Abstract
The mobile Digital Access to a Web-enhanced Network (mDAWN) program was implemented as an online, mobile self-management system to support patients with type-2 diabetes and their informal caregivers. Patients used wireless physiological sensors, received text messages, and had access to a secure web platform with health resources and semi-facilitated discussion forum. Outcomes were evaluated using (1) pre and post self-reported health behavior measures, (2) physiological outcomes, (3) program cost, and (4) in-depth participant interviews. The group had significantly decreased health distress, HbA1c levels, and systolic blood pressure. Participants largely saw the mDAWN as providing good value for the costs involved and found the program to be empowering in gaining control over their diabetes. mHealth programs have the potential to improve clinical outcomes through cost effective patient-led care for chronic illness. Further evaluation needs to examine integration of similar mHealth programs into the patient-physician relationship.
Introduction
Diabetes mellitus is a chronic illness which affects over 2.7 million Canadians1. 90% of those affected have Type-2 diabetes, a metabolic disorder characterized by the body’s inability to produce and/or use insulin effectively. In 2010, diabetes cost the Canadian health systems $11.7 billion. This number is expected to rise to $16 billion by 20201. The majority of these costs are attributed to health care costs (including those for complications such as cardiovascular disease, limb amputations and kidney disease), disability and work loss2, some of which could be delayed or avoided through improved diabetes management1
Though management can take a number of forms, a growing body of evidence suggests that technology-based outreach increases patient health literacy and access to quality care and there is specific evidence to demonstrate that self-management and monitoring can improve clinical outcomes as well as quality of life in patients with diabetes3–5. The extension from generalized technology to mobile technology is logical as almost 50% of Canadians are using mobile devices to access the internet, and an increasing amount of time is being spent on these devices as compared to traditional technologies such as computers or laptops6. Mobile technologies offer a number of benefits including portability and improved accessibility, and are a particularly interesting platform for health related uses due to their low cost relative to other health technologies and due to the availability of health related applications.
Mobile Health (mHealth) research has begun to document evidence of its potential for chronic disease and diabetes management7,8 and current literature demonstrates positive findings with respect to the use of mHealth tools in a variety of health contexts 9–12 however a recent environmental scan of fully implemented mHealth programs in Canada that use remote tracking and analysis of clinical data for patients’ management of diabetes identified a scarcity of fully implemented or sustained programs. We therefore conducted a feasibility study, mobile Digital Access to a Web-enhanced Network (mDAWN), to examine the use of mobile technologies (wireless monitoring sensors, social media, and text messaging) in an attempt to address this gap and to better understand the potential of mHealth programs to support both individuals with type-2 diabetes and their caregivers.
Methods
mDAWN was conducted to determine the feasibility and potential benefits of using mHealth technologies to support patient with diabetes patients and their caregivers. The program included three different mHealth “tools”; SMS, an online community, as well as wireless and manual monitoring devices and participants were enrolled for a 3 month period. mDAWN used an iterative approach to examine the use of mHealth technologies from a patient/caregiver perspective, conducting a pilot study (December 1, 2013 to February 28, 2014) to enrolled a group of patients and caregivers to trial the system and refine our protocol and monitoring strategy and a full intervention with program evaluation from May 1st, 2014 to July 31st, 2014.
Participants were recruited from the general public using convenience sampling. The pilot study engaged 24 participants: 13 patients and 11 caregivers (family members or friends identified as health supporters). During the study, two dyads dropped out due to injury and a lack of interest due to technology delivery delays. The second study group was comprised of 43 participants, 26 patients and 17 caregivers.
In order to be eligible to participate in the mDAWN program, participants needed to be able to receive text messages and access the internet. They also needed to speak English and be at least 18 years of age. Patient participants were offered an honorarium of $100 and caregivers were offered an honorarium of $75. Initially, it was required that patients select a caregiver to participate with them in the study; however, this requirement was removed as potential participants reported that they did not necessarily view the presence of a caregiver as important to their type-2 diabetes management. All participants completed an informed consent process.
Participants took part in a 3-month study during which they received access to a website with health resources and a semi-facilitated discussion forum, bi-weekly text messages with health tips and challenges, and three monitoring devices including a wireless blood pressure monitor which measured systolic and diastolic blood pressure as well as heart rate, a wireless weight scale, and a manual blood glucose meter.
