Diabetes Management Through Remote Patient Monitoring: The Importance of Patient Activation and Engagement with the Technology

Abstract

Background:

The documented efficacy and promise of telemedicine in diabetes management does not necessarily mean that it can be easily translated into clinical practice. An important barrier concerns patient activation and engagement with telemedicine technology.

Objective:

To assess the importance of patient activation and engagement with remote patient monitoring technology in diabetes management among patients with type 2 diabetes.

Methods:

Ordinary least squares and logistic regression analyses were used to examine how patient activation and engagement with remote patient monitoring technology were related to changes in hemoglobin A1c (HbA1c) for 1,354 patients with type 2 diabetes monitored remotely for 3 months between 2015 and 2017.

Results:

Patients with more frequent and regular participation in remote monitoring had lower HbA1c levels at the end of the program. Compared to patients who uploaded their biometric data every 2 days or less frequently, patients who maintained an average frequency of one upload per day were less likely to have a postmonitoring HbA1c > 9% after adjusting for selected covariates on baseline demographics and health conditions.

Conclusions:

Higher levels of patient activation and engagement with remote patient monitoring technology were associated with better glycemic control outcomes. Developing targeted interventions for different groups of patients to promote their activation and engagement levels would be important to improve the effectiveness of remote patient monitoring in diabetes management.

Introduction

With continuing advancements in digital technology, telemedicine has been increasingly used in the management of diabetes in the United States.¹ The American Telemedicine Association defines telemedicine (or telehealth) as “….the use of medical information exchanged from one site to another via electronic communication to enhance the clinical health status of patients.”² Relative to conventional diabetes management, telemedicine may allow for closer monitoring of blood glucose and improved medication adherence by patients.³ Several studies have reported that telemedicine can improve diabetes management through better metabolic and glycemic control.^4
–6 Meta-analyses of multiple randomized controlled trials suggest that patients with diabetes participating in telehealth programs experience more substantial reductions in hemoglobin A1c (HbA1c) compared to those receiving usual care.^7
–9

The documented efficacy and promise of telemedicine in diabetes management, however, does not necessarily mean that it can be easily translated into clinical practice. Besides reimbursement constraints for telemedicine services,¹⁰ an important barrier includes whether telemedicine can effectively engage patients for the purpose of disease management.

Patient engagement in the context of telemedicine represents “an effective way to foster patients' self-management of their diseases by supporting the use of technological devices that can facilitate self-monitoring skills.”¹¹ According to the Technology Acceptance Model,¹² patients' use of technology is influenced by two key factors as follows: perceived usefulness of the technology and perceived ease of use, which in turn impact patients' attitudes and intention of using the technology. Not all patients are attracted to the idea of resorting to telemedicine for disease management. While some patients might appreciate the convenience of digital health technologies, others might perceive using these technologies as a burden and would avoid them because of indifference, ignorance, technophobia, or simply a preference of physical over virtual encounters with health care providers.¹³ These challenges and concerns may result in reduced patient engagement (e.g., lower frequency and longevity of intervention system use) and consequential attenuated efficacy in diabetes management.

Closely related to patient engagement is the concept of patient activation, which has been defined as “understanding one's role in the care process and having the knowledge, skill, and confidence to manage one's health and health care.”¹⁴ A growing body of literature has documented the positive impact of patient activation on diabetes outcomes.^15
–17 There is also evidence that increases in patient activation may be associated with better diabetes self-management, including regular exercise and maintenance of a blood glucose diary.¹⁸ Despite these findings, little is known about how patient activation might be related to diabetes management through remote patient monitoring.

While patient activation is related to patient engagement, the latter is a broader concept encompassing interventions designed to increase activation and promote positive health behavior.¹⁹ In this study, we used “patient activation” to denote patients' knowledge and confidence in diabetes management and “patient engagement” to indicate the extent to which patient actually participated in diabetes management through remote monitoring. Based on recently collected data from a large cohort of patients with type 2 diabetes who used remote patient monitoring technology for a 3-month period, this study sought to address how patient activation and engagement with remote patient monitoring technology were related to diabetes management outcomes. The study also assessed whether and the extent to which patient activation and patient engagement with remote patient monitoring technology were correlated with each other. Implications of study findings to future design and implementation of remote patient monitoring programs for diabetes management were discussed.

