Program completion and glycemic control in a remote patient monitoring program for diabetes management: Does gender matter?

diabetes research and clinical practice

159 (2020) 107944

Contents available at ScienceDirect

Diabetes Research and Clinical Practice journal homepage: www.elsevier.com/locat e/dia bre s

Program completion and glycemic control in a remote patient monitoring program for diabetes management: Does gender matter? Tzeyu L. Michaud a,b,*, Mohammad Siahpush b, Keyonna M. King a,b, Athena K. Ramos a, Regina E. Robbins b, Robert J. Schwab c, Martina A. Clarke d, Dejun Su a,b a

Center for Reducing Health Disparities, College of Public Health, University of Nebraska Medical Center, Omaha, NE, USA Department of Health Promotion, College of Public Health, University of Nebraska Medical Center, Omaha, NE, USA c Division of General Internal Medicine, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE, USA d Division of Cardiovascular Medicine, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE, USA b

A R T I C L E I N F O

A B S T R A C T

Article history:

Aims: To examine gender differences in program completion and glycemic outcomes for

Received 11 July 2019

patients with type 2 diabetes (T2D) in a remote patient monitoring (RPM) program for dia-

Received in revised form

betes management.

13 November 2019

Methods: Based on data from an RPM program that enrolled post-discharge T2D patients

Accepted 19 November 2019

(n = 1645) in 2014–2017, logistic regression models were estimated to assess gender differ-

Available online 23 November 2019

ence in the likelihood of completing the three-month RPM program; whereas ordinary least squares (OLS) regression models were used to examine gender difference in post-RPM

Keywords: Program participation HbA1c Type 2 diabetes Telemedicine Gender disparities Engagement

hemoglobin A1c (HbA1c), controlling for demographics, baseline health status, including HbA1c, patient activation scores, and physiological data upload frequency for patients who had completed the program. Results: Among enrolled participants, men had lower odds of completing the three-month RPM program than women (adjusted odds ratio, 0.61; 95% confidence interval [CI], 0.39–0.95). However, among those who completed the program, men had lower post-RPM HbA1c than women ( 0.18; 95% CI,

0.33,

0.03) after controlling for baseline HbA1c and

other covariates. Conclusions: While female patients with T2D were more likely to complete the RPM program, they showed a higher glycemic level at the end of the program compared to male patients. To close gender disparities in health, interventions through telemedicine tailored towards women’s diabetes outcomes and men’s engagement level are warranted. Ó 2019 Elsevier B.V. All rights reserved.

1.

Introduction

Type 2 diabetes (T2D), commonly accompanied by cardiovascular disease (CVD) risk factors (e.g. hypertension, high

cholesterol, and smoking), has posed a major challenge to society, patients, and families, as an emerging global pandemic [1]. A recent study using the National Health and Nutrition Examination Survey data reported that diabetes and

* Corresponding author at: Center for Reducing Health Disparities, 984340 Nebraska Medical Center, Omaha, NE 68198, USA. E-mail address: [email protected] (T.L. Michaud). https://doi.org/10.1016/j.diabres.2019.107944 0168-8227/Ó 2019 Elsevier B.V. All rights reserved.

2

diabetes research and clinical practice

prediabetes (including undiagnosed cases) affected almost half of the US adult population. However the results also exhibited a sex bias with the potential underestimation for the prevalence of diabetes and prediabetes in women due to physiological differences between men and women [2,3]. Nevertheless, increasing evidence has shown that diabetes adversely affects women more than men despite higher prevalence in men [3,4]. Compared to men with the disease, women with T2D, regardless of menopausal status, had up to 27% higher excess risk of stroke and 44% higher excess risk of coronary heart disease [5–7]. They also experience depression and anxiety and thus lowered quality of life, which in turn may negatively affect attitudes towards diabetes selfmanagement and associated behaviors [8–10]. In addition to physiological and biological differences, behavioral differences in health-seeking, adherence to treatment and medication regimes, and access to healthcare may also contribute to gender disparities in diabetes care and/or outcomes [11–14,3]. Telemedicine (or telehealth), including remote patient monitoring (RPM), involves the use of telecommunications and virtual technology to deliver health care outside of traditional health-care facilities [15] and to enhance the clinical health status of patients [16]. Several studies have reported that telemedicine can improve diabetes management through better metabolic and glycemic control [17–20], and may also be cost-effective [21]. It is important to investigate gender-specific differences in diabetes outcomes to guide the necessary changes in healthcare provision [10], especially in the telemedicine setting as it has been widely used for diabetes management and self-management due to its accessibility. Although there has been some research examining patients with T2D who do not participate in telehealth intervention studies and no sex or gender differences have been found [22,23], it is less known how gender mediates the clinical outcomes in a RPM program for post-discharge diabetes management. This study aims to examine gender-specific differences in: (1) the completion of a 3-month RPM program among individuals with T2D, and (2) hemoglobin A1c (HbA1c) outcomes among completers, to close the research to practice gap.

