This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Diabetes, is properly cited. The complete bibliographic information, a link to the original publication on http://diabetes.jmir.org/, as well as this copyright and license information must be included.
Translation of diabetes self-management education and support (DSMES) into a digital format can improve access, but few digital programs have demonstrated outcomes using rigorous evaluation metrics.
The aim of this study was to evaluate the impact of a digital DSMES program on hemoglobin A1c (HbA1c) for people with type 2 diabetes.
A single-arm, nonrandomized trial was performed to evaluate a digital DSMES program that includes remote monitoring and lifestyle change, in addition to comprehensive diabetes education staffed by a diabetes specialist. A sample of 195 participants were recruited using an online research platform (Achievement Studies, Evidation Health Inc). The primary outcome was change in laboratory-tested HbA1c from baseline to 4 months, and secondary outcomes included change in lipids, diabetes distress, and medication adherence.
At baseline, participants had a mean HbA1c of 8.9% (SD 1.9) and mean BMI of 37.5 kg/m2 (SD 8.3). The average age was 45.1 years (SD 8.9), 70% were women, and 67% were White. At 4-month follow up, the HbA1c decreased by 0.8% (
This study provides early evidence that a digitally enhanced DSMES program improves HbA1c and disease self-management outcomes.
Over 34 million people in the United States have diabetes (9% of the adult population), and 1 in 4 health care dollars spent in the United States is for diabetes care [
DSMES is widely covered by private and public insurance, including Medicare, and is typically prescribed by a physician at diagnosis, when education gaps exist, or when the treatment plan is changed. The primary goal of DSMES is to help patients acquire the knowledge, skills, and abilities for diabetes self-care [
Despite the widely accepted benefits of DSMES, access remains a challenge. Only 43 states and 57% of counties in those states have accredited DSMES programs in the United States [
Technology-based DSMES programs have demonstrated a positive impact on hemoglobin A1c (HbA1c) in academic settings with noncommercially available programs [
The goal of this pilot study was to evaluate the impact of a digital DSMES program enhanced with deep lifestyle and behavior change support on HbA1c for people with T2DM and elevated HbA1c. We hypothesized that the digital DSMES program would be associated with greater improvements in HbA1c for people who were furthest away from their HbA1c goal (baseline HbA1c≥9.0%) at the start of the program. We further evaluated the impact of the digital DSMES program on cardiovascular and patient-reported outcomes, as cardiovascular risk factors are a frequent comorbidity of diabetes.
We invited members of an online health community to participate in this study (Achievement, Evidation Health Inc). Achievement is a web- and mobile-based community in the United States where members can connect their activity trackers, and fitness and health apps to the platform and, by logging activities, accumulate points that are redeemable for monetary rewards. Additionally, members self-report on various health conditions and are invited to participate in remote research opportunities as relevant studies become available. In this study, recruitment was targeted to members who had self-reported a diagnosis of T2DM. Invited members were linked to an online research study platform (Achievement Studies, Evidation Health Inc) where study eligibility was assessed using automated screener questions. Individuals who lived in the United States, were at least 18 years of age, self-reported a T2DM diagnosis, self-reported HbA1c of 7.5% or greater, had a BMI≥25 kg/m2 (≥23 kg/m2 if they self-identified as Asian), and had access to a computer or smartphone to participate in the digital DSMES program were eligible for the study.
If deemed eligible after completing the screener, potential participants continued in the online study platform to sign an electronic informed consent form and completed an online baseline survey, which consisted of questions about their demographics, health and diabetes history, and patient-reported outcomes. They then completed a baseline visit at a Quest Diagnostics Patient Service Center (PSC) of their choosing. The baseline visit consisted of a venous whole blood draw, physical measurements (height, weight, waist circumference), resting blood pressure, and resting heart rate. After completing the PSC visit, potential participants were instructed to set up their account on the digital DSMES program. After completion of a signed electronic informed consent form, and both the PSC visit and program account setup, individuals were considered enrolled in the study. Participants were able to reach out to research staff with questions via email or phone through the online study platform before and during the enrollment process, and could continue to reach out throughout the study.
During the study period, participants were encouraged to engage with the DSMES program. All participants were provided a cellularly connected weight scale that was linked to their program account. Participants who were advised to use monitoring devices in their diabetes self-care were provided cellularly connected blood pressure monitors and glucose meters. Participants were also able to access their own personal online study platform dashboard to complete study procedures and keep track of their progress throughout the study through the use of any web-enabled device. Approximately 4 months after enrollment, participants repeated the online survey and clinical outcome measures (HbA1c, blood pressure). Participants received compensation for completing each study-related task such as surveys and lab visits. This study was approved by the Western Institutional Review Board (Puyallup, WA).