Materials and Methods

Study Setting and Sample

The remote patient monitoring program was implemented at Nebraska Medicine, the top rated hospital in Nebraska.²⁰ Based on informed consent, patients with type 2 diabetes who had a recent hospital admission for any reason were recruited to the program no later than 1 month after hospital discharge. The study was exempt from IRB approval since it was set up as a hospital-based quality improvement program. The 3-month program included daily (7 days a week) remote monitoring of biometric data, including blood pressure, weight, and glucose level, and a minimum of one weekly phone call or instant calls from nurse coaches when an alert was indicated in the monitoring system.

During the remote patient monitoring, nurse coaches provided services, including medication adherence assessment, nutritional counseling, weight measurement, and disease self-management support. A certified diabetes educator provided participants with their HbA1c data and a virtual foot examination at baseline and at the completion of the program. Patients who completed the designated 3-month remote patient monitoring program received diabetic retinopathy screening at one of the three local community health centers. Primary care providers were provided with patient data uploaded throughout the entire intervention period. In total, 1,883 patients were enrolled in the program from 2015 to 2017, out of which 1,354 patients completed 3 months of remote patient monitoring and were included in this study. The leading causes of program attrition included loss of glucose meters, moving to another state, accidents, too busy to participate, and not receiving the information requested based on a qualitative assessment conducted by the study team.

Measurements

Our primary outcome variable is HbA1c, which was measured at both baseline and program completion by nurses in the project team and was divided into two clinical groups—HbA1c > 9% and HbA1c ≤ 9%. Poor glycemic control is usually defined as HbA1c > 9%.²¹

Demographic information included age at baseline, gender, and race (non-Hispanic white versus not White [the vast majority of non-White patients were African American patients]). Baseline body mass index was used to categorize participants into “not obese” (<30 kg/m²) and “obese” categories (≥30 kg/m²). High blood pressure at baseline was defined as >140/90 mmHg.

Patient Activation Measure-13 (PAM-13)²² was used to denote the degree of patient activation at both the baseline and the end of 3 months of remote patient monitoring. PAM-13 is a unidimensional, interval level Guttman-like scale consisting of 13 item questions used to measure patient knowledge, skill, and confidence for self-management of chronic conditions. Subsequent work on this scale further differentiated patients into four levels of activation based on the scoring of their responses: Level 1—the patient does not yet believe they are active or have an important role in managing their health (<47.0); Level 2—the patient lacks confidence and knowledge to take action to manage their health (47.1–55.1); Level 3—the patient is beginning to take action to manage their health (55.2–67.0); and Level 4—the patient is maintaining actions of managing their health over time (≥67.1).²³ In addition, changes in the PAM-13 activation scores were also calculated as the difference in scores between baseline and the completion of the program.

Patient engagement with remote monitoring technology was operationalized as the frequency patients uploaded their biometric data during the program,²⁴ which was calculated using the total number of uploads of biometric data during the whole program period for each patient divided by the total number of days the patient stayed in the program. Because the program completion date was defined as the day on which a patient returned the remote monitoring equipment, the number of intervention days did not always equal 90 days (or 3 months). Quintiles were created to classify the sample into five levels of upload frequency as follows: first quintile, 0–0.56 uploads per day; second quintile, 0.56–0.75 uploads per day; third quintile, 0.75–0.87 uploads per day; fourth quintile, 0.87–0.95 uploads per day; and fifth quintile, 0.95–1.49 uploads per day. Based on these quintiles, a dichotomous variable indicating the level of patient engagement was created with the first 3 quintiles labeled as low engagement and the latter two quintiles as high engagement.

Analysis

In the descriptive statistics, we reported the mean and standard deviation for continuous variables and the frequency and percentage for categorical variables. In the multivariate analysis, we used: (1) ordinary least squares regression to assess the association between HbA1c level at program completion and three focal explanatory variables: baseline patient activation, its changes during the program, and the frequency of biometric data uploading after controlling for the effect of selected demographics and baseline health conditions; (2) logistic regression to examine the association between the odds of having HbA1c > 9% at program completion and the three focal explanatory variables and other covariates on demographics and baseline health conditions; and (3) logistic regression to examine whether there was an association between patient activation and patient engagement after controlling for the effect of selected covariates. All analyses were conducted using Stata^® version 14 (StataCorp, College Station, TX). Two-tailed p-values of less than 0.05 were considered statistically significant.