2.

Materials and methods

2.1.

Study setting and study sample

We conducted a retrospective, observational analysis of data from a RPM program, which has been described previously [17,24,25]. In brief, patients with T2D with a hospital admission for any reason were recruited to enroll in the program to improve disease management no later than one month after discharge from Nebraska Medicine in Omaha, Nebraska, USA. Other program enrollment criteria included: (1) being 19 years of age or older; (2) able to use their own glucometer and self-administer insulin and/or other prescribed medications; (3) not pregnant; (4) having no history of addiction; (5) able to speak and read English; (6) having a discharge planned to home; and (7) able to express a basic understanding of and successfully use RPM equipment.

159 (2020) 107944

The three-month program entailed daily remote monitoring of biometric data, including blood pressure, weight, and glucose; and weekly phone calls or instant calls from nurse coaches when an alert was issued in the monitoring system due to out of range results. Throughout the course of the intervention, during weekly phone calls, nurse coaches provided individualized education services based on patients’ experiences and reported challenges which included diet, exercise, medication adherence, coping, and problem solving, following the American Diabetes Association (ADA) workbook for self-care [26]. Participants’ biometric data (HbA1c, weight and blood pressure), which was automatically uploaded to the monitoring portal via the remote monitoring device, was shared with their primary care providers using electronic health records. Remote monitoring equipment (Cardiocom, Medtronic Inc., Minneapolis, MN) included a cellular base unit, blood pressure cuff, blood glucose meter, weight scale, and necessary cords. From 2014 to 2017, a total of 1883 patients with T2D were enrolled. We excluded 4 patients from the analysis sample due to missing information on gender. We further excluded 234 patients from the sample due to missing baseline HbA1c data, indicating they had not started the RPM program. As a result, in this retrospective, observational study, 1645 enrolled patients were included in the post-hoc analysis. The RPM program was launched for the quality improvement purpose, not so much for research. As a result, the study protocol was waived by the Institutional Review Board.

2.2.

Outcome variables

2.2.1.

Completion status of RPM

Completion status of RPM, a dichotomous variable (Yes or No), was used to indicate whether enrolled participants had completed the three-month RPM program. In addition to daily uploading of biometric data starting from the date that the remote monitoring equipment was installed and activated in their home by medical assistants to the date that the equipment was returned, participants also had to visit one of three local community health centers where they received (1) a diabetic retinopathy screening, (2) a virtual nutritional counseling session, and (3) a foot exam from a Certified Diabetes Educator (CDE) assisted by an on-site medical assistant remotely when they approached the end of the 90-day program. For the virtual counseling session, specifically, a CDE would review participants’ overall glucose readings, dietary changes that are made by participants over the course of the 90 days as well as how overall perceived health had changed over this period. Moreover, CDEs would encourage and support participants in individual planning to continue making or maintaining positive changes in their ongoing nutritional plan after leaving the program. Participants were designated as completing the program upon completion of all the above activities.

2.2.2.

Hemoglobin A1c

HbA1c was measured at both baseline and program completion via self-uploading using the remote monitoring equipment.

diabetes research and clinical practice

2.3.

Predictive variables

Demographic information included age at baseline, gender, and self-reported race (i.e., white, African American, Hispanic, Asian, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander and other). We further categorized participants into non-Hispanic white or non-white as most of the study participants were non-Hispanic white (65%) and African Americans (28%). Primary health insurance was categorized into public insurance (Medicare and Medicaid) or private insurance in the regression analyses. High blood pressure, considered an important CVD risk factor, was defined as >140/90 mmHg [27]. Changes in body weight was calculated by using weight at end of RPM minus weight at baseline. We used the 13-item Patient Activation Measure (PAM) to measure patients’ knowledge, skill, and confidence for self-management of chronic conditions [28]. PAM is a Likert scale with five response categories ranging from 1 to 5: (1) strongly disagree, (2) disagree, (3) agree, (4) strongly agree, and (5) not applicable. The measure is scored on a theoretical 0– 100 scale with most patients falling in the 35–95 point range. The following four levels of activation, reflecting a developmental progression from passive receipt of care toward greater activation, have been previously identified : Level 1 the patient does not yet believe they are active or have an important role in managing their health (score < 47.0); Level 2 - the patient lacks confidence and knowledge to take action to manage their health (scores between 47.1 and 55.1); Level 3 - the patient is beginning to take action to manage their health (scores between 55.2 and 67.0); and Level 4 - the patient is maintaining actions of managing their health over time (scores  67.1) [29,30]. The change in the activation score was calculated from the difference between baseline and at the completion of the RPM program. Finally, we included the daily upload frequency in the analysis, which was calculated by using the total number of uploads of biometric data during the program period divided by the total number of intervention days. Our previous work indicated higher daily data upload frequency was associated with better glycemic control outcomes [24].

2.4.

