Develop a PowerPoint Presentation in regards to the issue of Advanced Practice: Telehealth in Managing Chronic Diseases: Diabetes and Hypertension.

Effect of a Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Patients With Persistently Poor Type 2 Diabetes Control A Randomized Clinical Trial Matthew J. Crowley, MD, MHS; Phillip E. Tarkington, MD; Hayden B. Bosworth, PhD; Amy S. Jeffreys, MStat; Cynthia J. Coffman, PhD; Matthew L. Maciejewski, PhD; Karen Steinhauser, PhD; Valerie A. Smith, DrPH; Moahad S. Dar, MD; Sonja K. Fredrickson, MD; Amy C. Mundy, NP; Elizabeth M. Strawbridge, RD; Teresa J. Marcano, MSN, RN; Donna L. Overby, RN; Nadya T. Majette Elliott, MPH; Susanne Danus, BS; David Edelman, MD

IMPORTANCE Persistently poorly controlled type 2 diabetes (PPDM) is common and causes poor outcomes. Comprehensive telehealth interventions could help address PPDM, but effectiveness is uncertain, and barriers impede use in clinical practice.

OBJECTIVE To address evidence gaps preventing use of comprehensive telehealth for PPDM by comparing a practical, comprehensive telehealth intervention to a simpler telehealth approach.

DESIGN, SETTING, AND PARTICIPANTS This active-comparator, parallel-arm, randomized clinical trial was conducted in 2 Veterans Affairs health care systems. From December 2018 to January 2020, 1128 outpatients with PPDM were assessed for eligibility and 200 were randomized; PPDM was defined as maintenance of hemoglobin A1c (HbA1c) level of 8.5% or higher for 1 year or longer despite engagement with clinic-based primary care and/or diabetes specialty care. Data analyses were preformed between March 2021 and May 2022.

INTERVENTIONS Each 12-month intervention was nurse-delivered and used only clinical staffing/resources. The comprehensive telehealth group (n = 101) received telemonitoring, self-management support, diet/activity support, medication management, and depression support. Patients assigned to the simpler intervention (n = 99) received telemonitoring and care coordination.

MAIN OUTCOMES AND MEASURES Primary (HbA1c) and secondary outcomes (diabetes distress, diabetes self-care, self-efficacy, body mass index, depression symptoms) were analyzed over 12 months using intent-to-treat linear mixed longitudinal models. Sensitivity analyses with multiple imputation and inclusion of clinical data examined the impact of missing HbA1c measurements. Adverse events and intervention costs were examined.

RESULTS The population (n = 200) had a mean (SD) age of 57.8 (8.2) years; 45 (22.5%) were women, 144 (72.0%) were of Black race, and 11 (5.5%) were of Hispanic/Latinx ethnicity. From baseline to 12 months, HbA1c change was −1.59% (10.17% to 8.58%) in the comprehensive telehealth group and −0.98% (10.17% to 9.19%) in the telemonitoring/care coordination group, for an estimated mean difference of −0.61% (95% CI, −1.12% to −0.11%; P = .02). Sensitivity analyses showed similar results. At 12 months, patients receiving comprehensive telehealth had significantly greater improvements in diabetes distress, diabetes self-care, and self-efficacy; no differences in body mass index or depression were seen. Adverse events were similar between groups. Comprehensive telehealth cost an additional $1519 per patient per year to deliver.

CONCLUSIONS AND RELEVANCE This randomized clinical trial found that compared with telemonitoring/care coordination, comprehensive telehealth improved multiple outcomes in patients with PPDM at a reasonable additional cost. This study supports consideration of comprehensive telehealth implementation for PPDM in systems with appropriate infrastructure and may enhance the value of telehealth during the COVID-19 pandemic and beyond.

TRIAL REGISTRATION ClinicalTrials.gov Identifier: NCT03520413

JAMA Intern Med. doi:10.1001/jamainternmed.2022.2947 Published online July 25, 2022.

Visual Abstract

Supplemental content

Author Affiliations: Author affiliations are listed at the end of this article.

Corresponding Author: Matthew J. Crowley, MD, MHS, Division of Endocrinology, Department of Medicine, Duke University School of Medicine; Durham VA Center of Innovation to Accelerate Discovery and Practice Transformation, Durham VAMC HSR&D (152), 508 Fulton St, Durham, NC 27705 (matthew. [email protected]).

Research

JAMA Internal Medicine | Original Investigation

(Reprinted) E1

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

P atients with persistently poor type 2 diabetes (T2D) con- trol disproportionately experience negative outcomes.1-3

We have defined persistently poorly controlled diabe- tes (PPDM) as maintenance of hemoglobin A1c (HbA1c) level of 8.5% or greater for more than 1 year despite receiving clinic- based diabetes care; 10% to 15% of all patients with T2D meet PPDM criteria.4,5 Given their high risk for complications and costs,1-3 patients with PPDM represent a compelling popula- tion for care delivery redesign.

Drivers of PPDM, including unavailable blood glucose data, medication nonadherence, suboptimal diet/activity, com- plex medications, and depression,6-12 can be difficult to ad- dress in the clinic setting.13,14 By facilitating contact outside of clinic, telehealth could improve outcomes in PPDM. Tele- health strategies targeting individual factors underlying poor T2D control reduce HbA1c level vs clinic-based care by 0.3% to 0.6%.15-19 Although such HbA1c changes may not suffice for patients with PPDM, combining multiple strategies into com- prehensive telehealth interventions could produce greater im- provement. However, multicomponent T2D interventions have achieved variable results.20-25

Beyond this uncertain effectiveness, other barriers have hindered implementation of comprehensive telehealth for PPDM in practice. Implementation barriers include interven- tion designs reliant on research-funded staff and resources, in- sufficient electronic health record (EHR) integration of pa- tient data, and uncertain reimbursement.26-29 Before comprehensive telehealth can become a real-world solution for PPDM, approaches are needed that are unambiguously ef- fective and also explicitly designed for feasible implementa- tion. The upsurge in telehealth use during the COVID-19 pan- demic has only strengthened the case for considering comprehensive telehealth as a means to address PPDM.30

We sought to address barriers to practical use of compre- hensive telehealth for PPDM by evaluating a comprehensive telehealth intervention in a randomized clinical trial (RCT). This intervention combined 5 strategies targeting contributors to PPDM: telemonitoring, self-management support, diet/ activity support, medication management, and depression sup- port. To facilitate eventual implementation, we explicitly de- signed the intervention for delivery by clinical Veterans Health Administration (VHA) Home Telehealth (HT) nurses using ex- isting clinical resources.

Methods The protocol for this active-comparator, parallel-arm RCT (NCT03520413) has been published5 and appears in Supplement 1. We compared a practical, comprehensive telehealth intervention with a simpler telehealth approach, consisting of telemonitoring and care coordination, which is already available in VHA practice for T2D. A usual-care comparator was deemed inappropriate because clinic-based care is by definition insufficient for PPDM.31

This RCT was conducted at 2 VHA sites (Durham, North Carolina, and Richmond, Virginia). Institutional review boards at both sites approved the study. This study followed the Con-

solidated Standards of Reporting Trials (CONSORT) reporting guideline.