Results

As indicated in Table 1, the average age of the 1,354 patients who completed the 3-month remote patient monitoring program was 59.6 years. Close to 55% of the patients were female and about two thirds were non-Hispanic white. Out of the 439 minority patients, 365 of them were African American. The average HbA1c at baseline was 7.7% (plus mmol), which fell to 7.1% by the end of the remote patient monitoring program (p < 0.001). The percentage of patients with HbA1c > 9% decreased from 20% at baseline to 10.6% at program completion. Close to three quarters of the patients were obese, and one quarter was hypertensive at baseline.

Table 1.

A Description of the Variables Used in the Analysis Based on Data from 1,354 Patients with Type 2 Diabetes Who Participated in a 3-Month Remote Patient Monitoring Program from 2015 to 2017

VARIABLES	NO. OF CASES	PERCENTAGE	MEAN (SD)
Age at baseline	1,354		59.6 (11.8)
Gender
Male	609	45.1
Female	741	54.9
Race
Non-Hispanic Whites	914	67.5
Minorities (not White)	439	32.5
HbA1c at baseline	1,354		7.7 (2.0)
HbA1c at the end of the program	1,336		7.1 (1.5)
HbA1c > 9 at baseline
No	1,083	80.0
Yes	271	20.0
HbA1c > 9 at the end of the program
No	1,193	89.4
Yes	141	10.6
Obesity status at baseline
Not obese (BMI <30)	342	25.8
Obese (BMI ≥30)	986	74.2
High blood pressure at baseline
No	1,022	75.5
Yes	332	24.5
PAM-13 score at baseline
Level 1 (≤47)	123	9.4	41.9 (6.2)
Level 2 (47.1–55.1)	245	18.7	50.6 (2.2)
Level 3 (55.2–67.0)	447	34.1	60.0 (3.8)
Level 4 (≥67.1)	497	37.9	80.3 (9.7)
Change in PAM-13 score from the baseline to the end of the program	1,256		5.4 (16.4)
Number of uploads per day
First quintile (0–0.56)	270	20.0	0.36 (0.14)
Second quintile (0.56–0.75)	271	20.0	0.66 (0.05)
Third quintile (0.75–0.87)	273	20.2	0.81 (0.04)
Fourth quintile (0.87–0.95)	268	19.8	0.91 (0.02)
Fifth quintile (0.95–1.49)	271	20.0	0.98 (0.04)

BMI, body mass index; HbA1c, hemoglobin A1c; PAM-13, Patient Activation Measure-13; SD, standard deviation.

Baseline patient activation data showed that ∼9%, 19%, 34%, and 38% of participants were at Levels 1, 2, 3, and 4, respectively. The average increase in the PAM-13 score from baseline to the end of the program was 5.4. As for upload frequency, the average number of uploads per day for patients in the first quintile was 0.36 and 0.66, 0.81, 0.91, and 0.98 for the second, third, fourth, and fifth quintiles, respectively (Table 1).

There was a negative correlation between baseline patient activation and postprogram HbA1c (Table 2). Overall, as the patient activation levels increased, their postprogram HbA1c decreased, although the difference in HbA1c between Level 3 and Level 4 activation was small. Relative to patients with Level 1 activation at baseline, the average postprogram HbA1c of patients with Level 4 activation at baseline was 0.32 percentage points lower (p = 0.023), after controlling for the effect of other covariates. Positive changes in patient activation from the baseline to the end of the program were also associated with better HbA1c outcomes. On average, for each point increase in the PAM-13 score, the estimated postmonitoring HbA1c decreased by 0.01 percentage points (p = 0.016), after controlling for the effect of other covariates in the regression.

Table 2.