The blood pressure monitor and weight scale were purchased from blipcare (www.blipcare.com), a subsidiary of the Carematix company, as they fulfilled the study target of consumer-marketed technology with wireless connectivity capabilities. Measurements from these devices were automatically uploaded to a secure website, accessible to patients and their caregivers, when taken and could be viewed in a number of different graphical and report formats. The blood glucose meter was standard and patients were asked to measure their blood glucose twice daily and enter these results into an electronic tracking form designed by the study team. Patients and caregivers also had access to a web-portal with health resources and a semi-facilitated discussion board. Access to this webportal was limited to participants and study team members and discussion were started both by study team members and by participants. In addition to this, participants received text messages twice a week over the three month intervention period; the first message was designed to be informative, including a health tip and a link to a matching resource on the website and the following text provided a challenge or a question on the same topic, encouraging participants to interact with other study participants using the web-portal.
During the three month study period, evidence of the mDAWN’s impact was gathered from four sources:
Self-reported health measures
Physiological outcomes in weight, blood pressure, heart rate, and blood glucose
Program Cost and Feasibility
In-depth post-study interviews with participants
Self-reported health measures
Patients and caregivers were administered an online survey at the beginning and again at the end of the three month interventions. Nine different measures, selected by a literature scan of similar programs and through review by the clinicians on the research team, were administered to participants. We chose a large number of measures to scrutinize which ones could best reflect the impact of mHealth activities. A statistical and theoretical analysis led to the selection of the following measures for pre-post comparison:
Health Distress: measures how much distress a person experiences directly related to their health in the past month. It is more specific to distress stemming from health issues rather than a general measure of anxiety or depression. The lower the score, the less distressed a person is about their health.
eHealth Literacy (eHEALS): measures the ability to read, understand and communicate about health information in regards to internet resources to make informed health decisions. High eHealth literacy is indicated by a high score on this measure.
Diabetes Empowerment – Short Form (DES-SF): measures psychosocial self-efficacy of people with diabetes, dissatisfaction, readiness to change, and goal attainment. The higher score, the better the patients’ sense of empowerment when it comes to managing their diabetes.
Quality Adjusted Life Year (QALY) Index: a measure of disease burden that assesses the quality and quantity of life lived in one year ranging from 0 to 1.0, with 1.0 being a full quality life year lived. It is typically used in assessing the monetary value of a medical intervention. The Euroqol 5D-5L (EQ-5D-5L) and United Kingdom look-up table was used to obtain the QALY values for this study.
Two measures were not included beyond a preliminary analysis, the Patient Activation Measure (PAM) and Patient Health Questionnaire (PHQ-9) due to high correlations which indicated overlap with other measures. The PAM is a broad measure of confidence with health management and is not diabetes-specific. The PHQ-9 measures depression, but not in relation to a person’s current view of their health otherwise. Social support scales were excluded as patients did not identify with being in need of a caregiver. (Measures of social support that were considered: Multidimensional Scale of Perceived Social Support (MSPSS), Zarit Burden Interview, and Kingston Caregiver Scale).
Physiological outcomes in weight, blood pressure, heart rate, and blood glucose
Data was analyzed from patients’ weight, systolic blood pressure, and blood glucose measurements. Additionally, patients visited their physicians before and after the study to provide a measure of HbA1c. Weight and systolic blood pressure data were automatically uploaded to the Carematix website when a patient took a measurement. The Carematix website had a patient facing site where the participants could securely sign-on and review their self-monitoring data as well as an administrative side where the research team could view data from all participants.
Participants were asked to obtain their glycated hemoglobin (HbA1c) from their physician prior to the mDAWN study and at the end of the three month program. Blood glucose measurements were collected by patients manually and then emailed to the research team. Patients were instructed to measure their blood glucose twice a day at alternating times every other day. Specifically, patients would measure when they woke up in the morning and before they went to bed on one day and then the following day they would measure before lunch and then again before dinner. This resulted in an aggregate view of blood glucose at four times a day: (1) at wake-up (morning), (2) before lunch time (midday), (3) before dinner (evening), (4) before bed (late evening).
Program Cost and Feasibility
An environmental scan was conducted to identify research or clinical programs similar to mDAWN for comparison in implementation and costs. The search was limited to programs in Canada that use technology to improve health outcomes for people with diabetes by enabling remote tracking and analysis of clinical data.