Population We recruited patients with PPDM, defined as the following: di- agnosed T2D (International Statistical Classification of Dis- eases and Related Health Problems, Tenth Revision E11); 2 or more HbA1c values of 8.5% or greater during the prior year with none less than 8.5%; and at least 1 appointment during the prior year with a primary care clinician or diabetes specialist (en- docrinologist or other diabetes clinician). Exclusion criteria in- cluded the following: refusal to enroll in VHA HT (because both study interventions were delivered by HT nurses); factors mak- ing HbA1c reduction potentially inadvisable (age >70 years, metastatic cancer/life expectancy <5 years, recent cardiovas- cular disease complications, or prior hypoglycemic seizure/ coma); lack of telephone access; dementia, psychosis, or sub- stance use disorder; pregnancy; receiving dialysis or skilled nursing care; insulin pump use; or continuous glucose moni- tor use (unless also willing to submit self-monitored blood glu- cose [SMBG] data per HT protocol).

Race was determined by self-report and categorized as Asian, American Indian or Alaska Native, Black or African American, Native Hawaiian or Other Pacific Islander, White, other, or unknown. Because of low numbers in the other cat- egories, race was ultimately presented as Black or African American, White, or other race. Self-reported ethnicity was as- sessed using a single question: “Are you of Latino/a or His- panic origin or descent?” Race and ethnicity data were col- lected to facilitate generalizability assessment and to determine whether intervention effectiveness varied by these factors.

Recruitment and Enrollment After EHR screening, a research assistant mailed opt-out let- ters to potential participants, then conducted phone screen- ing. Eligible participants provided informed consent and un- derwent in-person baseline assessment; consented patients with a baseline HbA1c level of less than 8.5% were excluded

Key Points Question Compared with a simpler telehealth approach (telemonitoring and care coordination), can a practical, comprehensive telehealth intervention improve outcomes among patients whose type 2 diabetes remains persistently poorly controlled despite clinic-based care?

Findings In this randomized clinical trial of 200 adults with persistently poorly controlled type 2 diabetes, hemoglobin A1c

level improved by 1.59% at 12 months among those randomized to receive the comprehensive telehealth intervention, compared with 0.98% for the telemonitoring/care coordination group.

Meaning A comprehensive telehealth intervention improved outcomes in persistently poorly controlled type 2 diabetes compared with a simpler telehealth intervention; because it was explicitly designed for feasible use in clinical practice, this approach may warrant implementation in systems that need to improve diabetes control in which the requisite infrastructure is available.

Research Original Investigation Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes

E2 JAMA Internal Medicine Published online July 25, 2022 (Reprinted) jamainternalmedicine.com

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

before randomization. Given the high proportion of men in the VHA population, we oversampled women, aiming to achieve greater than 20% in the randomized population.

Randomization and Blinding Randomization was stratified with blocks of 2; stratification variables were site, prior VHA HT use, and preenrollment dia- betes specialty care (endocrinologist or other diabetes clini- cian). The computer-generated randomization sequence was accessible only to study statisticians. Patients received ran- domization assignments by phone from the project coordina- tor within 1 week of consent. Because participants received in- formation about both interventions during consent, they were not blinded to randomization arm. Research assistants col- lecting outcome data were blinded to participant randomiza- tion status.

Interventions Both 12-month interventions were delivered by clinical HT nurses rather than research staff. Although experienced with telehealth-based disease care, these nurses had no special- ized diabetes training. Individual nurses delivered only 1 study intervention, with no crossover. Each intervention was deliv- ered by 1 nurse in Durham, while in Richmond, 2 nurses de- livered comprehensive telehealth and 5 delivered telemoni- toring/care coordination.

Following randomization, all participants enrolled in VHA HT. Patients enrolled in HT used a telehealth device (Medtronic), blood glucose meter (Abbott), and connector cable; once connected to the blood glucose meter, the tele- health device automatically transmitted SMBG data to HT. Af- ter HT enrollment, participants began their assigned interven- tion; both groups also continued care with existing clinicians. Patients’ HbA1c goals were individualized per American Dia- betes Association guidelines.32

Comprehensive Telehealth Intervention This intervention comprised 5 nurse-delivered components (Figure 1)5: telemonitoring, self-management support, diet/ activity support (together with a study dietitian), medication management (together with a study medication manager), and depression support (together with a study psychiatrist). The

study dietitian, medication managers, and psychiatrists were clinicians, not research staff. Intervention nurses completed a single training session and received a manual. Nurses deliv- ered the intervention to participants during 26 every-2-week telephone encounters, which the nurses scheduled directly with participants; nurses could additionally be reached for acute issues. Clinical information was tracked using tem- plated EHR notes.

For the telemonitoring component, participants transmit- ted SMBG data up to 4 times daily based on their medication regimens but could monitor less frequently per nurse discre- tion. During each of the 26 scheduled encounters, the nurse reviewed SMBG data, reconciled medications, and assessed self-reported medication adherence.

For the self-management support component, interven- tion nurses delivered module-based self-management educa- tion during 16 of the 26 scheduled encounters. Each module covered a unique topic addressing knowledge and/or self-efficacy.5

For the diet/activity support component, a dietitian called participants with a body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) of 25 or greater to develop an individualized diet plan postenroll- ment. Plans were tailored to patient preferences and targeted to greater than 5% weight loss via a deficit of 500 to 750 calo- ries per day.33 Patients were also encouraged to maintain 150 minutes or more of moderate to vigorous activity weekly.34

During each of the 26 scheduled encounters, the nurse re- viewed progress. An additional dietitian phone follow-up could be arranged for patients not meeting goals.

For the medication management component, each site used 2 to 3 diabetes specialists (physicians, clinical pharma- cists, or nurse practitioners). After each of the 26 scheduled encounters, the intervention nurse forwarded an EHR-based summary note to the medication manager. The medication manager considered treatment changes with guidance from a medication protocol, which targeted a fasting glucose level of 90 to 150 mg/dL and preprandial glucose level of 140 to 180 mg/dL (tailoring permitted based on HbA1c goal/hypoglyce- mia; to convert glucose level to mmol/L, multiply by 0.0555). The medication manager conveyed recommendations via an EHR note addendum, which the nurse implemented; medi-

Figure 1. Comprehensive Telehealth Intervention Design

Scheduled phone encounter (15-30 min per encounter, every 2 wk × 26 wk)

Templated report compiled and documented in EHR

1. Telemonitoring Nurse reviews SMBG data, medications, adherence

2. Self-management support Nurse delivers self-management module

3. Diet/activity support Nurse supports individualized diet plan

and activity plan

4. Medication management Report sent to study medication manager through EHR after encounters, changes

implemented by nurse

5. Depression support Nurse screens for depression every 12 wk,

facilitates study psychiatrist assessment for positive screens

Adapted with permission from Kobe et al.5 EHR indicates electronic health record; SMBG, self-monitored blood glucose.

Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes Original Investigation Research

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 25, 2022 E3

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

cation managers did not routinely contact participants. Pri- mary clinicians were alerted to changes via the EHR.

For the depression support component, participants with a Personal Health Questionnaire-8 (PHQ-8) score of 10 or higher at baseline or subsequent screening entered the depression pro- tocol, which was supported by 1 psychiatrist at each site and provided pharmacologic and nonpharmacologic options.35 All patients receiving depression support had PHQ-8 follow-up ev- ery 8 weeks, with treatment changes as needed.

