Over half of all deaths, and many illnesses, can be attributed to three modifiable risk factors: tobacco use, overweight/obesity, and physical inactivity. There are clear links between these modifiable factors and heart disease, cancer, chronic lung disease, and stroke which continue to be the leading causes of death in the United States. VHA has made significant improvements in controlling conditions that lead to heart disease, cancer and stroke (e.g., hypertension and hyperlipidemia). We have not, however, done as well in addressing the underlying behavioral factors (e.g., obesity, tobacco use, and physical inactivity).
The objective of this study was to determine if a telephone-based Shared Decision Making (SDM) intervention, using VHA's web-based HealtheLiving Assessment and a Prevention Coach would increase patient activation and enrollment in prevention programs compared to usual care. The Specific Aims include: to determine whether the intervention improved activation and participation of patients in prevention programs; to determine whether patients participating in the intervention had greater reduction in cardiovascular risk; and to conduct process evaluations of the intervention and its implementation to inform future dissemination and implementation.
Two-site, two-arm effectiveness-implementation Hybrid Type 1 Intervention-Implementation trial. The study was performed at the Durham and Ann Arbor VA Medical Centers. Eligible Veterans had at least one modifiable risk factor (obese, inactive, or tobacco user) but were not currently enrolled in a prevention service. The brief telephone intervention was conducted by a prevention coach and used the output from VHA's HealtheLiving Assessment to engage Veterans in a conversation where individual preferences were matched to behaviors and choices for specific prevention services (either in the VA or community) were offered. The resulting prevention action plan was documented in the medical record. Outcomes were obtained at 1 month and 6 months after enrollment. The co-primary outcomes were change in the Patient Activation Measure (PAM) and proportion enrolled in effective prevention services. The secondary outcome is 10-year risk of coronary events, as measured by Framingham Risk Score.
Recruitment started in November 2014 and ended in May 2016. Of patients who screened eligible by phone, 79% consented (n=436) of whom 96% (n=417) were randomly assigned to one of the two arms.
On average, participants were 56 years old (SD=12.2). A large percentage of participants were African American (40%), most were male (85%), and over 25% of our participants reported inadequate income to cover bills at the end of the month reflecting a high proportion of financial vulnerability. In terms of health status, 69% of patients reported their general health as excellent, very good, or good. Eighty percent of participants met enrollment criteria because of weight (BMI 30), 50% for being moderately/vigorously active for less than 150 minutes/week, and 39% were active smokers. Approximately half of the participants had more than one eligible risk factor. The mean Health Risk Assessment (HRA)-generated health age was 60 years while the participants' average chronological age was 56 primarily reflecting their excess cardiovascular risk.
Adherence to the Intervention
Of the 208 participants randomized to the intervention, 194 participants (93%) completed the first coaching call, and 182 participants (88%) completed the second; 14 participants (7%) of intervention participants did not complete either call. The mean duration of the first coaching call was 34 minutes (SD 14 minutes), and the mean duration of the second coaching call was 12 minutes (SD 10 minutes).
At six months, intervention participants reported higher enrollment in a prevention program, 51% vs 29% (OR=2.5; 95% CI: 1.7, 3.9, p <0.0001) and higher participation in a prevention program, 40% vs. 23% (OR=2.3; 95% CI: 1.5, 3.6; p<.001) compared to usual care participants. There was no difference in PAM scores between groups at the one month assessment. However, at six months after baseline, intervention participants showed greater improvement in PAM scores than usual care participants (mean difference=2.5; 95% CI: 0.2,4.7; p=0.03).
The FRS showed almost no change over time from baseline to six months for both groups (mean difference=0.7; 95% CI: -0.7,2.2; p=0.33).
Results were robust to missing data assumptions; in our sensitivity analyses for the primary outcome of enrollment, the OR with 95% CI ranged from 2.1 (1.4, 3.2) with all missing assumed to be not enrolled to 2.7 (1.8, 3.9) for all missing assumed to be enrolled.