Predictors of HbA1c at Program Completion Based on Results from Ordinary Least Squares Regression (n = 1,225)

VARIABLES	COEFFICIENT	p -VALUE
Age at baseline	0.01	0.030
HbA1c at baseline	0.39	<0.001
Gender (ref = female)
Male	−0.19	0.007
Race/ethnicity (ref = non-Hispanic whites)
Minorities (not white)	0.20	0.010
Baseline BMI category (ref = not obese)
Obese	0.14	0.076
Having high blood pressure at baseline (ref = no)	0.22	0.008
Baseline patient activation score
Level 1 (≤47)
Level 2 (47.1–55.1)	−0.16	0.259
Level 3 (55.2–67.0)	−0.35	0.008
Level 4 (≥67.1)	−0.32	0.023
PAM score change between baseline and the end of the remote patient monitoring program	−0.01	0.016
Number of uploads per day
First quintile (0–0.56) (ref)
Second quintile (0.56–0.75)	−0.20	0.070
Third quintile (0.75–0.87)	−0.33	0.003
Fourth quintile (0.87–0.95)	−0.40	<0.001
Fifth quintile (0.95–1.49)	−0.40	<0.001

The results also revealed a negative association between frequencies of patient data uploads during the program and postprogram HbA1c. Relative to patients with an average upload of 0.56 times per day or less, patients who were in the third quintile (with an average of uploads between 0.75 and 0.87 times per day) had an estimated average postprogram HbA1c of 0.33 point lower after controlling for other covariates in the regression (p = 0.003). A similar reduction in HbA1c can also be observed as the upload frequency increased to the fourth and fifth quintiles, although there was virtually no difference between these two quintiles.

Older patients were more likely to have a higher HbA1c at the end of the program. For each year of age increase at baseline, the postprogram HbA1c increased by 0.01 (p = 0.03). Not surprisingly, one of the most significant predictors of postprogram HbA1c was baseline HbA1c. The results also revealed significant gender and racial disparities in postprogram HbA1c. The average HbA1c for male patients was 0.19 points lower than that for female patients (p = 0.007). Ethnic/racial minority patients had an average postprogram HbA1c of 0.2 point higher than non-Hispanic White patients (p = 0.01). Having hypertension at baseline was associated with higher HbA1c at the end of the program, whereas being obese at baseline was not significantly associated with HbA1c levels at program completion.

In terms of the odds of having a postprogram HbA1c greater than 9%, the findings were similar to what we found with the ordinary least squares regression regarding the effect of baseline age, HbA1c, race, and gender, as indicated in Table 3. However, there were also several notable differences. Being hypertensive at baseline was not significantly associated with the odds of having a postprogram HbA1c greater than 9%. Similarly, patient activation at baseline was not significantly predictive of the odds of having a postprogram HbA1c greater than 9%. However, positive changes to baseline PAM score were associated with lower odds of having a postprogram HbA1c greater than 9% (odds ratio [OR] = 0.98, 95% confidence interval [CI: 0.97–1.00], p = 0.031). The association between upload frequency and having HbA1c greater than 9% was less consistent than the corresponding findings in the ordinary least squares regression: being in the third (OR = 0.48, 95% CI [0.25–0.90], p = 0.022) or fifth upload quintiles (OR = 0.31, 95% CI [0.14–0.67], p = 0.002) was associated with lower odds of having a postprogram HbA1c greater than 9% relative to being in the first upload quintile.

Table 3.

Predictors of HbA1c > 9 at Program Completion Based on Results from Logistic Regression (n = 1,224)

VARIABLES	OR	95% CI	p -VALUE
Age at baseline	1.01	0.99–1.03	0.162
HbA1c at baseline	1.69	1.53–1.86	<0.001
Gender (ref = female)
Male	0.52	0.34–0.81	0.004
Race/ethnicity (ref = non-Hispanic Whites)
Minorities (not white)	1.61	1.06–2.47	0.027
Baseline obesity status (ref = not obese)
Obese	1.40	0.85–2.32	0.189
Having high blood pressure at baseline (ref = no)
Yes	1.18	0.74–1.87	0.637
Baseline patient activation score
Level 1 (≤47) (ref)
Level 2 (47.1–55.1)	0.68	0.32–1.45	0.316
Level 3 (55.2–67.0)	0.49	0.24–1.01	0.054
Level 4 (≥67.1)	0.70	0.32–1.51	0.356
PAM score change from baseline to the end of the program	0.98	0.97–1.00	0.031
Number of uploads per day
First quintile (0–0.56) (ref)
Second quintile (0.56–0.75)	0.68	0.38–1.21	0.186
Third quintile (0.75–0.87)	0.48	0.25–0.90	0.022
Fourth quintile (0.87–0.95)	0.64	0.35–1.20	0.163
Fifth quintile (0.95–1.49)	0.31	0.14–0.67	0.002

CI, confidence interval; OR, odds ratio.