This scan focused on three information areas: peer-reviewed literature (from 2008 to 2014), grey literature and key informant outreach. Grey literature included a review of national and provincial health organization publications, press releases, internal reports and other online documentation. Sixty inquiries were made of key informants around organizational use of mHealth programs for diabetes management.
Patient participants were also given a cost-feasibility survey at the end of their post survey that asked about annual household income, what devices the participants’ were most likely to pay for, how much they would pay to use mHealth devices to manage their type-2 diabetes and the perceived value of an mDAWN-like program.
In-depth post-study interviews with participants
Patient and caregiver participants were invited to take part in a structured interview about their overall program experience with mHealth and how it could fit into their interactions with the health care system as well as specific experiences with the program content (resources, text messages, discussion board, and monitoring devices).
Results
Self-Reported Health Measures
Two pre and post-test comparisons were done, one for patients only, and one for patients / caregiver dyads. Patients’ pre and post test scores were compared on the selected self-report health measures, HbA1c, health distress, diabetes empowerment (DES-SF), and eHealth literacy (eHEALS). Dyads were analyzed for comparisons between patients and caregivers as well as pre and post-test measures on health distress, eHealth literacy (eHEALS) and QALY.
A one-way within groups multivariate analysis of variance (MANOVA) was conducted to assess patient pre and post - test differences on HbA1c, health distress, diabetes empowerment, and eHealth literacy. There was a statistically significant difference between pre and post testing, Wilks’ Lambda = .391, F (4, 11) = 4.28, p = .025. The effect size for the overall model using partial eta squared was .61 indicating a very large effect. Each dependent variable in the model had a partial eta squared of over .138, which is a generally accepted criteria (Cohen, 1988) for a large effect. HbA1c, health distress and diabetes empowerment scores all saw changes in a positive direction, with HbA1c and health distress lowering and diabetes empowerment increasing. HbA1c decrease is not only statistically significant, but also clinically significant in dropping from 7.41 to 6.77 (see Appendix A, table 1), with 7 as the clinically acceptable upper limit of normal. eHealth literacy decreased from pre to post test. See Table 1 in Appendix A for the means, standard deviations, and partial eta squared of each variable.
A mixed within and between groups MANOVA was conducted to assess differences between caregivers and patients on pre and post-tests of health distress, eHealth literacy, and quality-adjusted life year (QALY). There was a statistically significant interaction between patients and caregivers across pre and post-tests, Wilks’ Lambda = .533, F (3, 12) = 3.5, p < .05. The partial eta squared was .467 indicating a large effect. Patients reported a significant decrease in health distress, whereas caregivers reported an increase with a large effect size of .405 (partial eta squared). Similarly, QALY scores increased for patients and decreased for caregivers with a moderate effect size of .079. Both patient and caregiver eHealth literacy scores decreased, but with no effect size to report (partial eta squared = 0). See Table 2 in Appendix A for patient and caregiver means and standard deviations.
Physiological Outcomes
Weight and systolic blood pressure were each analyzed with a one-way repeated measures analysis of variance (ANOVA) and blood glucose was analyzed using a two-way repeated measures ANOVA to examine measurement time and changes across the three months of the intervention.
There was no significant decrease in the group’s average weight (Wilk’s Lambda = .859, F (2, 14) = 1.15, p = .346), though the group lost an average of 3.51 pounds. Additionally there was a significant negative correlation between number of weigh-ins a participant completed and their weight difference by the end of the study (r = −.69, p < .01) indicating that the more a participant weighed themselves, the more weight they could be expected to lose.
There was a significant decline in systolic blood pressure across the three months, Wilk’s Lambda = .49, F (2, 11) = 5.7, p < .02. A multivariate partial eta squared .51 was found, indicating a large effect. Further examination using a trend analysis showed that the significant drop occurred from month 1 to month 2 with systolic readings climbing back up again slightly in the third month (quadratic trend F (1, 12) = 7.79, p < .016, partial eta squared = .394).
Blood glucose data was analyzed according to the four measurement times (morning, midday, evening, and late evening) across the three mDAWN months to elucidate if was there a difference over the three months in each of the four measurements. Data was analyzed using a two-way repeated measures ANOVA. There was no interaction effect for time and month (Wilk’s Lambda = .80 F (6, 60) = 1.15, p = .35).
