The prevalence of diabetes continues to grow in the United States, affecting an estimated 8.3% of the population (25.8 million people) (National Diabetes Fact Sheet, 2011). In addition, the national prevalence of diabetes is estimated to affect approximately 26.5% of the population by the year 2050 (Boyle, Thompson, Gregg, Barker, & Williamson, 2010). Lifestyle modification and behavior changes are key components of diabetes management, especially type 2 diabetes (Norris, Lau, Smith, Schmid, & Engelgau, 2002).
The Internet and related mobile technologies present a widely accessible, 24-hour means to promote disease management and facilitate behavior modification (Kaufman, 2010). These social media environments provide popular venues in which patients gain health-related information. Thus, web- based strategies provide a viable option for facilitating diabetes self-management. Consequently, the implementation of web-based interventions to assist with diabetes management has exploded over the past decade (Chomutare, Fernandez-Luque, Arsand, & Hartvigsen, 2011; Osborn, Mayberry, Mulvaney, & Hess, 2010). To date, the majority have focused specifically on using web-based technology to facilitate the glucose monitoring process, allowing patients to upload monitoring data so their physician can adjust the dosage of insulin or medication (Chomutare et al., 2011; Harris, Hood, & Mulvaney, 2012; Yu et al., 2012). Generally these types of intervention have shown enhanced patient–provider communication, medication adherence and ultimately an improvement in glucose control.
A review of Internet diabetes programs published in 2011 identified over 137 web-based mobile applications, with most focused on insulin titration and very few focused on lifestyle modification (Chomutare et al., 2011). Several disease-specific information exchanges also now exist on Facebook and other online social networking sites. Greene, Choudry, Kilabuk and Shrank (2011) identified the fifteen largest Facebook groups focused on diabetes management, in which patients with diabetes, family members, and their friends use Facebook to share personal clinical information, to request disease-specific guidance and feedback, and to receive emotional support. In addition to web-based interventions for diabetes management, interest in mobile health applications for self-management of diabetes is growing. Chomutare and colleagues (2011) found 60 diabetes applications on iTunes for iPhone in July 2009, but by February 2011 the number had increased by more than 400% to 260.
Healtheo360 is one such web-based intervention designed to assist individuals in managing their diabetes. Healtheo360 was designed as a “caring network” where patients living with chronic conditions, such as diabetes, cancer, and Alzheimer’s, can go to find support, encouragement and information. Integrating social media with healthcare, healtheo360 is promoting the voice of patients and caregivers through its video health platform in which patients, their caregivers, family members and friends share their stories through self-generated videos so that others may find inspiration, motivation and support. Members and visitors can explore a diverse assembly of specialized support groups and connect with people just like them. This community can also follow others on their individual journey, gain access to invaluable resources, learn something new, share their knowledge and most importantly, help others in need. Comments are allowed and encouraged to promote lively discussions. A member- generated video library provides longitudinal health journals that cover symptoms, diagnosis, treatment, remission, and various health conversations. Healtheo360 recently commissioned two patient studies designed to measure the advantages of Virtual Social Therapy® for type 1 and type 2 diabetics via healtheo360’s interactive video platform. Results of the two studies are discussed below.
The first research study sought to determine the effects of participation in Virtual Social Therapy® via the healtheo360 community for patients with type 1 and type 2 diabetes. Patients from the healtheo360 community were recruited online via the website and compared with new participants recruited from other diabetes communities, social networks, and the online yellow pages. The study involved the completion of a paper questionnaire at the end of six months participation, and comparing results to the survey completed by diabetic subjects involved with healtheo360.