The primary outcome of this study was change in HbA1c from baseline to 4 months, as well as changes in HbA1c based on starting HbA1c values. Secondary outcomes included changes in cardiovascular risk factors (blood pressure, total cholesterol [TC]) among those who started the study with elevated risk factors, in addition to changes in diabetes distress and medication adherence from baseline to 4 months.
At baseline, participants completed an assessment at the PSC that included 13 mL venous whole blood specimen collection under sterile conditions by a trained phlebotomist. The nonfasting blood specimens were processed for HbA1c and a lipids panel (TC, high- and low-density lipoprotein [HDL, LDL], and TC/HDL ratio). A trained technician collected blood pressure after a 5-minute quiet resting period with legs uncrossed using an automatic blood pressure monitor and size-adjustable cuff. Height was measured to the nearest centimeter using a calibrated stadiometer with the participant in stocking feet. Weight was measured using a calibrated scale with the participant in light clothing and no shoes. Waist circumference was measured in whole units (inches) using a nonstretchable measuring tape above the first layer of clothing. BMI was calculated from weight in kilograms divided by height in meters squared. Results were sent by Quest Diagnostics and accessed by the research team via secure file transfer. Participants received copies of their results both via secure email and mail.
Participants completed an online survey of patient-reported outcomes including the Diabetes Distress Scale (DDS), a 17-item scale of different dimensions of distress and burden related to diabetes, which has been shown to have reliability and validity [
The original protocol planned for a repeat assessment using identical methods 4 months after enrollment. However, the 4-month assessments were scheduled to begin in April of 2020, during the height of the COVID-19 pandemic [
Omada for Diabetes is a digitally enhanced DSMES program designed to build self-management skills and support diabetes management between outpatient visits with primary care providers and specialists to ensure that users achieve their health targets (eg, HbA1c, blood pressure, cholesterol) and obtain health maintenance services (eg, screening for neuropathy and retinopathy). The program offers disease education, comprehensive lifestyle self-management support (ie, support for weight loss, dietary changes, physical activity increases), support for involvement in members’ current medication regimen, and support for use of monitors or trackers for their blood sugar and blood pressure, which are often used to inform small modifications in food intake, physical activity, medication, or communication with health care providers. Participants used a technology-enabled platform with a portable interface to a variety of personal mobile devices. All participants received a cellularly connected BodyTrace weight scale, and if needed, a blood glucose monitor (3G BioTel Care, Telcare LLC, Concord, MA) was also provided. Participants were assigned to a CDCES who provided individualized coaching around the American Association of Diabetes Educators 7 self-care behaviors [
The study was powered to detect a clinically meaningful 0.5% reduction in the primary outcome of HbA1c. With an estimated standard deviation of 1.8 and power set to 90%, the minimal sample size needed was 162. To allow for potential 20% loss to follow up and 10% of lab HbA1c values being below 7.5% at baseline, a total of 186 participants were planned for enrollment.
Descriptive statistics are presented to describe the demographics and baseline health status of participants. Baseline correlations using Pearson and Spearman correlation coefficients were examined to determine variables (age, gender, BMI) that could potentially confound HbA1c outcomes. No significant correlations were detected; therefore, paired
We analyzed outcomes using complete case analysis for those who returned 4-month clinical and patient-reported survey data. Using multiple imputation, with an imputation of baseline values for primary and secondary outcomes for those with missing data at 4 months, we found that outcomes were similar in magnitude and statistical significance using both analytic methods. Therefore, we present our findings on the sample using results from the complete case analysis.