Data analysis

Means and standard deviations (SD) were calculated for continuous variables, whereas frequencies and percentages were calculated for categorical variables. Significant gender differences in demographics and clinical characteristics were determined using the t tests for continuous data and chisquared tests for categorical data. In the multivariable analysis, logistic regression models were used to calculate adjusted odds ratios (ORs) to determine if there were gender differences in the likelihood of completing the three-month RPM program. Among participants who had completed the program, we conducted ordinary least squares (OLS) regression analysis to determine whether there was a gender difference in the HbA1c level at the completion of RPM, controlling for demographics, clinical characteristics, patient activation scores, and daily upload

159 (2020) 107944

3

frequency. Changes in body weight between baseline and the end of RPM were not included in the logistic regression analysis as a predictive variable. We were not able to collect weight data at the end of RPM for participants who did not complete the program. All analyses were conducted using Stata version 14 (StataCorp, College Station, TX, USA). Two-tailed p values of less than 0.05 were considered statistically significant for this study.

3.

Results

3.1.

Participant characteristics

Of the 1645 patients with T2D enrolled in the RPM program, approximately 55% were women (Table 1). Compared to men, women tended to be younger (58.0 ± 12.3 vs. 59.7 ± 12.3) and a higher proportion of them were minorities (39% versus 27%). Women had a higher percentage of public health insurance coverage compared to men (50% versus 45%) and a higher mean baseline BMI (36.8 ± 9.4 vs. 34.3 ± 7.6). Men had a higher mean baseline HbA1c (7.9 ± 2.1 vs. 7.7 ± 2.0), mean baseline weight (216.5 ± 69.0 vs. 194.8 ± 65 .2), and greater prevalence of hypertension (27% versus 26%) than women. Women and men shared similar composition from Level 1 to Level 4 PAM scores and the mean daily upload frequency was 0.6 with standard deviation of 0.3. For the 1350 participants who completed the RPM program (out of total 1645, 82% completion rate), there were similar patterns of gender-specific differences in baseline characteristics except for daily upload frequency and post-RPM HbA1c. Among this subgroup, men had higher mean daily upload frequency than women (0.8 ± 0.2 vs. 0.7 ± 0.2) and the mean postRPM HbA1c was lower in men than women (7.0 ± 0.1 vs. 7.2 ± 0.1). When comparing the baseline characteristics for the enrolled patients by gender, we found significant differences between women and men in terms of age, race, primary health insurance, and weight (Table 1). Moreover, we found significant differences by gender when comparing baseline characteristics for those who completed the RPM program except for variables of primary health insurance, HbA1c and the status of high blood pressure at baseline. We also found significant differences by the RPM completion status in terms of baseline characteristics, with the exception of the PAM score category, when comparing those who completed the program vs. those who did not complete the program (Appendix Table A1).

3.2.

Gender difference on the completion of RPM

Table 2 shows the logistic regression results of the likelihood of completing the RPM program in terms of participants’ gender, while controlling for other characteristics. The odds of male participants completing the program was lower than for female participants (OR, 0.61; 95% confidence interval [CI], 0.39–0.95). We did not find a significant difference on the completion of RPM in terms of race/ethnicity.

4

diabetes research and clinical practice

159 (2020) 107944

Table 1 – Participant characteristics by gender. Variable

RPM completion status, N (%) Not completed Completed Age at baseline (year), mean (SD) Race, N (%) Non-Hispanic Whites African Americans Othersa Unknown/missing Primary health insurance at baseline, N Medicare Medicaid Private Unknown/missing HbA1c at baseline, mean (SD) HbA1c at the end of RPM, mean (SD) Weight at baseline, mean (SD) Weight at the end of RPM, mean (SD) BMI at baseline, mean (SD) BMI at the end of RPM, mean (SD) High blood pressure at baseline, N (%) No Yes Unknown/missing PAM score category at baseline, N (%) Level 1 (47) Level 2 (47.1–55.1) Level 3 (55.2–67.0) Level 4 (67.1) Unknown/missing Daily upload frequency, mean (SD)

All enrolled patients

Patients who completed the RPM

Women n = 906

Men n = 739

165 (18) 741 (82) 58.0 (12.3)

130 (18) 609 (82) 59.7 (12.3)

550 (61) 355 (39) 1 (0)

536 (73) 202 (27) 1 (0)

(%) 401 (44) 50 (6) 379 (42) 76 (8) 7.7 (2.0) – 194.8 (65.2) – 37.1 (10.0) –

316 (43) 16 (2) 335 (45) 72 (10) 7.9 (2.1) – 216.5 (69.0) – 34.6 (8.1) –

671 (74) 234 (26) 1 (0.1)

540 (73) 199 (27) 0

77 (9) 168 (18) 285 (32) 338 (37) 38 (4) 0.7 (0.3)

70 (10) 134 (18) 253 (34) 266 (36) 16 (2) 0.7 (0.3)

