A Social Media Intervention for Exercise Motivation and Cardiac Rehabilitation Adherence

There is no research that addresses the use of Facebook as a tool to affect autonomous motivation in cardiac rehabilitation. The purpose of the current feasibility study is to determine the feasibility of using a Facebook intervention, providing education, peer support and provider support, to affect change in motivation and self-determination for exercise, and adherence to cardiac rehabilitation in patients with CHD during a 12-week Phase II cardiac rehabilitation program. It is hypothesized that: 1. Scores for motivation for exercise overall will increase for patients exposed to a Facebook intervention and across individual motivational subtypes (regulations) relative to a comparison group who receive educational handouts and emails. 2. Percentage of cardiac rehabilitation sessions attended will be higher relative to a comparison group who receive educational handouts and emails. 3. Engagement in the private Facebook group (number of "hits" and "likes") will predict number of cardiac rehabilitation sessions attended and the change in motivation. The feasibility of a larger trial will be based on sample size and participants' engagement in the Facebook group. Methodology Design This is a prospective, randomized controlled pilot trial to evaluate the feasibility of using a social media intervention to affect change in motivation for exercise and adherence to cardiac rehabilitation. Setting and Sample: The setting for this study will be in the outpatient cardiac rehabilitation at the main campus of a large tertiary care center in Northeast Ohio and several satellite facilities in the region. All patients receive an individualized exercise prescription based on functional capacity at intake. Most patients will be able to attend up to 3 sessions per week for a total of 36 sessions. All patients who are current and regular Facebook users, have qualified for cardiac rehabilitation (diagnosed with CHD), and are entering cardiac rehabilitation at the main campus or one of the regional satellite hospitals of this tertiary care center, will qualify to participate in the study prior to beginning Phase II cardiac rehabilitation. Current Facebook users were chosen . Regular use will be defined as logging onto Facebook at least 2 times in the last month. Inclusion criteria will include both men and women 18 years of age or older and live within 100 miles of the main campus of this tertiary care center. Participants must be able to read and understand English in order to complete the questionnaires: the Psychological Need Satisfaction in Exercise Scale (PNSE) \[37\] and the Behavioral Regulation in Exercise Questionnaire-3 (BREQ-3). There will be no exclusion based on secondary diagnosis; however, participants must be able to exercise well enough to qualify to take part in cardiac rehabilitation. Measures: Change in motivation for exercise, will be measured at baseline and post-intervention using the BREQ-3. The BREQ-3 is a 24 question validated instrument that measures forms of intrinsic and extrinsic regulation of exercise behavior and is based on self-determination theory. Psychometrics were completed for the BREQ-2 by Markland and Tobin. Cronbach's alpha reliabilities were: amotivation = 0.83, external regulation = 0.79, introjected regulation = 0.80, identified regulation = 0.73, and intrinsic regulation = 0.86. The BREQ-3 includes 5 additional questions in addition to those on the BREQ-2 and has a new subscale for integrated regulation \[33\]. The subscales (regulations) of the BREQ-3 are used to calculate a relative autonomy index (RAI). Each question is answered on a 5 point Likert scale (0-4) and represents one of the regulations. The regulations are weighted then summed to give a single score. The resulting index gives an indication of individual respondent's self-determination for exercise. The RAI will place individual motivational subtypes or behavioral regulations on the self-determination continuum from amotivated (lacking intention to exercise) to intrinsically motivated (self-determined or autonomously motivated). The PNSE will be used to assess need satisfaction with exercise. This scale was designed to assess the perception of psychological need satisfaction associated with self-determined motivation for exercise and consists of 18 items on a 6 point Likert scale, with 3 subscales measuring perceived competence, autonomy, and relatedness. The scale showed high internal consistency (Cronbach \>0.90) \[37\]. The percentage of cardiac rehabilitation sessions attended will be measured at the time of cardiac rehabilitation completion or dropout. It will calculated by dividing the number of sessions attended in a 3 month period of time by the total number of sessions allowed by insurance, and multiplying by 100. Facebook engagement, will be assessed by measuring the number of "likes" by individuals on the private Facebook group. "Likes" (the number of times a participant clicks "like" on any of the Facebook posts) will be counted and, along with "hits" will be used to examine the association between engagement in the intervention, and cardiac rehabilitation adherence and change in motivation. A post-intervention questionnaire will be given to determine number of "hits". The participants will be asked to circle the number of times they accessed the private Facebook group per week: 0, 1-5, 6-10, 11-15 or \> 15 times. The questionnaire will also be used to collect qualitative data on participants' perceptions of the intervention, including whether they felt supported in their care, more in touch with providers, whether or not they chatted with other Facebook members and if the Facebook group affected their exercise behaviors. The questionnaire will use a Likert scale (1, "not at all"-5, "quite a bit") for all questions and a section for comments. Patient characteristics will include key demographic variables (age, gender, race, employment, distance to cardiac rehabilitation, socioeconomic status), engagement (number of "hits" and "likes"), and key clinical