A total of 231 study packets were completed during the 9-month data collection period. Characteristics and frequencies for gender, ethnic origin, highest level of education, work status, marital status of members and non-members in Study 1 are included in Table 1. In Study 1, there were 103 members of the healtheo360 community and 128 non-members. The member and non-member samples were similar in terms of gender with the member group consisting of 70 (68.0%) females and 33 (32.0%) males and the non-member group consisting of 83 (64.8%) females and 45 (35.2%) males. The ethnic background of the non-members was as follows: 10 (7.8%) Asian or Pacific Islander, 22 (17.2%) Black not Hispanic, 11 (8.6%) Hispanic, 1 (0.8%) Filipino, 1 American Indian/Alaskan Native (0.8%), 78 (60.9%) White not Hispanic, and 5 (3.9%) Other. For the member group, the ethnic background was similar with: 5 (4.9%), Asian or Pacific Islander, 23 (22.3%) Black not Hispanic, 10 (9.7%) Hispanic, 62 (60.2%) White not Hispanic, and 2 (1.9%) Other. Education was also similar among members and non- members. For nonmembers, the highest level of education completed was: 3 (2.9%) some high school, 8 (7.8%) high school graduate, 33 (32.0%) some college/university, 46 (44.7%) college/university graduate, 13 (12.6%) post-graduate education. For members, the highest level of education completed was: 2 (1.6%) some high school, 13 (10.2%) high school graduate, 49 (38.3%) some college/university, 49 (38.3%) college/university graduate, and 15 (11.7%) post-graduate education. Current work status reported by non-members was: 1 (0.8%) high school student, 21 (16.4%) full-time college/university student, 3 (2.3%) part-time college/university student, 56 (43.8%) work full-time (35 hours or more per week), 17 (13.3%) work part-time (less than 35 hours per week), 4 (3.1%) stay at home parent, 10 (7.8%) retired, and 16 (12.5%) neither work nor school. In contrast, current work status reported by members was: 8 (7.8%) full-time college/university student, 4 (3.9%) part-time college/university student, 51 (49.5%) work full-time (35 hours or more per week), 15 (14.6%) work part-time (less than 35 hours per week), 6 (5.8%) stay at home parent, 9 (8.7%) retired, and 10 (9.7%) neither work nor school. Marital status among the members and non-members was also comparable. For the non-members, 40 (31.3%) reported being married, 3 (2.3%) separated, 3 (2.3%) widowed, 72 (56.3%) single, and 10 (7.8%) divorced. For members, 40 (38.8%) reported being married, 2 (1.9%) separated, 3 (2.9%) widowed, 51 (49.5%) single, and 7 (6.8%) divorced. Among members, 34 (33.0%) had type 1 diabetes and 69 (67.0%) had type 2 diabetes. Among non-members, 46 (35.9%) had type 1 diabetes and 82 (64.1%) had type 2 diabetes.
The survey used in Study 1 was based on a diabetes questionnaire developed by the Stanford Patient Education Research Center at the Stanford University School of Medicine, which is available online at: http://patienteducation.stanford.edu/research/diabquest.pdf. Subjects covered in the questionnaire include: 1) General Health, 2) Symptoms, 3) Daily Activities, 4) Physical Activities, 5) Confidence About Doing Things, 6) Diet, 7) Medical Care, 8) Medications, 9) Mood, and 10) Sleep. Respondents rated most items on a Likert scale specifying their level of agreement or disagreement. Other items were rated as “yes”/ “no” or “don’t know.”
An independent subjects design was used in the first study. The study involved comparing healtheo360 online diabetes community members with six month participation to diabetic nonmembers.
There were 231 participants in the final sample for the first study, 103 were members of the healtheo360 online diabetes community and 128 were non-members.
Results of the General Health section of the survey, indicated that members of the healtheo360 group stated that their health interfered more with their daily activities than non-members with a significant difference in the scores for interference for members (M = 7.058, SD = 4.452) and non- members (M = 5.210, SD =4.154); [t(229) = -3.254, p = .001] (see Table 3). Also, there was a significant difference for exercise between members (M = 6.515, SD = 4.614) and non-members (M = 5.023, SD = 4.098); [t(229) = -2.598, p=.01] (see Table 3). Hence, despite indicating that their health interfered more with their daily activities, members of the healtheo360 community reported significantly more time engaged in physical activities
When queried about their confidence to manage health and diabetes-related activities of daily living such as eating, diet, exercise, and blood sugar, members of the healtheo360 community did not differ significantly from non-members (p = .184). Following the questions related to confidence, participants were questioned more specifically about diet, medical care, medications, mood, and sleep. Regarding diet, healtheo360 members reported better eating (p = .070) but worse mood (p = .062), than nonmembers, with these results being marginally statistically significant. However, in being prepared for doctor’s visits and better sleep, the differences between the groups were not significant.