Program Cost and Feasibility
The environmental scan of diabetes-focused mHealth programs revealed that there are very few published evaluations of remote monitoring programs- while a number of pilot programs featured remote monitoring, none appeared to have a mobile component or reported outcomes beyond the pilot stage. Outreach to key informants supported this finding, with most stating that they were unaware of any mobile health programs for diabetes being used within their organizations. Many were familiar with mobile apps for diabetes management, and multiple respondents expressed interest in learning more about the outcomes of mDAWN, however, no baseline was found to provide a cost comparison point for the mDAWN program.
Patient participants in the second study group also completed a cost survey at the end of the program. In this survey, participants were provided with information about the variable costs of the program and asked questions designed to help the study team understand the perceived value of a program such as mDAWN. 24 participants answered the cost survey, though not all respondents answered all questions. Key findings are illustrated in Table 4
Table 4.
Cost Survey Results
| Household Income | # of responses |
|---|---|
| 0 to $20,000 | 1 |
| $20,000 to $40,000 | 4 |
| $40,000 to $60,000 | 7 |
| $60,000 to $80,000 | 3 |
| More than $80,000 | 6 |
| The total cost per person of the mDAWN study is $25 per month for the monitoring service, $35 a month for blood glucose test strips, and $365 for the monitoring devices (weight scale, blood pressure monitor, and blood glucose meter). If you were to purchase an mDAWN-like system yourself, you would spend $365 at start-up and then $60 a month after that. Please choose the statement below that best describes how the program cost suits you. | # of responses |
| I am willing to pay for this, but I can’t afford it. | 6 |
| I can’t afford it and I wouldn’t want to pay for it if I could. | 4 |
| I am willing to pay for this, but it would be a financial burden. | 3 |
| I would only want to pay for parts of this (for example, just the weight scale). | 5 |
| I can afford this and I would pay for it. | 4 |
| I can afford this and it would save me money in other ways. How would it save you money? | 0 |
| If the British Columbia Medical Services Plan (MSP) covered some of the cost for home monitoring systems, would you take part and what would be the most you could pay? | # of responses |
| Yes, but the maximum I could pay is $0 | 10 |
| Yes, but the maximum I could pay is $40 – $60 a month | 6 |
| Yes, but the maximum I could pay is $60–$80 a month | 1 |
| Yes, but the maximum I could pay is $80–$100 a month | 2 |
| Wouldn’t participate even if costs were covered. | 2 |
In-Depth Post Study Interviews with Participants
All participants indicated an overall positive experience with the mDAWN program, finding their participation beneficial. Common themes that emerged from the participant interviews were the creation of self-care routines and habits, increased self-awareness, and participant empowerment. Participating in the mDAWN program increased awareness of diabetes and of the self, related to diabetes. Particularly, participants expressed an increased understanding of how their lifestyle can affect their glucose control, weight or blood pressure measurements. More importantly, through their experiences in the program participants were able to connect and apply the information provided to their own lives. In turn, making these connections allowed participants to effect positive changes in their lifestyle including exercise and eating habits and to achieve goals such as reducing HbA1c levels, weight or blood pressure measurements.
Participants’ experience in the program also increased their sense of empowerment. Throughout the program participants gained an appreciation that they were ultimately responsible in caring for their diabetes. Participants felt that the program provided the right amount of accountability, incentive, and support to allow them to achieve their goals. Some participants indicated that participating in the program increased their feelings of control and level of confidence in their ability to care for themselves. Being generally participant-led, the mDAWN program gave participants the tools and knowledge while still allowing considerable freedom. This allowed participants to take ownership over the positive changes they achieved while participating in the program.
This new sense of self-confidence carried over into their expectations of future interactions with their doctor, as patients noted that they were now able have a more knowledgeable conversation and were more confident in knowing which questions to ask. In essence, participants felt they were now more “activated” patients who would be more involved in making decisions about their care with their doctors.