Telemonitoring/Care Coordination Intervention Participants transmitted SMBG data and received automated self-management information daily by phone. Participants re- ceived nurse calls for alert SMBG values and could reach nurses as needed for acute issues but did not complete scheduled en- counter calls. Participants also received care coordination, in- cluding communication about upcoming appointments, notification of primary clinicians regarding acute needs, and preappointment compilation of SMBG data for review by primary clinicians. Diabetes medication management was not formally integrated into this intervention; instead, medica- tion adjustments were at the discretion of existing clinicians during or outside of scheduled encounters. Of note, because telemonitoring/care coordination is considered routine HT practice, no training was required for nurses delivering this intervention.

Outcomes Outcome assessments were performed by blinded research as- sistants at 3-month intervals for 12 months. Full assessments were competed at 0, 6, and 12 months, with additional HbA1c- only assessments at 3 and 9 months.

The primary outcome was HbA1c level. Secondary out- comes were diabetes distress,36 diabetes self-care,37

self-efficacy,38 BMI, and depression symptoms.39 Adverse events were assessed by structured self-report40; incidence of blood glucose level less than 70 mg/dL was also examined using SMBG data transmitted to HT in both arms.

Intervention Costs Intervention costs were examined in both arms. Labor costs included all intervention nurse, dietitian, medication man- ager, and psychiatrist time spent delivering the intervention; capital costs included HT equipment, telephone service costs, overhead, and supplies.

Fidelity Assessment Fidelity assessment for the comprehensive telehealth inter- vention included nurse tracking of encounters and time using online software. The principal investigator and project coor- dinator conducted periodic shadowing of nurses and bian- nual case review meetings with the medication managers.5 Up- dated medications were tracked at outcome visits.

Influence of COVID-19 Pandemic on Outcome Ascertainment In March 2020, VHA announced a pandemic-related restric- tion on in-person research interactions. Survey data collec-

tion continued by phone when possible, but study measure- ment of HbA1c level and BMI was interrupted. In May 2020, we received permission to resume collection of HbA1c level given its importance to diabetes management; however, BMI could not be measured after March 2020, which affected the 12-month time point for 169 participants. Despite the pandemic, intervention delivery continued uninterrupted at both sites.

Statistical Analyses Sample Size Per previous data,31 we used an α level of .05, 80% power, 20% dropout, within-patient correlation 0.55, SD 1.6, and baseline HbA1c level of 10.3% to estimate that 100 participants per arm would detect a clinically significant HbA1c difference of 0.6% at 12 months.41 Power estimates were derived by generating 1000 stimulated data sets with these assumptions and fitting linear mixed models to assess the effect difference at 12 months.

Analytic Approach All analyses were intention-to-treat and performed using SAS, version 9.4 (SAS Institute).42 Linear mixed longitudinal mod- els were used for all primary and secondary outcomes.43 The primary outcome model included fixed effects for linear, qua- dratic and cubic time, time-by-arm interaction terms, and ran- domization stratification variables, and random effects for in- tercept and linear time (see eMethods in Supplement 2). The covariance structure was determined using Akaike informa- tion criteria.44 Our primary inference was on the estimated be- tween-arm 12-month HbA1c difference. As a post hoc sensitiv- ity analysis, we included baseline covariates with between- arm differences in the primary model. Also, to explore a dose- response effect of the comprehensive telehealth intervention, we conducted a descriptive post hoc analysis examining HbA1c

change among participants completing more than 20 vs 20 or fewer encounters.

For secondary outcomes, fixed effects included dummy- coded time effects for each time point and time-by-arm inter- action terms. Given the pandemic’s hindrance of BMI ascer- tainment, BMI was analyzed only at 0 and 6 months. To account for within-participant repeated measures, we fit an unstruc- tured covariance. We descriptively analyzed intervention en- gagement (encounter completion, SMBG transmission), ad- verse events, and costs.

Missing Data Our analyses implicitly accommodated missingness when re- lated to prior outcome data or other baseline model covari- ates defined as missing at random (MAR). As a sensitivity analy- sis for the primary model, we also multiplied imputed missing HbA1c data using a Markov chain Monte Carlo algorithm in- corporating additional variables to strengthen the MAR as- sumption (see eMethods in Supplement 2). With the pandem- ic’s influence on outcome ascertainment, we conducted another sensitivity analysis fitting our primary model with in- clusion of additional clinical HbA1c measurements obtained from the EHR.

Research Original Investigation Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes

E4 JAMA Internal Medicine Published online July 25, 2022 (Reprinted) jamainternalmedicine.com

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

Data Monitoring Committee Our data monitoring committee comprised the study statisti- cians (A.S.J. and C.J.C.) and 3 independent experts. This com- mittee met at 6-month intervals during the study and exam- ined recruitment, retention, randomization, adverse events, and outcomes.

Results Participants, Retention, and Fidelity Participants were enrolled from December 2018 through Janu- ary 2020; participant contact concluded in January 2021. Of 1128 individuals assessed, 257 were consented, and 200 were randomized (Figure 2); most of those consented but not randomized were excluded for baseline HbA1c level of less than 8.5%. Randomized participants were similar to those declining participation (eTable 1 in Supplement 2). Of those ran- domized, 101 were allocated to comprehensive telehealth and 99 to telemonitoring/care coordination; all were analyzed in their randomized group, and there was no between-arm crossover.

Per Table 1, participants had baseline mean (SD) age of 57.8 (8.2) years, HbA1c level of 10.2% (1.3%), and BMI of 34.8 (6.7). A total of 45 (22.5%) participants were women; 144 (72.0%) were Black, 11 (5.5%) were Hispanic/Latinx, 42 (21.0%) were White, and 14 (7.0%) were of other race. Prior to enrollment, 136 (68.0%) participants had received diabetes specialty care, and 12 (6.0%) had received HT care. Baseline characteristics were generally balanced across arms; moderate between- arm imbalances were noted in race, medication use, BMI, and social support.

Overall, 150 of 200 (75%) participants completed the 12- month assessment for HbA1c level, and 137 (68.5%) for survey- based outcomes (Figure 2). Participants in the comprehen- sive telehealth arm completed an average of 19.6 of 26 possible encounters; 33 participants completed 20 or fewer encoun- ters, and 14 completed 10 or fewer. Mean (SD) encounter time was 17.0 (11.2) minutes. Telemonitoring, self-management sup- port, diet/activity support, and medication management were delivered during all completed comprehensive telehealth en- counters; 30 of 101 (29.7%) participants initiated depression support based on elevated PHQ-8 score at baseline, and 23 of 90 (25.6%) at 6 months. Because the telemonitoring/care co- ordination intervention did not involve scheduled encoun- ters, encounter metrics were not tracked. Both groups expe- rienced changes in medication use during the study (eTable 2 in Supplement 2); descriptively, the comprehensive tele- health group had greater increases in use of glucagon-like pep- tide-1 receptor agonists (+15% vs +8%) and dipeptidyl pepti- dase-4 inhibitors (+5% vs +1%) from baseline to 12 months.