P-value

Women n = 741

Men n = 609

– – 59.0 (11.7)

– – 60.5 (11.8)

449 (61) 246 (33) 45 (6.0) 1 (0.1)

463 (76) 117 (19) 29 (4.8) 0

347 (47) 29 (4) 312 (42) 53 (7) 7.6 (1.9) 7.2 (0.1) 208.6 (56.3) 209.1 (2.0) 36.8 (9.4) 36.3 (8.7)

275 (45) 12 (2) 277 (46) 45 (7) 7.8 (2.0) 7.0 (0.1) 230.6 (58.2) 232.8 (2.2) 34.3 (7.6) 33.8 (6.8)

561 (76) 180 (24) –

459 (75) 150 (25) –

58 (8) 136 (18) 237 (32) 279 (38) 31 (4) 0.7 (0.2)

65 (11) 108 (18) 209 (34) 216 (36) 11 (2) 0.8 (0.2)

0.744

0.007 <0.001

–

0.003

0.099 <0.001 <0.001

0.017 <0.001

0.152

0.589

0.151

0.135

P-value

0.135 0.042 <0.001 <0.001 <0.001 <0.001 0.885

0.039

0.032

RPM, remote patient monitoring; BMI, body mass index; PAM, patient activation measurement. a Others included Hispanics, Asians, American Indian or Alaska Natives, Native Hawaiian or other Pacific Islanders, and individuals with two or more races.

3.3.

Gender difference on the HbA1c outcome

OLS regression results showed that among those who completed the RPM program, the post-RPM HbA1c was 0.18-point (95% CI, 0.33, 0.03) lower in male participants compared to female participants, controlling for baseline participant characteristics, activation, and other clinical outcomes (Table 3). Stratified by gender, for every pound of weight lost, there was a 0.018 and 0.017-point reduction, respectively, on the HbA1C measured at the end of the RPM among women and men (data not shown). We did not find a significant difference on the post-RPM HbA1c outcome in terms of race/ethnicity.

4.

Discussion

In this study, we aimed to examine the role of gender in the completion of a three-month RPM program for patients with T2D and its further impact on the post-RPM HbA1c among those who completed the RPM program. Among 1645 patients enrolled in the RPM program, 82% (withdrawal rate of 18%) completed the program and showed overall improved HbA1c level. This finding is comparable to a review study including

37 home telehealth studies for patients with heart failure and chronic obstructive pulmonary disease which found an average withdrawal (participants who agreed to participate but declined later) rate of 20%, with variation between 4% and 55% for the individual studies [31]. The most common reasons of withdrawal based on the findings from the review study were participants not wanting to use the telehealth device, health deterioration, and technical problems. Indeed, patient and/or clinician acceptance has been identified as one of the most important influences on the future implementation of telehealth [31–33]; and some patients prefer face-toface care, rather than healthcare delivered remotely, an essential component of telehealth interventions [22]. As demonstrated in our study, participants who completed the RPM were those who had higher technology engagement (daily upload frequency) and had a higher level of knowledge, skill, and confidence for disease self-management (measured by baseline PAM scores) than those who did not. Interestingly, one of the 21 studies included in the aforementioned review study reported detailed demographics for those who withdrew and found 100% were male. This is comparable with what we have found; male participants were more likely to discontinue the RPM than female participants.

diabetes research and clinical practice

Table 2 – Adjusted Odds ratios for the effect of gender and other covariates on completing the PRM program, n = 1395. Variable

OR

95% CI

P-value

gender (ref = female) Male Age at baseline A1c at baseline

0.61 1.01 0.98

(0.39, 0.95) (0.99, 1.03) (0.89, 1.09)

0.030 0.219 0.734

Race/ethnicity (ref = non-Hispanic white) 0.87 (0.55, 1.37) Not whitea

0.540 b

Primary health insurance (ref = Public insurance ) Private 0.92 (0.57, 1.49) 0.740 Weight at baseline 1.02 (1.02, 1.02) <0.001 Having high blood pressure at baseline (Ref = No) Yes 0.78 (0.48, 1.26) 0.303 Baseline patient activation score Level 2 (47.1–55.1) 0.99 Level 3 (55.2–67.0) 1.12 Level 4 (67.1) 1.12 Daily upload frequency 156.36

(ref = Level 1 [ (0.42, 2.32) (0.52, 2.42) (0.52, 2.42) (67.78, 360.69)

47]) 0.980 0.776 0.773 <0.001

a

Not white included African Americans, Hispanics, Asians, American Indian or Alaska Natives, Native Hawaiian or other Pacific Islanders, and others. b Public health insurance included Medicare and Medicaid.