To further analyze potential differences between members and non-members of the healtheo360 community, two scales, state and action were created. The state scale included the questions that assessed worry, interference, mood, and confidence, items all assessing what state participants were in, and the action scale included the questions that assessed exercise, eating, and preparation for doctor visits, items all assessing the actions participants were doing to improve their state. Results indicated that although members were in a worse state than non-members (p=.013), they are taking better actions than non-members (p= .001). Hence, participation in the healtheo360 group contributed to greater awareness and more positive behaviors for members compared to non- members. Evidently individuals joining the healtheo360 community are doing so in order to help themselves develop more positive, healthy behaviors and manage their diabetes better. There were no differences between state and action for individuals with type 1 diabetes compared to type 2 diabetes.
Study 2 improved upon Study 1 by requiring minimum standards of participation in Virtual Social Therapy® via the healtheo360 community for patients with type 1 and type 2 diabetes. As a result, participants who joined the community but then did not utilize its resources were excluded from the study. It also improved on Study 1 by utilizing a repeated measures design. Patients from the healtheo360 community were recruited online via the community and compared with new participants recruited from other diabetes communities, social networks, and the online yellow pages. The study involved the completion of a paper questionnaire prior to joining the healtheo360 interactive video platform and at the end of participation six months later.
A total of 90 study packets were completed during the 9-month data collection period. Characteristics and frequencies for gender, ethnic origin, highest level of education, work status, marital status of members and non-members in Study 2 are included in Table 2. In Study 2, there were 90 members of the healtheo360 community who participated in the study. In the total sample, there were 33 males (36.7%) and 57 females (63.3%). Ethnic origin was as follows: 8 (8.9%) Asian or Pacific Islanders, 19 Black not Hispanic (21.1%), 7 Hispanics (7.8%), 1 (1.1%) Filipino, and 51 (56.7%) White not Hispanic. Regarding work status, participants identified themselves as: 6 (6.7%) full-time college/university student, 4 (4.4%) part-time college/university student, 47 (52.2%) work full-time (35 hours or more per week), 14 (15.6%) work part-time (less than 35 hours per week), 6 (6.7%) stay at home parent, 5 (5.6%) retired, and 8 (8.9%) neither work nor school. Also shown in Table 2 is the level of education which included: 2 (2.2%) some high school, 7 (7.8%) high school graduate, 35 (38.9%) some college/university, 34 (37.8%) college/university graduate, and 12 (13.3%) post-graduate education. Marital status among the member group varied with 29 (32.2%) married, 2 (2.2%) widowed, 52 (57.8%) single, and 7 (7.8%) divorced. There were 28 (31.1%) patients with Type 1 diabetes and 62 (68.9%) patients with Type 2 diabetes.
The survey used in Study 2 was the same as the one used in Study 1. It was based on a diabetes questionnaire developed by the Stanford Patient Education Research Center at the Stanford University School of Medicine, which is available online at: http://patienteducation.stanford.edu/research/diabquest.pdf. Subjects covered in the questionnaire include: 1) General Health, 2) Symptoms, 3) Daily Activities, 4) Physical Activities, 5) Confidence About Doing Things, 6) Diet, 7) Medical Care, 8) Medications, 9) Mood, and 10) Sleep. Respondents rated most items on a Likert scale specifying their level of agreement or disagreement. Other items were rated as “yes”/ “no” or “don’t know.”
A repeated measures/paired samples experimental design was used in the second study. T-tests were run for each of the seven scales of worry, symptoms, interference, exercise, confidence, eating and mood to compare pre-test scores to post-test scores. Six months elapsed between the completion of the pre-test and the post-test. Equivalence testing revealed that members with diabetes did not decline over time. A one-way confidence interval with α = .05 was used in conjunction to the standard deviations to calculate an effect size. Next, mood was used as an independent variable to compute correlations between mood and the other six scales assessing symptoms, worry, interference, exercise, confidence, and eating. Lastly, for each subject, pre and post-test scores were converted to z-scores which were then converted into categorical variables for each participant who experienced a mood change. Category 1 contained z-scores from -1 to -3 and contained scores of participants whose mood declined the most, category 2 contained z-scores from -1 to 0 and contained scores of participants whose mood declined somewhat, category 3 contained z-scores from 0 to 1 and contained scores of participants whose mood improved somewhat, and category 4 contained z-scores from 1 to 3 and contained scores of participants whose mood improved the most. Analysis of variance (ANOVA) tests were then computed using mood change as the independent variable and the six other scales as the dependent variables.