Primary Outcome Between baseline and 12 months, estimated HbA1c change was −1.59% (10.17% to 8.58%) in the comprehensive telehealth group and −0.98% (10.17% to 9.19%) in the telemonitoring/ care coordination group; the estimated difference of −0.61% (95% CI, −1.12% to −0.11%; P = .02) favored comprehensive tele-

health (Table 2, Figure 3). Three-way interactions between in- tervention arm, time, and key stratification variables were not statistically significant, indicating no evidence for differen- tial HbA1c effects over time based on preenrollment diabetes specialty care or study site.

Similar results were found on sensitivity analyses with MAR imputation using multiply imputed data sets (mean 12-month difference, −0.63%; 95% CI, −0.95% to −0.35%; P = .03) and inclusion of additional clinical HbA1c measures from the study period (n = 191 from 62 comprehensive tele- health and 63 telemonitoring/care coordination participants; mean 12-month difference, −0.50%; 95% CI, −0.99% to −0.01%; P = .04). Findings with baseline covariate adjust- ment (race; insulin, metformin, sulfonylurea, sodium- glucose cotransporter-2 inhibitor use; BMI; social support) were also similar (mean 12-month difference, −0.66%; 95% CI, −1.17% to −0.14%; P = .01).

On exploratory descriptive analyses (eTable 3 in Supple- ment 2), comprehensive telehealth patients who completed

Figure 2. Participant Flow

1128 Patients assessed for eligibility

257 Consented

871 Excluded 458 Ineligible for the study

1 Eligible, but goal sample reached

293 Refused to participate 119 Unable to contact

200 Randomized

101 Randomized to comprehensive telehealth intervention

52 Excluded from study 48 HbA1c <8.5%

1 Did not receive VHA care

1 Upcoming bariatric surgery 1 Hospitalized

1 Invalid HbA1c from laboratory 4 Unable to contact 1 Withdrew from the study

HbA1c

82 6 mo

101 Baseline 87 3 mo

BMI

78 9 mo 77 12 mo

15 12 mo

101 Baseline 67 6 mo

89 6 mo

Survey measures 101 Baseline

71 12 mo

99 Randomized to telemonitoring/ care coordination

HbA1c

87 6 mo

99 Baseline 85 3 mo

BMI

75 9 mo 73 12 mo

16 12 mo

99 Baseline 68 6 mo

91 6 mo

Survey measures 99 Baseline

66 12 mo

BMI indicates body mass index, calculated as weight in kilograms divided by height in meters squared; HbA1c, hemoglobin A1c; VHA, Veterans Health Administration.

Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes Original Investigation Research

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 25, 2022 E5

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

more than 20 encounters (n = 68) experienced greater HbA1c

reduction (1.84%) than those who completed 20 or fewer (n = 33, 0.79%).

Secondary Outcomes Per Table 2, the comprehensive telehealth group had a greater improvement than the telemonitoring/care coordination group at 12 months for diabetes distress (mean difference, −0.25; 95% CI, −0.42 to −0.07), diabetes self-care (mean difference, 0.51; 95% CI, 0.25 to 0.78), and self-efficacy (mean difference, 0.39; 95% CI, 0.07 to 0.71). There were no statistically significant be- tween-group differences in depressive symptoms (12 months) or BMI (6 months).

Adverse Events Adverse events were similar between arms. Comprehensive telehealth participants had 26 serious events (16 hospital- izations), 3 possibly study-related (episodes of ketoacidosis, hyperglycemia, and possible medication-related urinary infection). In the telemonitoring/care coordination group, there were 19 serious events (17 hospitalizations, 1 death), none deemed study-related. Among the 97 comprehensive telehealth participants with available SMBG data, 70 (72.1%) reported 1 or more SMBG values less than 70 mg/dL over 12 months, with a per-patient mean (SD) of 7.3 (9.8). Among the 89 telemonitoring/care coordination participants with available SMBG data, 61 (68.5%) reported 1 or more SMBG

Table 1. Baseline Characteristics, Overall and Stratified by Study Arm

Baseline characteristics

No. (%)

Overall (n = 200) Comprehensive telehealth (n = 101)

Telemonitoring/care coordination (n = 99)

Demographic characteristics

Age, mean (SD), y 57.8 (8.2) 57.7 (8.3) 57.8 (8.0)

Sex

Female 45 (22.5) 24 (23.8) 21 (21.2)

Male 155 (77.5) 77 (76.2) 78 (78.8)

Race

Black or African American 144 (72.0) 68 (67.3) 76 (76.8)

White 42 (21.0) 25 (24.8) 17 (17.2)

Other racea 14 (7.0) 8 (7.9) 6 (6.0)

Hispanic/Latino ethnicityb 11 (5.5) 6 (5.9) 5 (5.1)

Did not graduate high school 57 (28.5) 29 (28.7) 28 (28.3)

Currently married 91 (45.5) 45 (46.5) 46 (45.5)

Social supportc,d 191 (95.5) 98 (97.0) 93 (93.9)

Employed (full/part-time/self) 90 (45.0) 48 (47.5) 42 (42.4)

Study site

Durham 115 (57.5) 57 (56.4) 58 (58.6)

Richmond 85 (42.5) 44 (43.6) 41 (41.4)

Years with diabetes, mean (SD) 12.1 (7.7) 12.1 (8.0) 12.0 (7.5)

Prior diabetes specialty care 136 (68.0) 69 (68.3) 67 (67.7)

Prior Home Telehealth enrollment 12 (6.0) 6 (5.9) 6 (6.1)

Clinical measures

Baseline HbA1c, mean (SD), % 10.2 (1.3) 10.1 (1.2) 10.2 (1.4)

BMI, mean (SD) 34.8 (6.7) 34.5 (6.4) 35.2 (7.0)

Hypertension 166 (83.0) 81 (80.2) 85 (85.9)

Hyperlipidemiad 171 (85.5) 84 (83.2) 87 (87.9)

Tobacco use in past 6 mo 30 (25.4) 16 (30.2) 14 (21.5)

Metformin 160 (80.0) 78 (77.2) 82 (82.8)

Sulfonylurea 83 (41.5) 35 (34.7) 48 (48.9)

Thiazolidinedione 14 (7.0) 7 (6.9) 7 (7.1)

SGLT-2 inhibitor 22 (11.0) 15 (14.9) 7 (7.1)

GLP-1 receptor agonist 25 (12.5) 11 (10.9) 14 (14.1)

DPP-4 inhibitor 4 (2.0) 0 4 (4.0)

Insulin use 142 (71.0) 78 (77.2) 64 (64.6)

Psychosocial measures

Diabetes distress (DDS), mean (SD) 1.9 (0.8) 1.9 (0.7) 1.9 (0.9)

Diabetes self-care (DSMQ), mean (SD) 6.7 (1.6) 6.9 (1.5) 6.5 (1.7)

Self-efficacy (PCS), mean (SD) 5.2 (1.5) 5.2 (1.5) 5.2 (1.4)

Depression (PHQ-8) score, mean (SD)e 7.3 (5.7) 7.0 (5.2) 7.6 (6.1)

Abbreviations: BMI, body mass index, calculated as weight in kilograms divided by height in meters squared; DDS, Diabetes Distress Scale; DPP-4, dipeptidyl peptidase-4; DSMQ, Diabetes Self-Management Questionnaire; GLP-1, glucagon-like peptide-1; HbA1c, hemoglobin A1c; PCS, Perceived Competence Scale; PHQ-8, Patient Health Questionnaire-8; SGLT-2, sodium-glucose cotransporter-2. a Because of low numbers in the

Asian, American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, other, and unknown categories, these were combined into a single category, “other race.”

b One patient in the telemonitoring/care coordination group responded “Don’t know” to the Hispanic/Latino ethnicity question.

c Social support was assessed by asking, “Do you have someone you feel close to, someone you can trust and confide in?”

d One patient in the comprehensive telehealth group responded “Don’t know” to having high cholesterol and to having social support.

e One patient in the comprehensive telehealth group and 1 patient in the telemonitoring/care coordination group were missing the PHQ-8 score.