Corroborating previous studies showing that male patients generally had better blood pressure and glycemic control than female patients with similar characteristics [34–36], we found that among participants who completed the program, men had better HbA1c control than women (reduction of 0.8% vs. 0.4%). It is postulated that male patients might respond better

5

159 (2020) 107944

to coaching by a female nurse coach (psychosocial factors) [17] due to sex differences on biological and endocrine factors (e.g. sex hormones) as suggested in the previous studies [3,37–39]. Historically, men were more likely to refuse to participate in and/or discontinue the research studies/trials [40,23], however, once they sustained and completed the program they presented better outcomes as indicated in our study. This may be partially explained by structural factors that men are more likely to receive more intensive medical treatment than women with similar conditions [41] due to the relative socioeconomic advantages of men [42,43]. Conversely, in a post-hoc analysis using 13-year follow-up data of a cluster-randomized controlled trial targeting patients with newly diagnosed T2D, Krag and colleagues indicated that structured personal diabetes care led by general practitioners reduced all-cause mortality and diabetes-related death (after adjusting HbA1c) in women but not in men compared with routine care [43]. Despite these inconsistent findings, similar to our case, the authors concluded that gender difference in diabetes-related outcomes (HbA1c or diabetes-related deaths) should be interpreted in the context of social and cultural interaction in addition to biological issues. Additionally, the use of different disease management approaches (in-person structured personal care vs. RPM in our study) and other intervention program features (e.g. the length of the intervention and the healthcare system) may also contribute to the inconsistent findings. The gender differences in completion status and glycemic outcomes of our study also point to the importance of incorporating patients’ gender in clinical decision making, such as in the choice of diagnostic tests, medications, and other treatments, to improve treatment outcomes for female patients

Table 3 – Ordinary least square regression results of the association between gender and other covariates, and post-RPM HbA1c, n = 1104. Variable

Coefficient

95% CI

P-value

gender (ref = female) Male Age at baseline A1c at baseline

0.18 0.01 0.39

( 0.33, 0.03) (0.00, 0.02) (0.36, 0.43)

0.018 0.024 <0.001

0.20

(0.04, 0.35)

0.014

Primary health insurance (ref = Public insurance ) Private insurance 0.05 Weight at baseline 0.001 Weight change 0.02

( 0.10, 0.21) ( 0.00, 0.00) (0.01, 0.03)

0.506 0.430 <0.001

Having high blood pressure at baseline (ref = no) Yes

(0.08, 0.41)

0.003

( ( ( ( (

0.258 0.020 0.052 0.015 <0.001

Race/ethnicity (ref = non-Hispanic whites) Not whitea b

0.25

Baseline patient activation score (ref = Level 1 [ 47]) Level 2 (47.1–55.1) 0.16 Level 3 (55.2–67.0) 0.31 Level 4 (67.1) 0.28 PAM score changes during PRM 0.01 Daily upload frequency 0.77

0.45, 0.12) 0.58, 0.05) 0.56, 0.00) 0.01, 0.00) 1.11, 0.44)

a Not white included African Americans, Hispanics, Asians, American Indian or Alaska Natives, Native Hawaiian or other Pacific Islanders, and others. b Public health insurance included Medicare and Medicaid.

6

diabetes research and clinical practice

and to reduce gender disparity [12,10]. For example, studies have reported that women were more likely than men to have poor chronic medication adherence. They were also less likely to receive recommended medication treatment and monitoring for diabetes and cardiovascular conditions [41], and had a poor quality of diabetes care [14,44]. Moreover, earlier research also showed that women were more susceptible to the side effects of the pharmacological treatment for T2D, such as statins or insulin therapies [45,46]. The presence of gender-specific differences in predisposition, development, and clinical presentation of T2D are related to diversities in biology, culture, lifestyle, environment, and socioeconomic status [9–14,41]. Telemedicine application for diabetes management tailored by gender, especially among postmenopausal women with a high incidence of CVD found in these groups [6,47], and men’s engagement level, might lead to better population health outcomes. Some limitations need to be noted. First, our analysis may be subjected to potential selection bias. That is, patients who self-selected to enroll and complete the RPM were more motivated or have specific idiosyncrasies than those who declined or did not complete the program; and thus were more likely to have a better post-RPM HbA1c outcome. Indeed, in the bivariate analysis, completers and non-completers differed significantly in terms of baseline age, HbA1c, weight, and hypertension, race, primary health insurance, and daily upload frequency. Second, the list of possible covariates in our study is incomplete. Data regarding physical exercise, diet, alcohol consumption, or smoking history, which may all contribute to explaining the gender gap in the postHbA1c outcome, were not available. Finally, we were also not able to account for the effects of medication adherence, medication usages (e.g. glucose lowering and lipid-lowering drugs) dispensing, and dosage due to data constraints. This may adversely affect patients’ diabetes outcomes as women were less likely than men to be adherent with their use of chronic medications [41], and women and men could have different reaction and presentation to pharmacological treatments [45,46,48,49] as discussed above. However, some RPM program features may partially alleviate this issue. For example, participants in the RPM program were recruited after discharging from a hospitalization and one of the inclusion criteria of the RPM program was the ability to use their own glucometer and self-administer insulin and/or other prescribed medications. In this case, it is highly likely that partic-

159 (2020) 107944

ipants have been prescribed anti-hyperglycemic drugs (e.g. metformin) based on the ADA practice guideline and recommendations [50]. Furthermore, during the weekly phone calls throughout the program course, medication adherence was one part of the individualized education services provided by nurse coaches. As telemedicine becoming an increasingly adopted mode of service and education delivery for diabetes management, the present study calls for improving awareness of genderspecific risk factors related to patient engagement and glycemic outcomes. More research is needed to better understand gender differences in program participation and reaction to telemedicine interventions to promote more personalized diabetes care and to reduce gender disparities in diabetes care through telemedicine.