There were 90 participants in the final sample for the second study. Participants completed the pre-test prior to participation in the online network and the post-test at six months after joining. None of the seven t-tests revealed any significant differences between pre- and post-test scores on the questionnaire (worry p = .917, symptoms p = .403, interference p = .696, exercise p = .701, confidence p = .166, eating p = .641, mood p = .804). Non-significance indicates a difference was not found, it is not the same as proving there was no difference. But as diabetics tend to decline over time, we wanted to prove that those on healtheo360 did not. Equivalence testing allows us to prove that two groups do not differ to a practically significant extent. Equivalence testing using the 95% confidence interval shows the largest possible decline in terms of effect size. In Cohen's universally accepted guidelines, an effect size of .2 is small and hardly visible. In contrast an effect size would have to reach .5 to be medium. On the worry scale there was an effect size = .19, on the symptoms scale there was an effect size = .09, on the interference scale there was an effect size = .22, on the exercise scale there was an effect size = .22, on the confidence scale there was an effect size = .32, and on the eating scale there was an effect size = .13. All of these effect sizes indicate that the largest possible declines in a population of diabetics being on healtheo360 over six months, being on each of the scales small and not practically significant. Therefore, although participants cannot be shown to improve, they definitely did not get worse over time, a finding that can be generalized would be true of the population.
Next when mood change was the independent variable and the other six scales were the dependent variables correlations revealed several significant relationships. The correlation between mood change and symptom change was r = .332, p= .002 indicating that individuals whose mood improved had less symptoms. Also, there was a significant correlation between mood change and interference (r = .375, p = .001) indicating that as participant’s mood improved, they experienced less interference in their lives. Likewise, there was a significant correlation between mood and confidence (r = .209, p = .048), indicating that as participants’ mood improved, their confidence improved. Lastly, there was a significant correlation between mood change and eating (r = .229, p = .03), indicating that as
mood improved, participants’ eating improved. Mood change was then converted into a categorical variable. Analysis of variance (ANOVA) tests indicated that effect of mood change on exercise was not significant [F (3, 86) = .620, p = .604], but that the change in mood and the five other variables: symptoms [F (3,77)= 2.803, p=.045], worry [F(3, 86) = 2.852, p = .042], interference [F (3, 86) = 5.576, p = .002], confidence [F (3, 86) = 3.548, p = .018], and eating [F (3, 86) = 3.389, p = .022] were significant at the .05 level (see Table 4). Therefore, compared to participants whose mood did not improve, participants whose mood improved the most showed significant improvement in the other five areas. Consequently, a positive mood change resulting from participation in healtheo360 resulted in fewer symptoms, less worry, less interference, and better eating in participants.
It was initially hypothesized that healtheo360 participants would show improvement in behaviors and health in the areas of Symptoms, Interference, Confidence About Doing Things, Diet, Medical Care, Medications, Mood, and Sleep after 6 months of participation. Although members did demonstrate significantly more time engaged in physical activities and ate better, they were less confident and demonstrated poorer general health, more symptoms, more interference in their daily activities, and less confidence about their ability to manage their health and diabetes-related activities of daily living. However, it became obvious that members were in a worse state than non-members upon joining; their relatively poor state was the motivation for joining. The results also showed that although members were in a worse state than non-members, they are taking better actions than non- members. Therefore, participation in healtheo360 resulted in greater awareness and improved attempts at self-management. In these areas, participation in healtheo360 produced clear and unambiguous benefits. Results of the second study revealed that individuals with diabetes who participated in the healtheo360 community and met the minimal participation requirements did not decline over time. Not unimportant when considering that diabetics do tend to decline over time. And if their mood improved from pre-test to post-test, they did experience improvement in other areas including symptoms, interference, confidence, and exercise.
One lesson we might take from these two studies is that education for diabetes patients and the promotion of awareness appears to be critical. The results highlight the importance of education and suggest that education should be added to existing components of the healtheo360 Virtual Social Therapy® community. The results further suggest possible evidence of the “Hawthorne effect,” a phenomenon in which subjects in behavioral studies change their performance in response to being observed. Participation in the study may have caused the adjustment in the healtheo360 participants’ behaviors and their resultant improvements. Although participants in this study were assessed upon joining and then six months later, it would be interesting to assess more long-term effects. It is possible that additional positive behavior changes could occur in members given more time. Likewise, it is possible that greater participation in the healtheo360 community may have differential effects, so it would be useful to further examine participation level in future studies.
