Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website

CRE 12-285 – HSR Study

CRE 12-285
Randomized Controlled Trial of Group Prevention Coaching
David Edelman, MD MHS
Durham VA Medical Center, Durham, NC
Durham, NC
Funding Period: November 2013 - October 2018
Primary prevention of cardiovascular disease (CVD) has been challenging to achieve; evaluation of new strategies for CVD primary prevention is the focus of this project, which is part of a CREATE (Collaborative Research to Enhance and Advance Transformation and Excellence) Program addressing prevention.

Multifactorial behavioral interventions are interventions that, instead of attempting to change a single health behavior among patients in a population (e.g., increased physical activity, smoking cessation), allow individual patients to choose one or more among a number of behaviors, any of which may get them to a common clinical objective (e.g., lowering blood pressure). These interventions often use some of a set of common approaches, such as goal-setting and improving self-efficacy, to assist patients in behavior modification. A recent systematic review suggests multifactorial behavioral interventions are likely efficacious in secondary prevention of CVD, and these interventions are also efficacious in a number of chronic illnesses (e.g., hypertension, diabetes). While there are studies that assess the performance of these interventions in patients without a common illness, there is little literature directly measuring risk reduction in a primary prevention population by means of multiple modest behavior changes; our group has performed such a study, but in a population very unlike VA (mostly white, mostly female, all insured).

Problem solving (PS) is an approach to behavior change that has a long history in the mental health literature, but is much less used, and less well tested, in the prevention arena. PS is an evolution of cognitive-behavioral therapy that addresses internal barriers to changing behavior, urging patients to use specific techniques designed to prime them for success in any behavior change they propose to attempt. PS is well-tested in mental health, and there are small studies suggesting its utility in changing behaviors associated with improved physical health. It is traditionally done in an individual setting, but recent literature indicates promise in use in group settings. However, PS has unproven effectiveness both in group settings and among primary prevention patients.

Group visits have been shown to be an effective means of improving outcomes in a number of settings. Group visits can either involve medication management or be entirely focused on behavior change; self-management groups (SMGs) are behavior intervention groups that are conducted by a trained facilitator and focus on having patients teach each other how to adopt healthful behaviors (usually in chronic illness). Because there is no requirement for medication management, SMGs can be used efficiently in patients without medication-requiring chronic illness, as there is no attendant cost of employing a prescribing provider. SMGs can be used as a strategy to deliver a multifactorial behavior intervention in patients with a common illness. However, our study above is the only RCT we can find of SMGs in primary prevention.

Our primary objective was to determine the effectiveness of group prevention coaching (GPC), compared to VA usual care, in reducing CVD risk in a primary prevention population. Our primary hypothesis was that, among patients with >5% 10-year risk for a cardiovascular event, there will be at least a 2.0% difference between GPC and usual care in reduction of predicted rate of fatal or non-fatal myocardial infarction (MI), as estimated by the Framingham Risk Score (FRS). We sought to determine the effectiveness of GPC (compared to VA usual care) in: (a) increasing physical activity, and (b) making dietary improvements. We also sought to determine the role of group cohesion in predicting the success (or lack thereof) of GPC among intervention patients.

We performed a randomized, controlled trial (RCT) of our GPC intervention. Outcomes were be measured at baseline, 6 months, and 12 months after enrollment. The control group received usual care. We performed the study among patients at the Durham and Buffalo VAMCs.

The patients were 401 VA users without prior history of cardiovascular event, but with at least 5% risk of such an event, with at least 2% of that risk potentially reversible. Risk was calculated by Framingham Risk Score (see primary outcome). We excluded patients with known significant atherosclerotic disease (coronary artery disease, stroke, peripheral arterial disease). To meet these criteria, all patients in the study had hypertension with less-than-optimal control, OR elevated total cholesterol, OR were current smokers. We also excluded patients with other severe intercurrent illness or poor life expectancy, or who were currently engaged in formal efforts to improve a CV risk behavior, or who were cognitively impaired or unable to use a telephone.