Declaration of Competing Interest No competing financial interests exist.

Acknowledgements The remote patient monitoring program as described in this study was supported by Grant Number 1C1CMS331344 from the U.S. Department of Health and Human Services, Centers for Medicare & Medicaid Services (CMS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies. The research presented here was conducted by the awardee. Findings from this study might or might not be consistent with or confirmed by the findings of the independent evaluation contractor hired by CMS.

Funding source of this manuscript This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Appendix (see Table A1).

diabetes research and clinical practice

7

159 (2020) 107944

Table A1 – Comparison of demographics and clinical characteristics between RPM completers and non-completers.

Age at baseline (year), mean (SD) Race, N (%) Non-Hispanic Whites Not white Unknown/missing Primary health insurance at baseline, N (%) Public Private Unknown/missing HbA1c at baseline, mean (SD) HbA1c at the end of RPM, mean (SD) Weight at baseline, mean (SD) Weight at the end of RPM, mean (SD) BMI at baseline, mean (SD) BMI at the end of RPM, mean (SD) High blood pressure at baseline, N (%) No Yes Unknown/missing PAM score category at baseline, N (%) Level 1 (47) Level 2 (47.1–55.1) Level 3 (55.2–67.0) Level 4 (67.1) Unknown/missing Daily upload frequency, mean (SD)

Completers N = 1350

Non-completers N = 295

P-value

59.6 (11.8)

54.7 (13.7)

<0.001 0.011

912 (68) 437 (32) 1 (0)

174 (59) 120 (41) 1 (0)

663 (49) 589 (44) 98 (7) 7.7 (2.0) 7.1 (1.5) 218.5 (58.2) 219.6 (54.3) 35.7 (8.8) 35.2 (8.0)

120 (41) 125 (42) 50 (17) 8.2 (2.1) – 134.3 (69.4) – 37.1 (11.5) –

1020 (76) 330 (24)

191 (65) 103 (35) 1 (0)

123 (9) 244 (18) 446 (33) 495 (37) 42 (3) 0.8 (0.2)

24 (8) 58 (20) 92 (31) 109 (37) 12 (4) 0.3 (0.3)

<0.001

<0.001 <0.001

<0.001

0.822

<0.001

RPM, remote patient monitoring; BMI, body mass index; PAM, patient activation measurement.

Appendix A. Supplementary material Supplementary data to this article can be found online at https://doi.org/10.1016/j.diabres.2019.107944.

[8]

[9] R E F E R E N C E S

[10] [1] Unnikrishnan R, Pradeepa R, Joshi SR, Mohan V. Type 2 diabetes: demystifying the global epidemic. Diabetes Care 2017;66(6):1432–42. [2] Menke A, Casagrande S, Geiss L, Cowie CC. Prevalence of and trends in diabetes among adults in the United States, 1988– 2012 prevalence of and trends in diabetes among US adults prevalence of and trends in diabetes among US Adults. JAMA 2015;314(10):1021–9. https://doi.org/10.1001/jama.2015.10029 %J JAMA. [3] Mauvais-Jarvis F. Gender differences in glucose homeostasis and diabetes. Physiol Behav 2018;187:20–3. [4] Wild S, Roglic G, Green A, Sicree R, King H. Global Prevalence of Diabetes. Estimates for the year 2000 and projections for 2030. J Diabetes Care 2004;27(5):1047–53. https://doi.org/ 10.2337/diacare.27.5.1047 %. [5] Lyon A, Jackson EA, Kalyani RR, Vaidya D, Kim C. Sex-specific differential in risk of diabetes-related macrovascular outcomes. Curr DiabRep 2015;15(11):85. https://doi.org/ 10.1007/s11892-015-0662-x. [6] Garcia M, Mulvagh SL, Bairey Merz CN, Buring JE, Manson JE. Cardiovascular disease in women: clinical perspectives. Circ Res 2016;118(8):1273–93. [7] Huxley R, Barzi F, Woodward M. Excess risk of fatal coronary heart disease associated with diabetes in men and women:

[11]

[12]

[13]

[14]

[15]

[16]