Prevention Programs Chosen by Intervention Participants
We categorized prevention programs chosen by participants by type (diet/weight loss, exercise, or smoking cessation), and as VA-sponsored, or community-based. Among the 51% (n=91 of 177) of intervention patients who endorsed enrolling in a prevention program by 6 months, 52% selected diet or weight loss programs, 26% selected exercise programs, and 19% selected smoking cessation (3% remained uncharacterized). Overall, 55% of these participants selected VA programs and 45% selected non-VA programs.
This is the first trial to test an intervention that coupled results from an HRA with a telephone-delivered health coaching intervention designed to help participants enroll in a structured prevention program. Health coaches were successful in engaging participants in a discussion about prevention, guiding them to a program that matched their goals, which resulted in intervention participants enrolling and participating in programs to a significantly greater degree than participants not receiving coaching. Intervention participants also showed significant improvement in PAM scores, a measure that captures patients' knowledge, skills, and confidence in management of their health
The degree of prevention program enrollment observed in this study (over half enrolled, and 40% participating by six months) is much higher than reported in other studies that seek to engage primary care patients in prevention programs. In part this may be explained by integrating results from an HRA with health coaching. The health coaches guided patients to first understand their risk, the portion of it that was modifiable, and what they could do to improve their risk. They emphasized the difference between their current health state and their ideal health state, and linked that concept to participants' values and hopes for the future. Rather than coaching to change a specific health behavior selected by the participant, coaches helped participants set a specific goal to enroll in a structured program that best matched their circumstances and preferences. In this way, the intervention was able to address a wide array of risky behaviors (e.g., weight, tobacco use) with a focus on helping to motivate patients to engage in effective, structured prevention programs. This approach leverages already-available prevention programs and requires less time than would be necessary to address changing the behavior itself. Additionally, the intervention has great potential for adoption within clinical settings because it is a relatively low-resource approach. Lastly, this approach may increase reach because many patients prefer telephone coaching for its convenience and personal approach.
A strength of our study is that we designed and conducted the intervention using elements that are widely available in VA and other large healthcare systems (e.g., online HRA, telehealth coaches). VA and other health systems are working to understand how to best incorporate health risk assessment into routine primary care as well as encouraging effective prevention programs for their patients. Our current study shows that when coupled with brief health coaching, patients are more likely to enroll and participate in prevention programs compared to using an HRA alone, thereby taking patients beyond goal setting to action.
Next steps should concentrate on how best to incorporate this relatively low-resource intensive intervention into routine primary care practice which we are doing in the implementation extension.
- Sloan C, Stechuchak KM, Olsen MK, Oddone EZ, Damschroder LJ, Maciejewski ML. Short-Term VA Health Care Expenditures Following a Health Risk Assessment and Coaching Trial. Journal of general internal medicine. 2020 Jan 2.
- Nouri SS, Damschroder LJ, Olsen MK, Gierisch JM, Fagerlin A, Sanders LL, McCant F, Oddone EZ. Health Coaching Has Differential Effects on Veterans with Limited Health Literacy and Numeracy: a Secondary Analysis of ACTIVATE. Journal of general internal medicine. 2019 Apr 1; 34(4):552-558.
- Oddone EZ, Gierisch JM, Sanders LL, Fagerlin A, Sparks J, McCant F, May C, Olsen MK, Damschroder LJ. A Coaching by Telephone Intervention on Engaging Patients to Address Modifiable Cardiovascular Risk Factors: a Randomized Controlled Trial. Journal of general internal medicine. 2018 Sep 1; 33(9):1487-1494.
- Oddone EZ, Damschroder LJ, Gierisch J, Olsen M, Fagerlin A, Sanders L, Sparks J, Turner M, May C, McCant F, Curry D, White-Clark C, Juntilla K. A Coaching by Telephone Intervention for Veterans and Care Team Engagement (ACTIVATE): A study protocol for a Hybrid Type I effectiveness-implementation randomized controlled trial. Contemporary clinical trials. 2017 Apr 1; 55:1-9.