Research Original Investigation Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes

E6 JAMA Internal Medicine Published online July 25, 2022 (Reprinted) jamainternalmedicine.com

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

values less than 70 mg/dL over 12 months, with a per- patient mean (SD) of 6.9 (11.1).

Intervention Costs Per-patient intervention costs were $2465 for comprehen- sive telehealth and $946 for telemonitoring/care coordina- tion, for a between-arm difference of $1519 over 12 months.

Discussion We sought to promote practical use of telehealth for PPDM by comparing 2 approaches designed for feasible clinical imple- mentation: a comprehensive telehealth intervention and telemonitoring/care coordination. While both approaches improved HbA1c level, the comprehensive telehealth inter- vention produced a greater 12-month improvement in HbA1c

level and multiple secondary outcomes, without excess hypoglycemia.

These findings demonstrate that practically designed tele- health can be effective for patients whose T2D remains per- sistently poorly controlled despite clinic-based care. More- over, combining telehealth strategies to target multiple barriers to improvement lowers HbA1c level more than simpler ap- proaches like telemonitoring/care coordination. Given our ac-

tive-comparator design, we cannot exclude that the within- arm HbA1c effects (−1.59% for comprehensive telehealth and −0.98% for telemonitoring/care coordination) may partly re- flect regression to the mean; for example, the −0.98% HbA1c

improvement with telemonitoring/care coordination ex- ceeds the effect reported for telemonitoring in systematic re- views (−0.4% to 0.5% vs usual care).15,45-47 However, the rela- tive HbA1c benefit seen with comprehensive telehealth in this study (−0.61%) was not subject to regression to the mean. Im- portantly, the HbA1c reduction with comprehensive tele- health was durably retained through 12 months.

While the comprehensive telehealth intervention was more expensive (additional $1519 over 12 months), this incremen- tal cost is less than most branded glucose-lowering medica- tions, and the comprehensive approach came with added ben- efits for diabetes distress, self-care, and self-efficacy. Given the high complication rates characteristic of PPDM and the long- term cost benefits of HbA1c reduction,1,2 implementing com- prehensive telehealth in practice may represent an appropri- ate investment for health care systems in which the requisite infrastructure is or can be made available.

When we developed this study, our focus was on leverag- ing the VHA’s telehealth infrastructure to generate solutions for PPDM. Since then, the COVID-19 pandemic has driven a dra- matic upsurge in telehealth use worldwide. Currently,

Table 2. Estimated Outcome Means and Mean Differences for Comprehensive Telehealth (n = 101) and Telemonitoring/Care Coordination (n = 99) Arms by Time Pointa

Outcome

Estimated mean Estimated mean difference (95% CI) P value

Comprehensive telehealth

Telemonitoring/care coordination

HbA1c, %

Baseline 10.17 10.17 NA NA

3 mo 8.90 9.18 −0.29 (−0.48 to −0.09) NA

6 mo 8.54 9.03 −0.48 (−0.78 to −0.18) NA

9 mo 8.61 9.20 −0.60 (−0.96 to −0.22) NA

12 mo 8.58 9.19 −0.61 (−1.12 to −0.11) .02

Diabetes distress (DDS)b

Baseline 1.93 1.93 NA NA

6 mo 1.53 1.57 −0.04 (−0.18 to 0.09) NA

12 mo 1.43 1.67 −0.25 (−0.42 to −0.07) .007

Diabetes self-care (DSMQ)c

Baseline 6.67 6.67 NA NA

6 mo 8.15 7.92 0.22 (−0.07 to 0.51) NA

12 mo 8.34 7.83 0.51 (0.25 to 0.78) <.001

Self-efficacy (PCS)d

Baseline 5.20 5.20 NA NA

6 mo 6.09 5.84 0.24 (−0.06 to 0.54) NA

12 mo 6.31 5.92 0.39 (0.07 to 0.71) .02

BMI

Baseline 34.81 34.81 NA NA

6 mo 35.05 34.86 0.19 (−0.24 to 0.62) .39

Depression symptoms (PHQ-8)e

Baseline 7.32 7.32 NA NA

6 mo 6.54 6.06 0.48 (−0.72 to 1.69) NA

12 mo 4.64 5.80 −1.16 (−2.53 to 0.21) .10

Abbreviations: BMI, body mass index, calculated as weight in kilograms divided by height in meters squared; DDS, Diabetes Distress Scale; DSMQ, Diabetes Self-Management Questionnaire; HbA1c, hemoglobin A1c; NA, not applicable; PCS, Perceived Competence Scale; PHQ-8, Patient Health Questionnaire-8. a Missing data by time point for the

comprehensive telehealth group were as follows: HbA1c: 3 months n = 14, 6 months n = 19, 9 months n = 23, 12 months n = 22; survey measures: 6 months n = 12, 12 months n = 30; BMI: 6 months n = 34. Missing data by time point for the telemonitoring/care coordination group were as follows: HbA1c: 3 months n = 14, 6 months n = 12, 9 months n = 24, 12 months n = 26; survey measures: 6 months n = 10, 12 months n = 35; BMI: 6 months n = 33. No data points were missing at baseline.

b A lower score on the DDS indicates lower levels of diabetes distress, so is preferred.

c A higher score on the DSMQ indicates better diabetes self-care, so is preferred.

d A higher score on the PCS indicates higher self-efficacy, so is preferred.

e A lower score on the PHQ-8 indicates fewer depressive symptoms, so is preferred.

Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes Original Investigation Research

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 25, 2022 E7

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

telehealth, as used by most systems, implies intermittent video or phone appointments,30,48-50 with limited interim patient– clinician contact or transfer of patient-generated data into the EHR—in essence, “clinic-like” care, only delivered remotely. Just as clinic-based care does not fully address the factors un- derlying PPDM, “clinic-like” telehealth is also likely inad- equate. There has always been a sound argument for using comprehensive telehealth when clinic-based chronic disease care falls short; now that telehealth has gained wider accep- tance, systems have a clear mandate to maximize its value for those high-risk patients who respond insufficiently to clinic care. The present study provides evidence supporting com- prehensive telehealth for PPDM within the VHA, but also pre- sents a template for how other systems might use existing re- sources to improve the management of PPDM and other hard- to-treat conditions.

