, 2.98	0.025*

*Statistically significant. β, effect size; Ref, reference.

All 39 VDOP participants were sent a survey about their satisfaction and quality of life; 26 (67%) completed it. Overall participant satisfaction with VDOP, as well as veterans’ quality-of-life ratings before and after participating in VDOP, are reported in the Supplementary Material. Five veterans reported being satisfied and 21 reported being very satisfied with VDOP. Before starting VDOP, 10 veterans reported their quality of life as quite bad, bad, or as bad as it could possibly be. After VDOP, all 26 respondents reported their quality of life as quite good, good, or as good as it could possibly be.

Conclusions

We found that participation in VDOP, a 100% virtual multidisciplinary intensive intervention, was associated with significantly improved glycemic control and weight loss compared with those receiving standard endocrine consultant care. Participants in VDOP also reported improved quality of life after completion.

Numerous studies have demonstrated that the addition of telehealth interventions to usual diabetes care achieves superior glycemic control than usual care alone. Meta-analyses by Su et al. (18) in 2015 and Lee et al. (19) in 2017, including 55 and 107 randomized trials, respectively, found that a variety of telehealth strategies significantly improved glycemic control when added to usual care.

In alignment with our results, other intensive telehealth programs have shown substantial impact on glycemic control and weight loss. For instance, the Virta health program reported significant improvements in glycemic control and weight among persons with diabetes who followed a very-low-carbohydrate ketogenic diet under close medical supervision (9). Of 262 individuals who began therapy in that study, 218 completed the program. The researchers noted significant reductions in both HbA_1c (−1.3%) and weight (−13.8 kg) (9). This program was also piloted within the VA, and a recent evaluation after 2 years showed the intervention was associated with a significant reduction in weight (BMI −1.56 kg/m²), total monthly diabetes medication use, and per-patient per-month outpatient spending (20). However, no statistically significant differences in HbA_1c, blood pressure, outpatient visits, inpatient visits, or ED visits were observed (20).

Other virtual diabetes interventions designed and implemented within the VA have also reported significant benefits. The Advanced Comprehensive Diabetes Care (ACDC) program delivered multidisciplinary intensive telehealth services involving biweekly telehealth nurse visits, supported by a medication manager, as well as a dietitian (if overweight) and psychiatrist (if there was a positive screen for depression). In a randomized controlled trial, ACDC recipients had a significantly greater decrease in HbA_1c (−1.59% vs. −0.98%) after 12 months than the control group, which received telemonitoring and care coordination but no medication management (21). The ACDC group had a nonsignificantly greater increase in BMI (0.24 vs. 0.05 kg/m²). Another meaningful intervention named Diabetes Care Networks was recently examined with 87 veterans (22). Remote self-management education and medication management by collaboration of primary care pharmacists with diabetes specialists led to a significant decrease in HbA_1c, LDL level, and systolic blood pressure (22).

The VDOP’s positive impact on HbA_1c and weight loss adds to successful programs that have been reported outside and inside the VA, but with some noticeable differences. In contrast with the Virta program (9), VDOP does not use a very-low-carbohydrate ketogenic diet, because evidence supporting the safety and efficacy of this diet is still inconclusive (23). Additionally, within VDOP, the use of GLP-1 receptor agonists and SGLT2 inhibitors (known to increase the risk of euglycemic diabetic ketoacidosis in people following a ketogenic diet) are used by veterans with cardiovascular or renal disease as needed. When compared with other VA programs, such as ACDC or Diabetes Care Networks, differences in target participant population, model structure, and resource utilization are also noted, suggesting the possibility for synergic coexistence while not being mutually exclusive. Lastly, VDOP’s emphasis on the physical activity component may explain the significant weight loss seen in this study compared with conventional endocrine care or diabetes-focused programs despite the use of medication such as GLP-1 receptor agonists.

This analysis has several limitations that should be noted. First, VDOP was implemented as a quality improvement program rather than a randomized controlled trial. Outcomes were compared with a contemporaneous group of veterans receiving standard endocrinology care at the same VA medical center. Although the comparison group was selected from the same health care system and met similar eligibility criteria, unmeasured differences between groups may have influenced outcomes. Participation in VDOP was voluntary, raising the possibility of selection bias, because veterans who elected to enroll may have been more motivated to engage in diabetes self-management than those who chose conventional endocrinology care. Although baseline demographic and clinical characteristics were generally similar between groups, VDOP participants had higher baseline BMI and greater baseline body weight than the endocrine group, which may have increased the potential for greater absolute or relative weight loss. Although multivariable models adjusted for baseline values of the outcome of interest, residual confounding related to motivation, health behaviors, readiness for change, or other unmeasured factors remains possible. Second, our post-program survey was not a validated instrument for assessing participant satisfaction or quality of life. Third, because veterans entered the program at different times, not all participants had reached 12-month follow-up at the time of analysis, resulting in missing data.

Nonetheless, the significant findings of this program support VDOP as an effective tool for the management of uncontrolled diabetes and obesity in veterans. We believe incorporating a multidisciplinary team, frequent visits, team huddles, an emphasis on diet and physical activity in addition to medication, and self-monitoring of both activity and glucose offers a successful formula for achieving marked improvements in weight and glycemic control. We note that despite the requirement for intensive self-management, veterans reported improved quality of life and very high overall satisfaction.

Although a formal cost-effectiveness analysis was not performed, several observations suggest the potential for economic benefit. VDOP participants were less likely to experience hospitalization during the post-program period, a major driver of diabetes-related health care costs. The program also leveraged existing VA telehealth infrastructure and multidisciplinary clinical staff, minimizing the need for additional capital investment. However, VDOP required substantially greater clinical engagement during the active intervention period than did providing standard care (21 vs. 5 clinical notes per participant), reflecting the intensive nature of the program. Therefore, although the intervention may offer opportunities for downstream cost avoidance, it also requires greater upfront personnel resources than standard endocrine care. Future studies should formally evaluate implementation costs, staffing requirements, scalability, and cost-effectiveness.

In the future, we can plan for research that evaluates VDOP at other sites, including an assessment of economic implications as well as clinical outcomes. Although not all veterans are likely to have the motivation and comfort with technology to successfully participate in VDOP-like programs, integration of VDOP with less intensive but still successful programs would allow providers to tailor diabetes care interventions to the diverse population of U.S. veterans.

Conclusion

In summary, this analysis shows that the intensive, multidisciplinary VDOP was associated with significant improvements in diabetes control, weight loss, and life satisfaction in veterans with uncontrolled diabetes and obesity compared with standard endocrine care. Prospective trials evaluating implementation issues, economic implications, and clinical outcomes at a broader range of sites should be designed and used.

This article contains supplementary material online at https://doi.org/10.2337/figshare.33055055.

Funding Statement

This work was supported by the VA Office of Connected Care Virtual Care Consortium of Research (CORE) grant.