Subjects provided written informed consent and completed a measurement battery. They were randomized, 1:1 and stratified by site, to receive or not receive the GPC intervention. Patients were assessed 6 and 12 months after baseline by a blinded research assistant.

The GPC intervention was delivered to patients in the active arm. The basic structure was that of a group problem-solving intervention, with interval phone calls delivered between each group session to check in on goal progress and reinforce group learning. Groups met at weeks zero, 3, 6, 10, 14, and 18; a focus group convened at 24 weeks without any intervention content. Each patient was to be called once between each group session. Groups had 5-15 patients, but all but 2 groups had 8-12 patients. Problem-solving principles represented the content of the group sessions. Problem solving intends to teach patients to overcome internal barriers to healthful behaviors. The groups will be coordinated by a clinical psychologist hired for the research enterprise but trained and credentialed identically to a facility's Health Behavior Coordinator (HBC). The phone calls associated with the intervention were performed by a healthcare coach.

Our primary outcome variable was Framingham Risk Score. FRS is calculated using age, gender, blood pressure, total and HDL cholesterol, smoking status, diabetes status, and status of taking blood pressure medication. All components of the FRS were measured at each scientific assessment. We also measured physical activity by International Physical Activity Questionnaire, caloric intake by Block Brief 2000 Food Frequency Questionnaire, weight, self-efficacy by Patient Activation Measure (range, 0-100), group cohesion by the Group Cohesion Scale, and Problem Solving skill by the Social Problem Solving Inventory (range, 0-139). We also measured a number of socio-demographic covariates at baseline.

Analysis was intent-to-treat, and performed by linear mixed modeling, adjusting for group clustering in the GPC arm.

We randomized 202 patients to GPC and 199 to usual care. Mean age was 62.4 years, 10% were women, and 32% were black. Mean 10-year risk of cardiovascular event was 29.2% by Framingham Risk Score (FRS); baseline mean Patient Activation Measure (PAM) and Social Problem Solving Inventory (SPSI) scores were 60 and 105, respectively. Follow-up rate at 6-months was 62% in GPC and 64% in usual care. GPC patients attended a mean of 3.4 (of 6) group sessions; 43.6% attended 5 or 6 groups, and 14.9% did not attend any. Patients completed a mean of 3.9 (of 6 calls); 57.4% completed 5 or 6 calls, and 14.4% completed no calls. Among GPC patients, 47.7% chose to work on improved eating or weight loss, 17.9% chose to increase physical activity, 27.8% chose to quit smoking, and 6.6% chose another strategy for improving risk (medication adherence or cessation of non-tobacco substance). In the LMM analyses, there was no difference at 6 months between GPC and UC patients in FRS (mean difference 0.7%, 95% CI, -1.4%, 2.8%), PAM score (mean differences 1.3 points, 95% CI -1.3, 3.9), or SPSI (mean difference 1.6 points, 95% CI -0.8, 4.0).

Other outcome analyses (especially physical activity and diet content) are pending, as is a completers analysis.

Our intervention was not better than usual care in its impact on cardiovascular risk or patient activation. Patients only attended a moderate number of sessions.

External Links for this Project

NIH Reporter

Grant Number: I01HX001142-01

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


Journal Articles

  1. Nieuwsma JA, Wray LO, Voils CI, Gierisch JM, Dundon M, Coffman CJ, Jackson GL, Merwin R, Vair C, Juntilla K, White-Clark C, Jeffreys AS, Harris A, Owings M, Marr J, Edelman D. A problem-solving intervention for cardiovascular disease risk reduction in veterans: Protocol for a randomized controlled trial. Contemporary clinical trials. 2017 Sep 1; 60:42-50. [view]

DRA: Aging, Older Veterans' Health and Care, Health Systems, Cardiovascular Disease
DRE: Treatment - Efficacy/Effectiveness Clinical Trial, Prevention
Keywords: Outcomes - Patient, Risk Factors, Symptom Management, Adherence, Cardiovascular Disease
MeSH Terms: none

Questions about the HSR website? Email the Web Team

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.