Older patients were more likely to have a high level of participation in remote monitoring than younger patients (Table 4). For each additional year of age at the baseline, the odds of having a high level of participation increased by 3% (OR = 1.03, 95% CI [1.02–1.04]) after controlling for the effect of other covariates in the regression. Patients with higher baseline HbA1c were less likely to have a high level of engagement. On average, each additional point of baseline HbA1c was associated with a 9% reduction in the odds of having a high level of participation (OR = 0.91, 95% CI [0.85–0.97]). Male patients were more likely to have a high level of participation than female patients (OR = 1.35, 95% CI [1.06–1.71]). Neither baseline patient activation level nor changes in PAM score were significantly associated with the odds of having a high level of participation.

Table 4.

Correlates of High Level of Patient Participation Based on Logistic Regression (n = 1,226)

VARIABLES	OR	95% CI	p -VALUE
Age at baseline	1.03	1.02–1.04	<0.001
HbA1c at baseline	0.91	0.85–0.97	0.005
Gender (ref = female)
Male	1.35	1.06–1.71	0.016
Race/ethnicity (ref = non-Hispanic Whites)
Minorities (not white)	0.95	0.73–1.23	0.681
Baseline obesity status (ref = not obese)
Obese	0.89	0.68–1.17	0.394
Having high blood pressure at baseline (ref = no)
Yes	1.01	0.77–1.33	0.950
Baseline patient activation score
Level 1 (≤47) (ref)
Level 2 (47.1–55.1)	0.84	0.52–1.35	0.459
Level 3 (55.2–67.0)	0.89	0.57–1.40	0.890
Level 4 (≥67.1)	1.24	0.77–1.99	0.373
Change in PAM-13 score from baseline to the end of the program	1.01	1.00–1.02	0.078

High level of patient participation was defined as being in the fourth or fifth quintiles of upload frequency.

Discussion

The findings from this study highlight the importance of addressing and improving patient activation and engagement with remote patient monitoring technology before the full potential of this monitoring approach can be realized. In particular, we found that even among patients with type 2 diabetes who had consented to participate in remote patient monitoring, they differed from each other in terms of their activation and participation in the remote monitoring, both of which were associated with diabetes outcomes at the end of the intervention. Patients with higher levels of activation at baseline had lower HbA1c levels at the end of the program compared to their counterparts with lower levels of activation. Moreover, patients whose activation improved during the program also had better glycemic control at the end of the program. These effects remained statistically significant after controlling for demographics and other baseline health conditions. Consistent with previous studies documenting the importance of patient activation in diabetes management under conventional care,^15
–17 findings from our study confirmed the relevance of patient activation in diabetes management in the setting of remote patient monitoring.

Results of our study also point to the importance of patient engagement with the remote monitoring technology. Adjusted estimates indicated that patients who uploaded their data more frequently had better HbA1c control than their counterparts who uploaded less frequently. In particular, this effect was most distinctive when comparing patients who on average maintained daily uploading and those who had less frequent uploading. Consistent with these findings, one previous study investigated the effect of patient engagement in self-monitoring with a telemonitoring device on glycemic control among patients with type 2 diabetes and reported that the proportion of patients achieving target HbA1c at 6 months was significantly higher among frequent users than among infrequent users of the telemonitoring device.²⁵ It was also found that an increased frequency of self-monitoring of blood glucose was significantly correlated with a reduction in HbA1c at 6 months. Findings from our study contribute to this body of work by assessing the importance of patient engagement in a much larger sample, as well as by identifying correlates of patient engagement as measured by upload frequency.