variables (functional capacity as measured in METS,cardiac rehabilitation indication, hypertension, diabetes, hyperlipidemia and waist circumference), which will be obtained from the electronic medical record. Data Collection Procedures: Volunteers will be recruited from the main campus or regional satellite hospitals of this tertiary care center during their inpatient stay, at the intake visit for cardiac rehabilitation at the main campus and satellite facilities in the region, or via phone call after discharge and up to the first cardiac rehabilitation session. Volunteers will be screened for Facebook use and interest in the study, then an email link will be sent which will include an information sheet. The Facebook group will be private in the sense that those not in the group will not be able to see the content. Participants will be given a baseline BREQ-3 questionnaire and PNSE scale. Participants will then be randomized to Facebook versus comparison groups using blocked randomization. Intervention. The Facebook intervention will include peer support, education, provider support. These interventions are designed to minimize pressure, offer choices, and allow for peer interaction, and positive feedback in order to provide support for competence, autonomy and relatedness. Competence will be supported with use of educational posts. Autonomy support will come from provider posts. Relatedness will be supported by peer interaction and engagement in the Facebook group. Educational posts will cover topics that will encourage participants to practice preventive heart care. The educational portion of the intervention is designed to offer clear information and structure, thus supporting competence. The posts may be in the form of text, video and/or pictures and will include materials from the hospital's health library, the American Heart Association and the Center for Disease Control. Provider posts will include topics such as motivational quotes, encouragement, reminders to exercise independently, and reminders to contact providers with questions. These postings are designed to promote a sense of choice and help participants feel that providers see them as having a unique frame of reference thus being autonomy-supportive. Providers will be nurses on the research team, exercise physiologists nurse practitioners and physicians. Provider support will also include links to provider health chats, in which patients can chat online with providers at set dates and times. Peer interaction on Facebook will be as frequently as the participant chooses and will be monitored daily by the research team. Engagement in Facebook is designed to offer an opportunity for social inclusion and a sense of involvement. The comparison group will receive the same educational and provider support materials as the Facebook group but will receive it via email. Both groups will have the opportunity for weekly education classes and typical peer interactions. Upon cardiac rehabilitation completion or dropout, post-data will be collected. It is anticipated that this pilot will take up to one year and will be completed when 30 participants for each group have been obtained. Data Analysis Statistical Methods. This is a feasibility study and the sample size obtained will determine if the study is appropriately powered to detect the desired effect size. Patient characteristics will be summarized group using frequencies and percentages for categorical factors, and means and standard deviations for continuous measures. In order to examine the primary outcome, differences in change in motivation between groups, overall motivation using the RAI from the BREQ-3 will be evaluated using analysis of covariance (ANCOVA) models. Mean differences with 95% confidence intervals for group differences will be presented. Multivariate analysis of variance models will be used to evaluate differences in the change across individual motivation subtypes (regulations), using the BREQ-3, between groups overall. If significant, separate ANCOVA models for each subtype will be fit. Similar models will be used to compare changes in needs satisfaction scores, overall and separately among the three subscales, between groups. Two-sample t-tests will be used to compare number of sessions completed. As a secondary analysis, the relationships between patient characteristics, "hits" and "likes", and the outcome variables RAI change, number of sessions, and needs satisfaction change will be examined using t-tests and Pearson correlations. The correlation between changes in RAI and needs satisfaction will also be evaluated. Analyses will be performed using SAS software (version 9.4; Cary, NC). An overall significance level of 0.05 will be assumed for all tests. Sample Size. The investigators plan to enroll 30 patients in each group. In the first 9 months of 2016, cardiac rehabilitation at the main campus of this tertiary care center had approximately 170 patient intakes. It is assumed that there will be a similar number of patient intakes for a 9 month period in 2017. Based on Facebook participation rates for those over age 50 \[21\] and the high participation rates in previous research projects in this facility's cardiac rehabilitation, it is estimated that 40% may meet eligibility requirements and agree to participate. Allowing for use of the first 8 participants to establish the Facebook group, the estimated sample size would then be 60 total participants for randomization to study group who can then be included in analysis. With this sample size, there will be 86% power to detect large effect sizes (d=0.8) for our study outcomes \[40\]. The primary aim of this sample size determination is to evaluate whether the proposed intervention is feasible, and to estimate the differences that might exist so that a larger trial that would have adequate power to detect smaller differences could be designed based on what was learned in this pilot study. The sample size of 30 per group was chosen primarily to facilitate a large intervention group, since the value of the intervention is predicated upon interaction among the participants.