It may also be worthwhile to further examine the components of the existing healtheo360 Virtual Social Therapy® community and to consider providing patients with not only the opportunity to interact with other diabetic patients, but to also track their blood glucose, receive electronic reminders, schedule physician visits, and email their health care team. A previous review showed that patient satisfaction is highest when the web-based system provides patients with these opportunities (Brown,
Lustria & Rankins, 2007). Additional limitations of existing smartphone and web-based applications such as healtheo360 include their lack of personalized feedback, usability issues, and integration with patients and electronic health records (El-Gayar, Timsina, Nawar, & Eid, 2013). Unfortunately, such comprehensive medical and self-management programs have not been implemented widely outside of systems funded by government agencies (Brown et al., 2007), likely due to their cost. The cost of developing and maintaining comprehensive systems continues to be a challenge and lack of reimbursement for web-based treatments is also a major barrier to implementation (Brown et al., 2007). However, as demonstrated by the results of these two studies, diabetes patients who participate in online communities and interventions such as the healtheo360 community, experience many positive health, mood, and behavioral effects.
Consequently, these benefits would eliminate or reduce the need for more costly interventions and medical care for patients with diabetes, which would ultimately result in long-term cost-savings for insurance companies and other stakeholders. Therefore, insurance companies need to consider Virtual Social Therapy® web-based treatments such as healtheo360 to reduce cost and improve health among their covered lives.
Implications Based on Rational Emotive Behavior Therapy (REBT)
The purpose of the present studies was to determine whether participation in the healtheo360 community would improve symptoms, worry, interference, exercise, confidence, and eating behaviors of patients with diabetes. More specifically, the findings showed that although participants did not necessarily indicate symptom improvement, they did show improved eating, and engagement in physical activities. Likewise, participation in the healtheo360 group contributed to greater awareness and more positive behaviors for members, suggesting that individuals may be joining the healtheo360 community in order to help themselves develop more positive, healthy behaviors and manage their diabetes better. Positive mood change following participation in the healtheo360 community also resulted in fewer symptoms, less worry, less interference, and better eating in participants.
These results are consistent with the premise of Rational Emotive Behavior Therapy (REBT) developed by Albert Ellis in the mid ‘50s. REBT is based on the theory that emotional disturbance is the result of illogical and irrational ways of thinking (Ellis & Wilde, 2002). It takes an interactionist approach in purporting that thoughts, feelings, and behaviors interact and significantly influence each other. According to this model, a person is confronted with an unpleasant activating event (A), about which he or she has rational (i.e., adaptive) or irrational (i.e., maladaptive) beliefs (B). These beliefs generate cognitive, behavioral and emotional consequences (C). Because individuals’ cognitions, emotions, and behaviors are interdependent (Bernard, Ellis & Terjesen, 2006), it is critical for therapists and treatment professionals to better understand the relationship between individuals’ beliefs/thoughts, affect, and behavior. While irrational beliefs lead to negative emotional consequences, hindering a person from reaching his goals, rational beliefs will result in neutral or positive emotional consequences and will help the person achieve more and feel better.
It is not only external events alone that cause emotional disturbance, but the events in combination with a person’s perceptions and evaluations about them. The REBT approach attempts to show clients that at the core of their emotional and behavioral disturbances they have one or more irrational beliefs that they strongly hold and subsequently act on. Since the goal of REBT is to improve an individual’s behavior and emotional functioning, an REBT therapist may seek to identify an individual’s specific irrational belief systems in order to facilitate changing them to more appropriate,
rational beliefs. REBT clients are encouraged to dispute (D) their irrational beliefs, in order to acquire more efficient (E) or adaptive and rational beliefs, and subsequently feeling and behaving in a better manner. An REBT therapist then teaches the client how to deal with these irrational beliefs by actively and strongly disputing and challenging these beliefs. He or she can then replace them with more sensible, functional, rational beliefs and subsequently feel and behave in a better manner. Furthermore, if therapists and medical professionals are able to identify specific beliefs that lead to specific emotions (e.g. anger, anxiety, depression) and behaviors (i.e. academic failure, social avoidance) it may lead to greater efficiency in the delivery of psychological and medical services for diabetes patients and produce more enduring and meaningful changes.