Discussion
The mDAWN patient group was able to achieve lowered health distress, HbA1c, and systolic blood pressure as well as an increase in diabetes empowerment over the course of the 3 month study. During interviews, the group reported that they were able to achieve their health goals through a bolstered sense of empowerment. Additional findings in our study based on pre- and post- study self-report health measures are worthy of further consideration: the decrease in both patients’ and caregivers’ perceived eHealth literacy and the rise in caregiver distress. The study team can hypothesize that the decrease in eHealth literacy may be a case of participants ‘not knowing what they don’t know’ at the beginning of the study - feeling that they had a good handle on how technologies have then discovered new digital ways to improve health through the mDAWN program. This “broadening of horizons” could have affected their perception of their own eHealth literacy. The rise in caregiver stress found in the study might be caused by the more constant linkage of them with their patients with potential 24/7 connectivity, thereby triggering a sense of rising responsibility. Recent studies and news reports indicate that caregiver burden can be a source of negative health outcomes, and our current findings may suggest a more nuanced view: that taking on or imposing the role of caregiver may be a factor in negatively impacting health for the caregivers themselves. Further research should be conducted in order better understand these results.
In terms of cost, participants largely identified the mDAWN program as providing good value for money and indicated that hypothetically, they would be willing to pay the fees associated with a program like this. While willingness exists, more than half of the participants (56%) also indicated that based on their current income they would be either unable to afford the program, or that they would find it to be a significant financial burden. Questions regarding cost sharing between the patients and the health system funder, in our case the provincial medical program, found that 52% of participants interested in continuing the program would only find it feasible to do so if the cost was zero. It should further be noted here that over half of mDAWN participants had a household income of less than $60,000/year (below that of the national average). This suggests that in order for a program like this to be successful and feasible for all participants, cost-sharing with the health-system or subsidies must be considered. Cost considerations in future studies should include considering how to build a cost model comparison point which is flexible to changes in technology development that impact accessibility and price, as well as exploring strategies which support sustainable and affordable programs.
Results from the mDAWN feasibility study suggest that the mDAWN system, in the context of type-2 diabetes self-management, can improve clinical outcomes and overall wellbeing. Key aspects of the program’s success include the synergistic use of physiologic sensors and social media. This combination of monitoring devices and a secure social media platform empowered participants by providing them with a unique combination of autonomy and connectedness. The monitoring devices allowed patients to quantify their own health data and apply it recognize patterns in their health and understand how lifestyle choices were affecting their wellbeing. Sharing insights on the social media platform provided opportunities to learn from others, share and celebrate success, and created a sense of accountability to a community. This unique combination of autonomy and connection to a peer group was highly valued by participants and identified as a key part of increasing patient empowerment.
This finding supports the validity of the behavioral change model put forward by Bandura (1986) which asserts that that three key sequential steps are needed to affect behavior change, self-monitoring, self-evaluation, and behavior modification, and that these steps work in synchrony and in a continuous cycle of positive reinforcement13. Our findings in this study may be applicable to not only diabetes self-management, but also chronic disease self-management in general.
Limitations
The mDAWN program explored the feasibility of a mHealth based program for type-2 diabetes. A convenience sample was used which resulted in a study group with high pre-study levels of motivation, eHealth literacy, and activation and relatively low levels of health distress and perceived burden. While the study has addressed the key research questions, the findings of this study cannot at this point be generalizable to the general population. Further research needs to be conducted to fully understand:
What value can a mHealth program provide to patients with low levels of motivation or health -engagement? Recruitment for mDAWN attracted participants who were interested in improving their health. While mDAWN supported improved outcomes for these individuals, we are unable to say how this program might work with a less-motivated patient group.
Can those who might not be as familiar with mobile health technologies still find this system useful for them? For those who use social media relatively less than the mDAWN study group, is the social media component of this program of value to help them connect with their peers?
How can health professionals add value to this monitoring system? Our study did not include health professional participation and focused on the value of mhealth tools when used independently by patients. This is in contrast to many conventional remote patient monitoring studies which commonly put patients in a passive role of being managed by health professionals. While both strategies have benefit, there is a gap in knowledge regarding both the value that mhealth tools might offer to the health professional-patient partnership model, as well as a lack of information regarding what would be required to fit this sort of system into existing workflows. Further research is necessary to understand how health professionals can best support patients in the use of mHealth tools as part of an integrated self-management plan.
The next step in this evaluation would be to conduct a clinical trial to address the questions above, and also to allow for the quantification of the benefits seen in this feasibility study.
Appendix A
Table 1.