Although the recruitment goal was 162 participants with starting HbA1c above 7.5%, 32 of the first 100 participants’ laboratory HbA1c result was below the 7.5% threshold. Therefore, we changed the protocol to use the baseline HbA1c as a clinical criterion for the study and only accepted those with a lab HbA1c value of 7.5% or greater. We continued enrollment until we reached at least 162 participants with a baseline HbA1c of 7.5% or greater and allowed the 32 participants with a baseline HbA1c below 7.5% to remain in the study. The final enrolled sample was 195, including 163 with a baseline HbA1c of 7.5% or greater and 32 with a baseline HbA1c of less than 7.5%. Six participants were withdrawn from the study: 4 developed a medical condition that precluded participation and 2 requested to voluntarily withdraw. At post-test, 78.8% (n=149) of the remaining 189 participants completed the home test kit; 8 were not sent kits as they resided in states where the home test is not authorized for distribution, and 88.4% (n=167) completed the online questionnaire. Study completion was defined as a final HbA1c value or completion of the final online questionnaire. We compared baseline demographic and clinical values for participants who completed the 4-month data collection and those who were lost to follow up, and found no significant differences across any baseline characteristics. We define loss to follow up as incompletion of the primary outcome of HbA1c. See
Study participant flowchart. HbA1c: hemoglobin A1c.
Baseline characteristics of participants are shown in
Baseline participant characteristics (N=195).
Baseline characteristica | Value | |
Age (years), mean (SD) | 45.1 (8.9) | |
Female, n (%) | 136 (69.7) | |
|
||
White/Caucasian | 131 (67.2) | |
Black/African American | 32 (16.4) | |
Hispanic or Latino | 17 (8.7) | |
Asian | 6 (3.1) | |
American Indian or Alaska Native | 2 (1.0) | |
Native Hawaiian or other Pacific Islander | 1 (0.5) | |
Other | 6 (3.1) | |
BMI, mean (SD) | 37.5 (8.3) | |
Weight (pounds), mean (SD) | 235.6 (57.3) | |
Weight (kg), mean (SD) | 106.9 (26.0) | |
Hemoglobin A1c, mean (SD) | 8.9 (1.9) | |
Total cholesterol (mg/dL), mean (SD) | 178.9 (43.3) | |
Systolic blood pressure (mmHg), mean (SD) | 127.0 (16.1) | |
Diastolic blood pressure (mmHg), mean (SD) | 82.0 (10.4) | |
Diabetes Distress Score, mean (SD) | 2.7 (1.0) | |
Adherent to current medications, n (%) | 36 (18.5) |
aThere were no statistically significant differences across baseline characteristics among those with and without follow-up data.
Averaged across the 16 program weeks, participants used their blood glucose meter an average of 7.4 times per week. Participants weighed in an average of 4.9 times per week, interacted with their CDCES an average of 1.6 times per week, completed an average of 0.8 lessons per week, interacted with their peer groups an average of 0.9 times per week, tracked their physical activity 5.3 times per week, and tracked meals an average of 10.2 times per week.
Baseline to post-test changes in all study outcomes are shown in
Baseline to post-test changes in clinical outcomes (N=167).
Outcomes | n | Baseline | Post-test | Difference | 95% CI | |||
|
||||||||
|
HbA1cb (%) | 149 | 8.9 | 8.1 | –0.8 | –1.1 to –0.5 | <.001 | |
|
Weight (pounds) | 147 | 231.4 | 228.3 | –3.0 | –5.8 to –0.3 | .03 | |
|
Weight (kg) | 147 | 105.0 | 103.6 | –1.4 | –2.6 to –0.1 | .03 | |
|
5% weight loss (%) | 147 | 0.0 | 18.4 | 18.4 | 0.1 to 0.2 | <.001 | |
|
|
167 | 2.6 | 2.3 | –0.3 | –0.5 to –0.2 | <.001 | |
|
|
Emotional Burden | 167 | 2.7 | 2.4 | –0.3 | –0.5 to –0.1 | <.001 |
|
|
Physician-Related | 167 | 2.1 | 1.8 | –0.3 | –0.4 to –0.1 | .001 |
|
|
Regimen-Related | 167 | 3.0 | 2.6 | –0.4 | –0.6 to –0.3 | <.001 |
|
|
Interpersonal | 167 | 2.7 | 2.4 | –0.3 | –0.5 to –0.1 | .002 |
|
Medication adherence (%) | 158 | 20.3 | 31.0 | 10.7 | —c | .01 | |
|
||||||||
|
TCe (mg/dL) | 43 | 230.0 | 190.5 | –39.5 | –51.3 to –27.6 | <.001 | |
|
SBPf (mmHg) | 114 | 131.6 | 132.5 | 0.9 | –2.1 to 3.9 | .54 | |
|
DBPg (mmHg) | 114 | 84.7 | 82.0 | –2.7 | –4.3 to –1.0 | .002 |
aStudy participants with complete data from both baseline and 4-month time points.
bHbA1c: hemoglobin A1c.
c—:Not applicable.
dStudy participants who began the study with elevated cardiovascular risk factors.
eTC: total cholesterol.
fSBP: systolic blood pressure.
gDBP: diastolic blood pressure.