While our focus on examining practical, comprehensive telehealth specifically for PPDM is novel, the findings also add to the broader telehealth literature in T2D. The between-arm HbA1c effect we observed (−0.61%) is notable given our active- comparator design and the relatively modest effects reported in recent systematic reviews of telehealth interventions (mean HbA1c benefit vs usual care, −0.4%).51,52 In particular, prior RCTs examining comprehensive interventions for T2D have not low- ered HbA1c level20-25; these include studies that specifically

sought to examine multicomponent interventions vs usual care in pragmatic settings (non-VHA), with neutral results.20-22

Limitations Despite efforts to oversample women, the population demo- graphics may limit generalizability. However, the cohort’s high proportion of Black participants (72.0%) is a strength and may suggest that the pandemic-induced shift to telehealth need not exacerbate health care inequities.53,54 The studied interven- tions were designed for practical delivery within the VHA, which may limit generalizability to systems lacking capacity for nurse-delivered telehealth, integrated mental health, and dietitian services; nevertheless, the idea of designing tele- health interventions to leverage available resources is broadly applicable. The VHA costs may not fully generalize, but the be- tween-arm cost difference may be more translatable.

While clinical intervention delivery continued unim- peded during the pandemic, the missing data frequency was higher than expected at 12 months; however, the sensitivity analyses with multiple imputation and inclusion of clinical HbA1c data support the validity of the findings. Of note, par- ticipants could not be blinded to randomization status, which leaves potential for bias, especially pertaining to subjective sur- vey measures.

This study was not designed to evaluate the effective- ness of each individual component of the comprehensive in- tervention. Future mediator analyses will examine how each component contributes to the overall intervention effect. Fi- nally, the comprehensive telehealth intervention does not ac- count for all contributors to PPDM, including social determi- nants of health.

Conclusions Findings from this randomized clinical trial showed that com- pared with a simpler telehealth approach, a comprehensive telehealth intervention improved HbA1c level and other out- comes in patients with PPDM. Because this comprehensive telehealth intervention was delivered by clinical staff using ex- isting resources, it may warrant clinical implementation in sys- tems with appropriate infrastructure. More broadly, this study provides valuable comparative evidence that may help sys- tems maximize the value of telehealth during the COVID-19 pandemic and beyond.

ARTICLE INFORMATION

Accepted for Publication: May 31, 2022.

Published Online: July 25, 2022. doi:10.1001/jamainternmed.2022.2947

Author Affiliations: Durham Veterans Affairs Center of Innovation to Accelerate Discovery and Practice Transformation (ADAPT), Durham, North Carolina (Crowley, Bosworth, Jeffreys, Coffman, Maciejewski, Steinhauser, Smith, Strawbridge, Majette Elliott, Danus, Edelman); Division of Endocrinology, Department of Medicine, Duke University School of Medicine, Durham, North Carolina (Crowley); Central Virginia Veterans Affairs

Health Care System, Richmond (Tarkington, Fredrickson, Mundy, Marcano, Overby); Division of General Internal Medicine, Department of Medicine, Duke University School of Medicine, Durham, North Carolina (Bosworth, Maciejewski, Smith, Edelman); Department of Population Health Sciences, Duke University School of Medicine, Durham, North Carolina (Bosworth, Maciejewski, Steinhauser, Smith); Department of Biostatistics and Bioinformatics, Duke University School of Medicine, Durham, North Carolina (Coffman); Duke-Margolis Center for Health Policy, Duke University School of Medicine, Durham, North Carolina (Maciejewski); Greenville VA Health Care

Center, Greenville, North Carolina (Dar); Division of Endocrinology, Department of Medicine, Brody School of Medicine at East Carolina University, Greenville, North Carolina (Dar).

Author Contributions: Drs Crowley and Coffman had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Crowley, Tarkington, Bosworth, Steinhauser, Smith, Strawbridge, Majette Elliott, Edelman. Acquisition, analysis, or interpretation of data: Crowley, Tarkington, Jeffreys, Coffman, Maciejewski, Steinhauser, Smith, Dar, Fredrickson,

Figure 3. Estimated Trajectories by Arm for Hemoglobin A1c (HbA1c) Level, From Linear Mixed Models

11

10

9

8

Es tim

at ed

H bA

1c , %

Time, mo 0 3 6 9 12

Comprehensive telehealth

Telemonitoring/care coordination

Estimated difference at 12 mo (comprehensive telehealth – telemonitoring/care coordination): –0.61%; 95% CI, –1.12% to –0.11%

Error bars indicate 95% CIs.

Research Original Investigation Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes

E8 JAMA Internal Medicine Published online July 25, 2022 (Reprinted) jamainternalmedicine.com

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

Mundy, Marcano, Overby, Danus. Drafting of the manuscript: Crowley, Tarkington, Bosworth, Coffman, Steinhauser, Strawbridge. Critical revision of the manuscript for important intellectual content: Tarkington, Bosworth, Jeffreys, Coffman, Maciejewski, Steinhauser, Smith, Dar, Fredrickson, Mundy, Marcano, Overby, Majette Elliott, Danus, Edelman. Statistical analysis: Jeffreys, Coffman, Maciejewski, Smith. Obtained funding: Crowley, Smith. Administrative, technical, or material support: Crowley, Tarkington, Jeffreys, Maciejewski, Mundy, Strawbridge, Marcano, Majette Elliott, Danus. Supervision: Crowley, Tarkington, Bosworth, Smith, Edelman.

Conflict of Interest Disclosures: Dr Crowley reported grants from National Institutes of Health (1R01NR019594-01), VA Quality Enhancement Research Initiative (QUE 20-012), VA Office of Rural Health, and VA Health Services Research & Development (CDA 13-261) outside the submitted work. Dr Bosworth reported research support from Otsuka, Novo Nordisk, Sanofi, Improved Patient Outcomes, Boehringer Ingelheim, National Institutes of Health, and VA, as well as consulting fees from WebMD, Sanofi, Novartis, Otsuka, Abbott, Xcenda, Preventric Diagnostics, VIDYA, and the Medicines Company outside the submitted work. Ms Danus reported grants from Department of Veterans Affairs/VA Health Services Research and Novo Nordisk outside the submitted work. Dr Edelman reported personal fees from Department of Veterans Affairs (salary) outside the submitted work. No other disclosures were reported.

Funding/Support: This study was supported by a grant from Veterans Affairs Health Services Research and Development (VA IIR 16-213, Crowley PI). Dr Crowley acknowledges funding from the National Institutes of Health (1R01NR019594-01), the Veterans Affairs Quality Enhancement Research Initiative (VA QUE 20-012), and the Veterans Affairs Office of Rural Health; he was supported by a Career Development Award from Veterans Affairs Health Services Research and Development (CDA 13-261) during part of the study period. Drs Bosworth and Maciejewski were supported by Veterans Affairs Health Services Research and Development Senior Career Scientist Awards (VA HSR&D 08-027, VA HSR&D 10-391). The authors acknowledge in-kind support from the Durham Center of Innovation to Accelerate Discovery and Practice Transformation (VA CIN 13-410) within the Durham VA Health Care System.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of Veterans Affairs.

Meeting Presentation: Portions of the data reported in this manuscript were previously reported at the 81st American Diabetes Association Scientific Sessions; June 25-29, 2021; virtual.

Data Sharing Statement: See Supplement 3.