Statistics for patient self-reported health measures MANOVA
| Pre Mean | Pre SD | Post Mean | Post SD | Partial Eta Squared | |
|---|---|---|---|---|---|
| Health Distress | 12.40 | 4.82 | 9.53 | 3.29 | .368 |
| HbA1c | 7.41 | 1.42 | 6.77 | 1.05 | .279 |
| Diabetes Empowerment | 31.05 | 2.38 | 32.5 | 2.01 | .243 |
| eHealth Literacy | 25.04 | 3.69 | 23.73 | 4.24 | .221 |
Table 2.
Statistics for patient and caregiver self-reported health measures MANOVA
| Patients | ||||
|---|---|---|---|---|
| Pre Mean | Pre SD | Post Mean | Post SD | |
| Health Distress | 12.0 | 4.85 | 8.44 | 3.5 |
| QALY | .826 | .119 | .866 | .113 |
| eHealth Literacy | 25.89 | 3.79 | 23.56 | 4.59 |
| Caregivers | ||||
| Health Distress | 5.17 | 2.04 | 7.0 | 3.1 |
| QALY | .802 | .046 | .787 | .064 |
| eHealth Literacy | 30.0 | 6.13 | 28.0 | 3.03 |
References
- 1.Canadian Diabetes Association Diabetes: Canada at the Tipping Point [Internet] 2010. [cited 2015 Mar 3]. Available from: http://www.diabetes.ca/CDA/media/documents/publications-and-newsletters/advocacy-reports/canada-at-the-tipping-point-english.pdf.
- 2.uOttawa Facts & Figures: Diabetes [Internet] 2015. [cited 2015 Mar 9]. Available from: http://www.med.uottawa.ca/sim/data/Diabetes_e.htm.
- 3.Norris SL, Engelgau MM, Venkat Narayan KM. Effectiveness of Self-Management Training in Type 2 Diabetes: A systematic review of randomized controlled trials. Diabetes Care. 2001;24(3):561–587. doi: 10.2337/diacare.24.3.561. [DOI] [PubMed] [Google Scholar]
- 4.Norris SL, Lau J, Smith SJ, Schmid CH, Engelgau MM. Self-Management Education for Adults With Type 2 Diabetes: A meta-analysis of the effect on glycemic control. Diabetes Care. 2002;25(7):1159–1171. doi: 10.2337/diacare.25.7.1159. [DOI] [PubMed] [Google Scholar]
- 5.Funnell MM, Brown TL, Childs BP, Haas LB, Hosey GM, Jensen B, et al. National Standards for Diabetes Self-Management Education. Diabetes Care. 2009;32(S1):S87–S94. doi: 10.2337/dc09-S087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.comScore Canada: Digital Future in Focus 2014 [Internet] 2014. [updated 2010 Feb 10; cited 2015 Mar 9]. Available from: http://cwta.ca/wordpress/wp-content/uploads/2011/08/2014-Canada-Digital-Future-in-Focus_FINAL.pdf.
- 7.Fiordelli M, Diviani N, Schulz PJ. Mapping mhealth research: A decade of evolution. J Med Internet Res. 2013;15(5) doi: 10.2196/jmir.2430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Baron J, McBain H, Newman S. The impact of mobile monitoring technologies on glycosylated hemoglobin in diabetes: A systematic review. J Diabetes Sci Technol. 2012;6(5) doi: 10.1177/193229681200600524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rosser BA, Eccleston C. Smartphone applications for pain management. J Telemed Telecare. 2011;17(6) doi: 10.1258/jtt.2011.101102. [DOI] [PubMed] [Google Scholar]
- 10.Patrick K, Griswold WG, Raab F, Intille SS. Health and the mobile phone. American Journal of Preventative Medicine. 2008;35(2) doi: 10.1016/j.amepre.2008.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Evans WD, Abroms LC, Poropatich R, Nielsen PE, Wallace JL. Mobile health evaluation methods: The Text4baby case study. J Health Commun. 2012;17 doi: 10.1080/10810730.2011.649157. [DOI] [PubMed] [Google Scholar]
- 12.Whittaker R. Issues in mhealth: Findings from key informant interviews. J Med Internet Res. 2012;14(5) doi: 10.2196/jmir.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bandura A, Cervone D. Differential engagement of self-reactive influences in cognitive motivation. Organizational Behaviour and Human Decision Processes. 1986;38(1):92–113. [Google Scholar]