The lack of correlation between patient activation and actual program participation, as indicated by findings from our study, warrants further research and points to the potential need of addressing them as independent factors contributing to glycemic control outcomes when using remote patient monitoring technology. This can be accomplished through patient education and coaching both at baseline and throughout the remote patient monitoring period. As indicated by our study, it is possible to increase patient activation through remote patient monitoring and coaching within 3 months. Previous studies documented how increased patient activation can lead to better diabetes management or positive changes in health behaviors.^14,26

To improve patient participation in remote monitoring, our study suggests that patients who are younger, have a higher baseline HbA1c, and are female might need more attention and support. Relative to older patients, especially those who have retired, younger patients might have less time, which would allow them to maintain active participation in remote monitoring. Moreover, patients with higher baseline HbA1c could be frailer or having more comorbidities than those with lower baseline HbA1c, which might either hinder them from active participation in the program or compel them to seek more immediate or potent care such as visiting a doctor or emergency room. Our finding that male patients on average had better glycemic control than female patients is consistent with findings from previous studies.^27,28 Male patients were also more engaged in remote patient monitoring based on upload frequency. These disparities warrant further research before specific interventions can be developed and implemented to decrease them.

Several limitations of this study merit comments. First, our study included limited sociodemographic information, because we did not collect data on marital status, education, or social support, which may have impacted our outcome findings.²⁹ Second, the vast majority of the racial/ethnic minority patients in the study were African American, and thus, the results may not generalize to other racial/ethnic minority groups. Finally, the targeted intervention period for this study was 3 months, which means our study findings are more relevant for assessing the short-term effects of patient activation and engagement on glycemic control through remote patient monitoring. Despite these limitations, this study represents a rare effort in examining and establishing the importance of patient activation and engagement in glycemic control through short-term remote patient monitoring, and our findings can inform the design and implementation of similar programs in the future.

While the two terms “patient activation” and “patient engagement” are often used interchangeably,¹⁴ they can mean different aspects in diabetes management through remote patient monitoring. Having the knowledge, skills, and confidence in managing diabetes, as captured by patient activation, does not necessarily translate into active engagement and participation in remote patient monitoring. Identifying determinants of patient activation and engagement and developing targeted interventions for different groups of patients to promote activation and engagement levels will be important for improving the effectiveness of remote patient monitoring in the management of diabetes.

Footnotes

Acknowledgments

The authors have no financial or other conflict of interest to declare. The remote patient monitoring program as described in this study was supported by Grant Number 1C1CMS331344 from the 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 article. 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 in this study 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. The authors thank Grace Cai who provided editorial assistance to the article.

Disclosure Statement

No competing financial interests exist.

References

Klonoff

. Telemedicine for diabetes: Current and future trends. J Diabetes Sci Technol, 2016; 10:3–5.

American Telemedicine Association. What is telemedicine. Available at http://www.americantelemed.org/main/about/telehealth-faqs- (last accessed March 15, 2017 ).

Kleinman

, Shah

, Phatak

, Viswanathan

. Improved medication adherence and frequency of blood glucose self-testing using an m-health platform versus usual care in a multisite randomized clinical trial among people with type 2 diabetes in India. Telemed J E Health, 2017; 23:733–740.

Riazi

, Larijani

, Langarizadeh

, Shahmoradi

. Managing diabetes mellitus using information technology: A systematic review. J Diabetes Metab Disord, 2015; 14:49.

Holtz

, Lauckner

. Diabetes management via mobile phones: A systematic review. Telemed J E Health, 2012; 18:175–184.

Paré

, Moqadem

, Pineau

, St-Hilaire

. Clinical effects of home telemonitoring in the context of diabetes, asthma, heart failure and hypertension: A systematic review. J Med Internet Res, 2010; 12:e21.

Polisena

, Coyle

, McGill

. Home telehealth for chronic disease management: A systematic review and an analysis of economic evaluations. Int J Technol Assess Health Care, 2009; 25:339–349.

, Zhou

, Kelley

, et al. Does telemedicine improve treatment outcomes for diabetes? A meta-analysis of results from 55 randomized controlled trials. Diabetes Res Clin Pract, 2016; 116:136–148.