The protocols provide rich data on participants’ emotions and behaviors. However, these results may be limited to comparable contexts and comparable participants. Furthermore, because the diabetes patients who participated in the program were volunteers, the results may reflect volunteer characteristics. Because the study’s methods involved self-report, the participants may not have expressed their actual behaviors or state. Certainly an objective measure of state such as Hemoglobin A1c would have been an asset. Nevertheless the two studies produced provocative and promising results. Membership in healtheo360 was correlated with better eating and exercise behaviors, was associated with the conditions being stabilized over time and produced benefits all across the board if mood also improved.
Further research is needed to clarify the complex relationship between cognitions and irrational beliefs on the emotional and behavioral functioning among patients with diabetes. More studies are needed that investigate the relationship between these patients’ social-emotional functioning, behavior, and underlying thought processes. More research is needed to determine if these underlying thought processes or irrational beliefs contribute to particular behaviors in patients with diabetes and whether rational emotive behavior therapy (REBT) and/or rational emotive education (REE) improve their psychological and behavioral problems. It is recommended that future research on patients with diabetes include measures of their irrational beliefs, as well as an objective measure of their state.
Bernard, M.E., Ellis, A. & Terjesen, M. (2006). Rational-emotive behavioral approaches to childhood disorders: History, theory, practice and research. In A. Ellis & M.E. Bernard (Eds.), Rational Emotive Behavioral Approaches to Childhood Disorders: Theory, Practice and Research (pp. 3- 84). New York: Springer Science & Media.
Boyle, J. P., Thompson, T. J., Gregg, E. W., Barker, L. E., & Williamson, D. F. (2010). Projection of the year 2050 burden of diabetes in the US adult population: Dynamic modeling of incidence, mortality, and prediabetes prevalence. Population Health Metrics, 8, 29.
Brown, L.L., Lustria, L.A. & Rankins, J. (2007). A review of web-assisted interventions for diabetes management: Maximizing the potential for improving health outcomes. Journal of Diabetes Science and Technology, 1 (6), 892-902.
Chomutare, T., Fernandez-Luque, L., Arsand, E., & Hartvigsen, G. (2011). Features of mobile diabetes applications: Review of the literature and analysis of current applications compared against evidence-based guidelines. Journal of Medical Internet Research, 13, e65.
El-Gayar, O., Timsina, P., Nawar, N., & Eid, W. (2013). Mobile applications for diabetes self-management: Status and potential. Journal of Diabetes Science & Technology, 7 (1), 247-262.
Ellis, A. & Wilde, J. (2002). Case studies in rational emotive behavior therapy with children and adolescents. Upper Saddle River, NJ: Pearson Education.Greene, J.A., Choudry, N.K., Kilabuk, E., & Shrank, W.H. (2011). Online social networking by patients with diabetes: A qualitative evaluation of communication with Facebook. Journal of General Internal Medicine, 26 (3), 287- 292.
Harris, M. A., Hood, K. K., & Mulvaney, S. A. (2012). Pumpers, skypers, surfers and texters: Technology to improve the management of diabetes in teenagers. Diabetes, Obesity & Metabolism, 14, 967- 972.
Kaufman, N. (2010). Internet and information technology use in treatment of diabetes. International Journal of Clinical Practice, Supplement, 41–46.
National diabetes fact sheet: National estimates and general information on diabetes and pre-diabetes in the United States, 2011. (Accessed at http://apps.nccd.cdc.gov/DDT_STRS2/CountyPrevalenceData.aspx?mode=DBT.).
Norris, S. L., Lau, J., Smith, S. J., Schmid, C. H., & Engelgau, M. M. (2002). Self-management education for adults with type 2 diabetes: A meta-analysis of the effect on glycemic control. Diabetes Care, 25, 1159–1171.
Yu, C. H., Bahniwal, R., Laupacis, A., Leung, E., Orr, M. S., & Straus, S. E. (2012). Systematic review and evaluation of web-accessible tools for management of diabetes and related cardiovascular risk factors by patients and healthcare providers. Journal of the American Medical Informatics Association, 19, 514–522.