[17]

meta-analysis of 37 prospective cohort studies. BMJ 2006;332 (7533):73–8. Siddiqui MA, Khan MF, Carline TE. Gender differences in living with diabetes mellitus. Materia Socio-Med 2013;25 (2):140. Kautzky-Willer A, Harreiter J, Pacini G. Sex and gender differences in risk, pathophysiology and complications of type 2 diabetes mellitus. Endocr Rev 2016;37(3):278–316. https://doi.org/10.1210/er.2015-1137 %J Endocrine Reviews. The Lancet Diabetes & Endocrinology T. Sex disparities in diabetes: bridging the gap. The lancet Diabetes & endocrinology 2017;5(11):839. Homko CJ, Zamora L, Santamore WP, Kashem A, McConnell T, Bove AA. Gender differences in cardiovascular risk factors and risk perception among individuals with diabetes. Diabetes Educator 2010;36(3):483–8. Legato MJ, Johnson PA, Manson JE. Consideration of sex differences in medicine to improve health care and patient outcomes. JAMA 2016;316(18):1865–6. McGill J, Vlajnic A, Knutsen P, Recklein C, Rimler M, Fisher SJDR, et al. Effect of gender on treatment outcomes in type 2 diabetes mellitus. Diabetes Res Clin Pract 2013;102(3):167–74. Rossi MC, Cristofaro MR, Gentile S, Lucisano G, Manicardi V, Mulas MF, et al. Sex disparities in the quality of diabetes care: biological and cultural factors may play a different role for different outcomes: a cross-sectional observational study from the AMD. Annals Initiat 2013;36(10):3162–8. World Health Organization. Telehealth. https://www.who. int/sustainable-development/health-sector/strategies/ telehealth/en/. Accessed August 29 2019. American Telemedicine Association. What is telemedicine. 2012. http://www.americantelemed.org/learn. Accessed August 30 2016. Michaud TL, Siahpush M, Schwab RJ, Eiland LA, DeVany M, Remote HG, et al. Monitoring and clinical outcomes for

8

[18]

[19]

[20]

[21]

[22] [23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

diabetes research and clinical practice

postdischarge patients with type 2 diabetes. Popul Health Manage 2018;21(5). Riazi H, Larijani B, Langarizadeh M, Shahmoradi L. Managing diabetes mellitus using information technology: a systematic review. J Diab Metab Disorders 2015;14(1):1. Su D, Zhou J, Kelley MS, Michaud TL, Siahpush M, Kim J, et al. Does telemedicine improve treatment outcomes for diabetes? a meta-analysis of results from 55 randomized controlled trials. Diabetes Res Clin Pract 2016;116:136–48. Polisena J, Tran K, Cimon K, Hutton B, McGill S, Palmer K. Home telehealth for diabetes management: a systematic review and meta-analysis. Diabetes Obes Metab 2009;11 (10):913–30. Zhai Y-k, Zhu W-j, Cai Y-l, Sun D-x, Zhao J. Clinical-and costeffectiveness of telemedicine in type 2 diabetes mellitus: a systematic review and meta-analysis. Medicine 2014;93(28). Mair F, Whitten P. Systematic review of studies of patient satisfaction with telemedicine. BMJ 2000;320(7248):1517–20. Foster A, Horspool KA, Edwards L, Thomas CL, Salisbury C, Montgomery AA, et al. Who does not participate in telehealth trials and why? a cross-sectional survey. Trials 2015;16(1):258. Su D, Michaud TL, Estabrooks PA, Schwab RJ, Eiland LA, Hansen G, et al. Diabetes management through remote patient monitoring: the importance of patient activation and engagement with the technology. Telemed e-Health. 2018. Michaud T, Siahpush M, Estabrooks P, Schwab R, LeVan T, Grimm B et al. Association between weight loss and glycemic outcomes: a post-hoc analysis of a remote patient monitoring program for diabetes management. Telemedicine and e-Health. 2019 Aug. StayWell Company, American Diabetes A. Living well with diabetes: a self-care workbook. Yardley, PA: Krames Staywell, LLC; 2018. Kjeldsen SE. Hypertension and cardiovascular risk: general aspects. Pharmacol Res 2018;129:95–9. https://doi.org/ 10.1016/j.phrs.2017.11.003. Hibbard JH, Mahoney ER, Stockard J, Tusler M. Development and testing of a short form of the patient activation measure. Health Serv Res 2005;40(6p1):1918–30. Dixon A, Hibbard J, Tusler M. How do people with different levels of activation self-manage their chronic conditions? Patient: Patient-Centered Outcomes Res 2009;2(4):257–68. Greene J, Hibbard JH. Why does patient activation matter? An examination of the relationships between patient activation and health-related outcomes. J Gen Intern Med 2012;27 (5):520–6. Gorst SL, Armitage CJ, Brownsell S, Hawley MS. Home telehealth uptake and continued use among heart failure and chronic obstructive pulmonary disease patients: a systematic review. Ann Behav Med 2014;48(3):323–36. Broens TH, Huis in’t Veld RM, Vollenbroek-Hutten MM, Hermens HJ, van Halteren AT, Nieuwenhuis LJ, et al. Determinants of successful telemedicine implementations: a literature study. J Telemed Telecare 2007;13(6):303–9. Wade VA, Eliott JA, Hiller JE. Clinician acceptance is the key factor for sustainable telehealth services. Qual Health Res 2014;24(5):682–94. Nilsson PM, Theobald H, Journath G, Fritz T. Gender differences in risk factor control and treatment profile in diabetes: a study in 229 Swedish primary health care centres. Scand J Prim Health Care 2004;22(1):27–31. Go¨bl CS, Brannath W, Bozkurt L, Handisurya A, Anderwald C, Luger A, et al. Sex-specific differences in glycemic control and cardiovascular risk factors in older patients with insulintreated type 2 diabetes mellitus. Gend Med 2010;7(6):593–9.