Additional Contributions: The authors would like to thank the following additional individuals for their contributions to this study: Deborah H. Jeter, RN (coordinated activities at Richmond, Virginia site); Steven Szabo, MD, PhD, and Shivan Desai, MD (study psychiatrists); Mary P. Garrett, MSN, RN, and Theresa C. Wilmot, RN (study intervention nurses); Melissa Durkee, PharmD, Susan Bullard, PharmD, and Janette Hiner, NP (study medication managers); and Teresa Howard, AA (research assistant at Durham, North Carolina site). Additional contributors were not directly compensated, but study funding (VA IIR 16-213, Crowley PI) provided partial salary support for Mss Jeter, Garrett, Wilmot, and Howard.

REFERENCES

1. Gilmer TP, O’Connor PJ, Rush WA, et al. Predictors of health care costs in adults with diabetes. Diabetes Care. 2005;28(1):59-64. doi:10. 2337/diacare.28.1.59

2. McBrien KA, Manns BJ, Chui B, et al. Health care costs in people with diabetes and their association with glycemic control and kidney function. Diabetes Care. 2013;36(5):1172-1180. doi:10.2337/dc12-0862

3. Stratton IM, Adler AI, Neil HA, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ. 2000;321 (7258):405-412. doi:10.1136/bmj.321.7258.405

4. Alexopoulos AS, Jackson GL, Edelman D, et al. Clinical factors associated with persistently poor diabetes control in the Veterans Health Administration: a nationwide cohort study. PLoS One. 2019;14(3):e0214679. doi:10.1371/journal.pone. 0214679

5. Kobe EA, Edelman D, Tarkington PE, et al. Practical telehealth to improve control and engagement for patients with clinic-refractory diabetes mellitus (PRACTICE-DM): protocol and baseline data for a randomized trial. Contemp Clin Trials. 2020;98:106157. doi:10.1016/j.cct.2020.106157

6. Davidson J. Strategies for improving glycemic control: effective use of glucose monitoring. Am J Med. 2005;118(suppl 9A):27S-32S. doi:10.1016/j. amjmed.2005.07.054

7. Donnelly LA, Morris AD, Evans JM; DARTS/MEMO collaboration. Adherence to insulin and its association with glycaemic control in patients with type 2 diabetes. QJM. 2007;100(6): 345-350. doi:10.1093/qjmed/hcm031

8. Hill-Briggs F, Gary TL, Bone LR, Hill MN, Levine DM, Brancati FL. Medication adherence and diabetes control in urban African Americans with type 2 diabetes. Health Psychol. 2005;24(4):349- 357. doi:10.1037/0278-6133.24.4.349

9. American Diabetes Association. 7. Obesity management for the treatment of type 2 diabetes. Diabetes Care. 2017;40(suppl 1):S57-S63. doi:10. 2337/dc17-S010

10. Chiu CJ, Wray LA. Factors predicting glycemic control in middle-aged and older adults with type 2 diabetes. Prev Chronic Dis. 2010;7(1):A08.

11. Hartz A, Kent S, James P, Xu Y, Kelly M, Daly J. Factors that influence improvement for patients with poorly controlled type 2 diabetes. Diabetes Res Clin Pract. 2006;74(3):227-232. doi:10.1016/j. diabres.2006.03.023

12. de Groot M, Anderson R, Freedland KE, Clouse RE, Lustman PJ. Association of depression and diabetes complications: a meta-analysis. Psychosom Med. 2001;63(4):619-630. doi:10.1097/ 00006842-200107000-00015

13. Morrison F, Shubina M, Turchin A. Encounter frequency and serum glucose level, blood pressure, and cholesterol level control in patients with diabetes mellitus. Arch Intern Med. 2011;171(17): 1542-1550. doi:10.1001/archinternmed.2011.400

14. Rosenthal ES, Bashan E, Herman WH, Hodish I. The effort required to achieve and maintain optimal glycemic control. J Diabetes Complications. 2011;25 (5):283-288. doi:10.1016/j.jdiacomp.2011.01.002

15. Medical Advisory Secretariat. Home telemonitoring for type 2 diabetes: an evidence-based analysis. Ont Health Technol Assess Ser. 2009;9(24):1-38.

16. Medical Advisory Secretariat. Behavioural interventions for type 2 diabetes: an evidence-based analysis. Ont Health Technol Assess Ser. 2009;9(21):1-45.

17. Kempf K, Altpeter B, Berger J, et al. Efficacy of the telemedical lifestyle intervention program TeLiPro in advanced stages of type 2 diabetes: a randomized controlled trial. Diabetes Care. 2017; 40(7):863-871. doi:10.2337/dc17-0303

18. Pimouguet C, Le Goff M, Thiébaut R, Dartigues JF, Helmer C. Effectiveness of disease-management programs for improving diabetes care: a meta-analysis. CMAJ. 2011;183(2):E115-E127. doi: 10.1503/cmaj.091786

19. Atlantis E, Fahey P, Foster J. Collaborative care for comorbid depression and diabetes: a systematic review and meta-analysis. BMJ Open. 2014;4(4): e004706. doi:10.1136/bmjopen-2013-004706

20. Lauffenburger JC, Ghazinouri R, Jan S, et al. Impact of a novel pharmacist-delivered behavioral intervention for patients with poorly-controlled diabetes: the ENhancing outcomes through Goal Assessment and Generating Engagement in Diabetes Mellitus (ENGAGE-DM) pragmatic randomized trial. PLoS One. 2019;14(4):e0214754. doi:10.1371/journal.pone.0214754

21. Sarayani A, Mashayekhi M, Nosrati M, et al. Efficacy of a telephone-based intervention among patients with type-2 diabetes: a randomized controlled trial in pharmacy practice. Int J Clin Pharm. 2018;40(2):345-353. doi:10.1007/s11096-018- 0593-0

22. Edelman D, Dolor RJ, Coffman CJ, et al. Nurse-led behavioral management of diabetes and hypertension in community practices: a randomized trial. J Gen Intern Med. 2015;30(5): 626-633. doi:10.1007/s11606-014-3154-9

23. Lee JY, Chan CKY, Chua SS, et al. Telemonitoring and team-based management of glycemic control on people with type 2 diabetes: a cluster-randomized controlled trial. J Gen Intern Med. 2020;35(1):87-94. doi:10.1007/s11606-019- 05316-9

24. Crowley MJ, Powers BJ, Olsen MK, et al. The Cholesterol, Hypertension, And Glucose Education (CHANGE) study: results from a randomized controlled trial in African Americans with diabetes. Am Heart J. 2013;166(1):179-186. doi:10.1016/j.ahj. 2013.04.004

25. Tang PC, Overhage JM, Chan AS, et al. Online disease management of diabetes: engaging and

Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes Original Investigation Research

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online July 25, 2022 E9

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022

motivating patients online with enhanced resources-diabetes (EMPOWER-D), a randomized controlled trial. J Am Med Inform Assoc. 2013;20(3): 526-534. doi:10.1136/amiajnl-2012-001263

26. Adler-Milstein J, Kvedar J, Bates DW. Telehealth among US hospitals: several factors, including state reimbursement and licensure policies, influence adoption. Health Aff (Millwood). 2014;33(2):207-215. doi:10.1377/hlthaff.2013.1054

27. Glasgow RE, Lichtenstein E, Marcus AC. Why don’t we see more translation of health promotion research to practice? rethinking the efficacy-to-effectiveness transition. Am J Public Health. 2003;93(8):1261-1267. doi:10.2105/AJPH.93. 8.1261

