IIR 07-154
Expert System-Based Feedback in Sedentary Overweight Veterans
Mary Ann Sevick, ScD MS VA Pittsburgh Healthcare System University Drive Division, Pittsburgh, PA Pittsburgh, PA Funding Period: October 2008 - June 2013 Portfolio Assignment: Mobility, Activity, and Function |
BACKGROUND/RATIONALE:
Physical activity (PA) has many health benefits. Primary care settings are an ideal location for PA counseling because the majority of US adults see their primary care provider (PCP) on an annual basis and are ready to discuss with them topics of prevention and health promotion. However, recent reviews of PA counseling in primary care suggest that there are limited benefits from such interventions, benefits are not sustained, and PA interventions may be more effective if delivered outside the primary care setting (e.g., via telephone, mail, or in the community). Using computerized expert systems to deliver motivationally-tailored counseling messages by mail may be an effective and economical approach for engaging clinical populations in PA. OBJECTIVE(S): The objective of this study was to examine the effectiveness and cost-utility of a computer-based expert system PA print intervention delivered by mail to Veterans receiving their primary care at the VA Pittsburgh Healthcare System (VAPHS). The primary aim was to determine the impact of the intervention on PA. The secondary aims of the study were to estimate the impact of the intervention on biological measures (e.g., weight, serum lipids, fasting glucose and HbA1c), quality of life, and cost-per-Quality Adjusted Life Year (QALY) at 12 months. The research hypotheses were that Veterans randomized to the intervention group, compared to the attention-control group, would be more likely to achieve at least 150 minutes per week of PA and would devote more time to at least light-intensity PA at 6 and 12 months. METHODS: The study was designed as a prospective randomized controlled trial; 232 Veterans from VAPHS were randomized to intervention or attention-control groups after informed consent was obtained and baseline measures were completed. Participants had to have a self-reported physically-inactive lifestyle and a BMI >= 25.0 kg/m2. Those randomized to the intervention group received a 12-month intervention that involved completion of 9 PA surveys used in a computer-based expert system to generate individually-tailored feedback messages. Messages, which were based on the Stages of Motivational Readiness model and Social Cognitive Theory, were used to encourage Veterans to increase their level of PA. Those randomized to the attention control group met with a health educator who focused on a healthy lifestyle, with brief mention of increasing PA, followed by additional attention-control mailings regarding healthy lifestyle tips other than PA. Both groups received 12 months of routine primary care at VAPHS. PA was measured at baseline, 6 months, and 12 months using the Community Healthy Activities Model Program for Seniors (CHAMPS) questionnaire and a 7-day accelerometer measurement. Health-related quality of life was assessed at the same timepoints using the Medical Outcomes Short Form-36 Health-Status Survey v2 (SF-36). Assessments of effectiveness occurred at baseline, 6 months (89.7% response rate), and 12 months (87.5% response rate). Conditional logistic regression models were used to examine the direct relationship between treatment assignment and outcomes at 6 and 12 months. Cost analyses were performed from a payer perspective. FINDINGS/RESULTS: There were no statistically significant differences between the two groups at baseline, with the exception of the controls having significantly better perceived quality of life, as measured by the SF-36 (p<0.05). For most of the defined outcomes, no differences between the two groups were found. However, 4 accelerometer-based measurements were statistically significant (or borderline significant) between treatment groups. At both 6 and 12 months, the control group spent more minutes per day in sedentary (i.e., physically inactive) activity than the intervention group (6 mos: 514 versus 461, p=0.02; 12 mos: 516 vs. 464, p=0.054). The difference between minutes of activity from baseline to 6 months was also different for the two groups. The intervention group decreased the minutes per day spent in sedentary behavior, while the control group increased (-16 vs. 21, p=n.s.). The intervention group also slightly decreased minutes spent in at least light-intensity activity, while the control group decreased more dramatically (-3 vs. -32, p=0.03). Finally, the intervention group slightly increased minutes per day in at least moderate-intensity activity, with a corresponding decrease in the control group (1 vs. -7, p=0.02). Using conditional logistic regression: relative to the control group, participants in the intervention group had an odds ratio of 1.54 (95% CI: 0.56-4.23; p=0.40) of meeting the US Surgeon General's recommendations of 150 minutes/week (CHAMPS) of at least moderate-intensity PA at 6 months, and a significantly increased odds ratio of 2.86 (95% CI: 1.03-7.95; p=0.04) of meeting this recommendation at 12 months. A similar model using an accelerometer showed an increased odds ratio of the intervention group participating in at least moderate-intensity PA at both 6 and 12 months; however, the difference was not statistically significant. A negative binomial regression with bootstrap standard errors analysis revealed that the intervention group had significantly lower utilization rates than the control group in months 1-6 for outpatient visits (all cause), non-podiatry outpatient visits, and emergency room visits, but not for overnight hospitalizations. None of these were significantly different between 7 and 12 months, or overall (i.e., 1-12 months aggregated). Examining costs, our estimates suggest that the intervention cost was $298.74/intervention-group Veteran and $143.20/control-group Veteran. In comparison to the control group, an additional $155.54 per Veteran was incurred. No cost-utility analysis was performed as no improvements in SF-36 scores or utility gains were observed over the 12-month study period. IMPACT: The finding of reduction in sedentary activity is clinically significant in that several lines of evidence now suggest a relationship between sedentary behavior and the development of cardiovascular disease risk factors, chronic disease, and mortality. The smaller decline in at least light-intensity PA that we observed in the intervention group is clinically important, as recent studies show PA decline can result in marked short-term changes, including increased intra-abdominal fat, decreased aerobic fitness, and impaired metabolic factors. Increased moderate-intensity PA may contribute to prevention and/or treatment of conditions such as muscle atrophy, insulin resistance, increased fat mass, diabetes, cardiovascular disease, hypertension, colon cancer, osteoarthritis, and osteoporosis. This study shows the extent to which a modestly-priced, computer-based expert system counseling intervention is effective for engaging a broadly-defined clinical population of Veterans in increasing physical activity and reducing sedentary behavior. (The type and number of comorbid conditions in these primary care Veterans was extensive, and likely played a role in the small effect sizes observed.) External Links for this ProjectDimensions for VADimensions 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 PUBLICATIONS:Journal Articles
DRA:
Aging, Older Veterans' Health and Care, Health Systems Science
DRE: Prevention, Treatment - Efficacy/Effectiveness Clinical Trial Keywords: Cost, Exercise, Obesity MeSH Terms: none |