Baseline to post-test changes in hemoglobin A1c (HbA1c) based on starting HbA1c.
HbA1c category | n | Baseline | Post-test | Difference | 95% CI | |
<7.5% | 24 | 6.3 | 6.4 | 0.1 | –0.2 to 0.4 | .49 |
7.5%-7.9% | 24 | 7.7 | 7.4 | –0.3 | –0.6 to 0.1 | .18 |
8.0%-8.9% | 28 | 8.4 | 7.8 | –0.6 | –1.0 to –0.2 | .002 |
>9.0% | 73 | 10.4 | 9.0 | –1.4 | –1.8 to –0.9 | <.001 |
At baseline, 58.5% (114/195) of the participants had systolic or diastolic blood pressure above the normal range (<120 mmHg and <80 mmHg, respectively). There was no significant change in systolic blood pressure, whereas diastolic blood pressure decreased by an average of 2.7 mmHg (
In the total sample, diabetes distress significantly decreased from 2.6 at baseline to 2.3 at post-test (
The results of this study provide initial evidence that the enhanced digital DSMES program was effective for improving HbAlc, weight, diabetes distress, and medication adherence among a sample of people with T2DM and elevated HbA1c. Furthermore, those who were furthest from their HbA1c goal at the start of the program (baseline HbA1c≥9.0%) achieved the greatest improvement in HbA1c, with an average change of 1.4%.
We found an inconsistent impact on cardiovascular outcomes among participants who started the study with elevated risk factors, with some improvements in diastolic blood pressure and TC, but no improvements in systolic blood pressure. However, blood pressure at baseline was close to the nationally recommended goal for those with diabetes, and the program was not designed to address hypertension specifically. Engagement was strong as evidenced by the high frequency of use across the features of the digital platform.
These results are consistent with prior studies of digital DSMES programs (both academic and commercial) that showed improvements in HbA1c and psychosocial outcomes [
The clinical outcomes observed in this study meet or exceed those expected from traditional DSMES programs as set by the American Diabetes Association [
The improvements in medication adherence are encouraging given that this is a major challenge in diabetes management [
There were several limitations to this pilot study. First, this pilot study is limited by its single-arm design and therefore carries the typical challenges in a nonrandomized design of unknown causal inference. Future research will benefit from a control group comparison and a randomized design to allow for a maximally rigorous test of the intervention. Second, we had to change the study methodology for follow-up lab measurement due to COVID-19 by shifting to a self-collected blood specimen versus a phlebotomist-collected venipuncture specimen; this creates potential for measurement error between instruments. However, this risk is attenuated by the high correlation of the venipuncture HbA1c and dried blood spot method [
This study provides additional evidence that a digitally delivered DSMES program enhanced with deep lifestyle and behavior change support impacts HbA1c for people with T2DM and elevated HbA1c, showing the greatest benefit for those with higher blood glucose levels, and suggests benefits for weight loss and improvements in cardiovascular outcomes. Future research is needed to understand the potential impact of digital DSMES on long-term diabetes outcomes to meet the needs of the changing health care landscape.
certified diabetes care and education specialist
Diabetes Distress Scale
diabetes self-management education and support
hemoglobin A1c
high-density lipoprotein
low-density lipoprotein
Patient Service Center
Simplified Medication Adherence Questionnaire
type 2 diabetes mellitus
total cholesterol
The authors would like to thank Andrea Newcom, Bailey Peterka, Carolyn Salter, Danene Moberly, Melinda Merry, and Brieana Polk-Perez for their support of the project and work with participants. We would also like to thank Sara Cross and Anna Telthorst from Quest Diagnostics, and Kimberly Russell, Lisa Johnstone, Amber Hogue, and Maximo Prescott from Evidation Health for study management. Data included in this manuscript were presented in an abstract at the 20th Annual Diabetes Technology Meeting Virtual Poster Session on November 19, 2020. This study was funded by Omada Health, Inc.
FWA, RQ, CCS, MT, and CBJ are employees of Omada Health, Inc, and receive salary and stock options. CC and JJ are employees of Evidation Health, Inc, and receive salary. Evidation Health, Inc received funds from Omada Health, Inc to perform the study.