28. Scott Kruse C, Karem P, Shifflett K, Vegi L, Ravi K, Brooks M. Evaluating barriers to adopting telemedicine worldwide: a systematic review. J Telemed Telecare. 2018;24(1):4-12. doi:10.1177/ 1357633X16674087

29. Ross J, Stevenson F, Lau R, Murray E. Factors that influence the implementation of e-health: a systematic review of systematic reviews (an update). Implement Sci. 2016;11(1):146. doi:10.1186/ s13012-016-0510-7

30. Wosik J, Fudim M, Cameron B, et al. Telehealth transformation: COVID-19 and the rise of virtual care. J Am Med Inform Assoc. 2020;27(6):957-962. doi:10.1093/jamia/ocaa067

31. Crowley MJ, Edelman D, McAndrew AT, et al. Practical telemedicine for veterans with persistently poor diabetes control: a randomized pilot trial. Telemed J E Health. 2016;22(5):376-384. doi:10.1089/tmj.2015.0145

32. Draznin B, Aroda VR, Bakris G, et al; American Diabetes Association Professional Practice Committee. 6. Glycemic targets: Standards of Medical Care in Diabetes—2022. Diabetes Care. 2022;45(S1)(suppl 1):S83-S96.

33. Draznin B, Aroda VR, Bakris G, et al; American Diabetes Association Professional Practice Committee. 8. Obesity and weight management for the prevention and treatment of type 2 diabetes: Standards of Medical Care in Diabetes—2022. Diabetes Care. 2022;45(S1)(suppl 1):S113-S124.

34. Draznin B, Aroda VR, Bakris G, et al; American Diabetes Association Professional Practice Committee; American Diabetes Association Professional Practice Committee. 5. Facilitating behavior change and well-being to improve health outcomes: Standards of Medical Care in Diabetes—2022. Diabetes Care. 2022;45(S1)(suppl 1):S60-S82.

35. US Department of Veterans Affairs. VA/DoD clinical practice guideline for the management of major depressive disorder. Accessed May 4, 2022. https://www.healthquality.va.gov/guidelines/MH/ mdd/VADoDMDDCPGFINAL82916.pdf

36. Polonsky WH, Fisher L, Earles J, et al. Assessing psychosocial distress in diabetes: development of the diabetes distress scale. Diabetes Care. 2005;28 (3):626-631. doi:10.2337/diacare.28.3.626

37. Schmitt A, Gahr A, Hermanns N, Kulzer B, Huber J, Haak T. The Diabetes Self-Management Questionnaire (DSMQ): development and evaluation of an instrument to assess diabetes self-care activities associated with glycaemic control. Health Qual Life Outcomes. 2013;11:138. doi:10.1186/1477-7525-11-138

38. Williams GC, Freedman ZR, Deci EL. Supporting autonomy to motivate patients with diabetes for glucose control. Diabetes Care. 1998;21(10):1644- 1651. doi:10.2337/diacare.21.10.1644

39. Kroenke K, Strine TW, Spitzer RL, Williams JB, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1-3):163-173. doi:10.1016/j. jad.2008.06.026

40. Bent S, Padula A, Avins AL. Brief communication: Better ways to question patients about adverse medical events: a randomized, controlled trial. Ann Intern Med. 2006;144(4):257- 261. doi:10.7326/0003-4819-144-4-200602210- 00007

41. Sacks DB, Arnold M, Bakris GL, et al; National Academy of Clinical Biochemistry; Evidence-Based Laboratory Medicine Committee of the American Association for Clinical Chemistry. Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Diabetes Care. 2011;34(6):e61-e99. doi:10.2337/ dc11-9998

42. Little R, Kang S. Intention-to-treat analysis with treatment discontinuation and missing data in clinical trials. Stat Med. 2015;34(16):2381-2390. doi: 10.1002/sim.6352

43. Fitzmaurice GM, Laird NM, Ware JH. Applied Longitudinal Analysis. 2nd ed. Wiley; 2011. doi:10. 1002/9781119513469

44. Akaike H. A new look at the statistical model identification. IEEE Trans Automat Contr. 1974;19 (6):716-723. doi:10.1109/TAC.1974.1100705

45. Zhu X, Williams M, Finuf K, et al. Home telemonitoring of patients with type 2 diabetes:

a meta-analysis and systematic review. Diabetes Spectr. 2022;35(1):118-128. doi:10.2337/ds21-0023

46. Kim Y, Park JE, Lee BW, Jung CH, Park DA. Comparative effectiveness of telemonitoring versus usual care for type 2 diabetes: a systematic review and meta-analysis. J Telemed Telecare. 2019;25(10): 587-601. doi:10.1177/1357633X18782599

47. Lee PA, Greenfield G, Pappas Y. The impact of telehealth remote patient monitoring on glycemic control in type 2 diabetes: a systematic review and meta-analysis of systematic reviews of randomised controlled trials. BMC Health Serv Res. 2018;18(1):495. doi:10.1186/s12913-018-3274-8

48. Baum A, Kaboli PJ, Schwartz MD. Reduced in-person and increased telehealth outpatient visits during the COVID-19 pandemic. Ann Intern Med. 2021;174(1):129-131. doi:10.7326/M20-3026

49. Heyworth L, Kirsh S, Zulman D, Ferguson JM, Kizer KW. Expanding access through virtual care: the VA’s early experience with COVID-19. NEJM Catalyst. July 1, 2020. Accessed January 18, 2022. https://catalyst.nejm.org/doi/full/10.1056/cat.20. 0327

50. Artandi M, Thomas S, Shah NR, Srinivasan M. Rapid system transformation to more than 75% primary care video visits within three weeks at Stanford: response to public safety crisis during a pandemic. NEJM Catalyst. April 21, 2020. Accessed January 18, 2022. https://catalyst.nejm.org/doi/full/ 10.1056/CAT.20.0100

51. Hangaard S, Laursen SH, Andersen JD, et al. The effectiveness of telemedicine solutions for the management of type 2 diabetes: a systematic review, meta-analysis, and meta-regression. J Diabetes Sci Technol. Published online December 26, 2021. doi:10.1177/19322968211064633

52. Faruque LI, Wiebe N, Ehteshami-Afshar A, et al; Alberta Kidney Disease Network. Effect of telemedicine on glycated hemoglobin in diabetes: a systematic review and meta-analysis of randomized trials. CMAJ. 2017;189(9):E341-E364. doi:10.1503/cmaj.150885

53. Litchfield I, Shukla D, Greenfield S. Impact of COVID-19 on the digital divide: a rapid review. BMJ Open. 2021;11(10):e053440. doi:10.1136/bmjopen- 2021-053440

54. Eruchalu CN, Pichardo MS, Bharadwaj M, et al. The expanding digital divide: digital health access inequities during the COVID-19 pandemic in New York City. J Urban Health. 2021;98(2):183-186. doi: 10.1007/s11524-020-00508-9

Research Original Investigation Comprehensive Telehealth Intervention vs Telemonitoring and Care Coordination in Type 2 Diabetes

E10 JAMA Internal Medicine Published online July 25, 2022 (Reprinted) jamainternalmedicine.com

© 2022 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Duke Medical Center Library User on 07/25/2022