Marcolino

, Maia

, Alkmim

MBM

, Boersma

, Ribeiro

. Telemedicine application in the care of diabetes patients: Systematic review and meta-analysis. PLoS One, 2013; 8:e79246.

10.

Weinstein

, Lopez

, Joseph

, et al. Telemedicine, telehealth, and mobile health applications that work: Opportunities and barriers. Am J Med, 2014; 127:183–187.

11.

Barello

, Graffigna

, Vegni

, Bosio

. The challenges of conceptualizing patient engagement in healthcare: A lexicographic literature review. J Particip Med, 2014; 6:e9.

12.

Davis

. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q, 1989; 13:319–339.

13.

Lupton

. The digitally engaged patient: Self-monitoring and self-care in the digital health era. Soc Theory Health, 2013; 11:256–270.

14.

Hibbard

, Greene

. What the evidence shows about patient activation: Better health outcomes and care experiences; fewer data on costs. Health Aff (Millwood), 2013; 32:207–214.

15.

Remmers

, Hibbard

, Mosen

, Wagenfield

, Hoye

, Jones

. Is patient activation associated with future health outcomes and healthcare utilization among patients with diabetes?. J Ambul Care Manage, 2009; 32:320–327.

16.

Rask

, Ziemer

, Kohler

, Hawley

, Arinde

, Barnes

. Patient activation is associated with healthy behaviors and ease in managing diabetes in an indigent population. Diabetes Educ, 2009; 35:622–630.

17.

Begum

, Donald

, Ozolins

, Dower

. Hospital admissions, emergency department utilization and patient activation for self-management among people with diabetes. Diabetes Res Clin Pract, 2011; 93:260–267.

18.

Hibbard

, Mahoney

, Stock

, Tusler

. Do increases in patient activation result in improved self-management behaviors?. Health Serv Res, 2007; 42:1443–1463.

19.

James

Health policy brief: patient engagement. Health Affairs. 2013. Available at http://www.healthaffairs.org/healthpolicybriefs/brief.php?brief_id=86. (last accessed October 22, 2018 ).

20.

U.S. News Rankings. Best Regional Hospitals. 2017. Available at https://health.usnews.com/best-hospitals/area/ne (last accessed May 8, 2018 ).

21.

Centers for Medicare and Medicaid Services. Medicare Shared Savings Program Quality Measure Benchmarks for the. 2015. Reporting Year. Available at https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/Downloads/MSSP-QM-Benchmarks-2015.pdf (last accessed October 22, 2018 ).

22.

Hibbard

, Mahoney

, Stockard

, Tusler

. Development and testing of a short form of the patient activation measure. Health Serv Res, 2005; 40(6p1):1918–1930.

23.

Dixon

, Hibbard

, Tusler

. How do people with different levels of activation self-manage their chronic conditions?. Patient, 2009; 2:257–268.

24.

Agboola

, Havasy

, Myint-U

, Kvedar

, Jethwani

. The impact of using mobile-enabled devices on patient engagement in remote monitoring programs. J Diabetes Sci Technol, 2013; 7:623–629.

25.

Lee

M-K

, Lee

K-H

, Yoo

S-H

, Park

C-Y

. Impact of initial active engagement in self-monitoring with a telemonitoring device on glycemic control among patients with type 2 diabetes. Sci Rep, 2017; 7:3866.

26.

Harvey

, Fowles

, Xi

, Terry

. When activation changes, what else changes? The relationship between change in patient activation measure (PAM) and employees' health status and health behaviors. Patient Educ Couns, 2012; 88:338–343.

27.

Kautzky-Willer

, Kamyar

, Gerhat

, et al. Sex-specific differences in metabolic control, cardiovascular risk, and interventions in patients with type 2 diabetes mellitus. Gend Med, 2010; 7:571–583.

28.

Göbl

, Brannath

, Bozkurt

, et al. Sex-specific differences in glycemic control and cardiovascular risk factors in older patients with insulin-treated type 2 diabetes mellitus. Gend Med, 2010; 7:593–599.

29.

Barnett

, Chumbler

, Vogel

, Beyth

, Qin

, Kobb

. The effectiveness of a care coordination home telehealth program for veterans with diabetes mellitus: A 2-year follow-up. Am J Manag Care, 2006; 12:467–474.