159 (2020) 107944

[36] Kautzky-Willer A, Kamyar MR, Gerhat D, Handisurya A, Stemer G, Hudson S, et al. Sex-specific differences in metabolic control, cardiovascular risk, and interventions in patients with type 2 diabetes mellitus. Gend Med 2010;7 (6):571–83. [37] Arnetz L, Ekberg NR, Alvarsson M. Sex differences in type 2 diabetes: focus on disease course and outcomes. In: Diabetes, metabolic syndrome and obesity: targets and therapy. p. 409. [38] Ding EL, Song Y, Malik VS, Liu S. Sex differences of endogenous sex hormones and risk of type 2 diabetes A systematic review and meta-analysis. JAMA 2006;295 (11):1288–99. https://doi.org/10.1001/jama.295.11.1288 %J JAMA. [39] Ding EL, Song Y, Manson JE, Rifai N, Buring JE, Liu S. Plasma sex steroid hormones and risk of developing type 2 diabetes in women: a prospective study. Diabetologia 2007;50 (10):2076–84. https://doi.org/10.1007/s00125-007-0785-y. [40] Edwards L, Thomas C, Gregory A, Yardley L, O’Cathain A, Montgomery AA, et al. Are people with chronic diseases interested in using telehealth? a cross-sectional postal survey. J Med Internet Res 2014;16(5) e123. [41] Manteuffel M, Williams S, Chen W, Verbrugge RR, Pittman DG, Steinkellner A. Influence of patient sex and gender on medication use, adherence, and prescribing alignment with guidelines. J Women’s Health. 2014;23(2):112–9. [42] Holge-Hazelton B, Malterud K. Gender in medicine – does it matter? Scand J Public Health 2009;37(2):139–45. https://doi. org/10.1177/1403494808100271. [43] Krag MØ, Hasselbalch L, Siersma V, Nielsen AB, Reventlow S, Malterud K, et al. The impact of gender on the long-term morbidity and mortality of patients with type 2 diabetes receiving structured personal care: a 13 year follow-up study. Diabetologia 2016;59(2):275–85. [44] Franzini L, Ardigo` D, Cavalot F, Miccoli R, Rivellese A, Trovati M, et al. Women show worse control of type 2 diabetes and cardiovascular disease risk factors than men: results from the MIND. IT study group of the italian society of diabetology. Nutri, Metabol Cardiovasc Dis 2013;23(3):235–41. [45] Kautzky-Willer A, Harreiter J. Sex and gender differences in therapy of type 2 diabetes. Diabetes Res Clin Pract 2017;131:230–41. [46] Gutierrez J, Ramirez G, Rundek T, Sacco RL. Statin therapy in the prevention of recurrent cardiovascular events a sexbased meta-analysis. Arch Intern Med 2012;172(12):909–19. https://doi.org/10.1001/archinternmed.2012.2145. [47] Kannel WB, Hjortland MC, McNamara PM, Gordon T. Menopause and risk of cardiovascular disease. Ann Intern Med 1976;85(4):447–52. [48] Schu¨tt M, Zimmermann A, Hood R, Hummel M, Seufert J, Siegel E, et al. Gender-specific effects of treatment with lifestyle, metformin or sulfonylurea on glycemic control and body weight: a german multicenter analysis on 9 108 patients. PPmP Psychotherapie Psychosomatik Medizinische Psychol 2015;65(12):622–6. https://doi.org/10.1055/s-00351559608. [49] Zhang R, Zhao L, Liang L, Xie G, Wu Y. Factors explaining the gender disparity in lipid-lowering treatment goal attainment rate in Chinese patients with statin therapy. Lipids Health Dis 2012;11(1):59. https://doi.org/10.1186/1476-511x-11-59. [50] American Diabetes Association. 9. Pharmacologic Approaches to Glycemic Treatment: Standards of Medical Care in Diabetes- 2019. Diabetes Care. 2019;42(Supplement 1): S90-S102. doi:10.2337/dc19-S009.