Stage-based expert systems to guide a population of primary care patients to quit smoking, eat healthier, prevent skin cancer, and receive regular mammograms
Introduction
Treating multiple health behavior risks on a population basis is one of the most promising approaches to enhancing health [1] and reducing health care costs [2]. Unfortunately, research to date has not demonstrated consistent efficacy of multiple behavior change interventions on a population basis [3], [4], [5], [6]. This study reports on a clinical trial conducted with a population of primary care patients that applied computer-based expert system reports tailored to stages of change for quitting smoking, adopting low fat diets, safer sun exposure, and preventing relapse from regular mammography screening.
To reduce chronic diseases and health care costs, national organizations like the National Institutes of Health and the Robert Wood Johnson Foundation (RWJF) have given special priority to population-based multiple behavior change research. A thorough literature review sponsored by RWJF evaluated the efficacy of multiple behavior change programs for both disease prevention and disease management [7]. No consistent support could be found for the efficacy of changing multiple health behaviors on a population basis for either disease prevention or disease management. Only one study out of 39 produced significant effects on each of three or more targeted behaviors. The research literature was limited, however, by quasi-experimental designs, action-oriented interventions designed for the small minority of populations prepared to take action on multiple behaviors, low participation rates, and a lack of the most promising behavior interventions, such as computer-generated tailored communications that can provide individualized and interactive interventions on a population basis.
A special supplement of the American Journal of Preventive Medicine reported on the RWJF project, “Addressing Multiple Behavioral Risk Factors in Primary Care.” This report highlights the need for research on multiple behavior interventions in populations of patients [8]. The prevalence and importance of multiple behavioral risks in primary care populations were documented [9], [10], [11], [12]. There was a consensus across authors and stakeholders that interventions need to be provided for multiple behaviors [13], [14]. The review of research to support such interventions, however, was limited almost entirely to studies that targeted only single behaviors [15]. Across these studies, there were multiple behaviors that were treated. The implication was that interventions designed to treat multiple behaviors in the same study should be successful. The present study tested that specific hypothesis.
There are three leading approaches to treating multiple behaviors in patient populations: (1) Counseling patients in the practice; (2) Diagnosing multiple risks and prescribing behavior medicines that patients can apply at home (such as, Internet programs); and (3) Reaching out to patient populations at home and providing home-based interventions directly to patients. The RWJF supplement focused primarily on counseling strategies in the office, but serious concerns were raised about the barriers, such as time, for implementing such practices. Interactive technology was viewed as a potential solution to such barriers [16]. The present project provided interactive technology interventions by reaching out directly to patient populations at home.
The RWJF project also recommended that given the consequences of multiple behavior risks, it is important that programs produce population impacts and not just clinical efficacy [17]. Impact equals efficacy times participation [18]. The Minnesota Heart Health Project (MHHP) generated no significant population outcomes [3] because of low participation rates in the interventions with highest efficacy [19]. With smoking, for example, nearly 90% of both the treatment and control communities saw mass media communications about quitting smoking in the past 12 months. Only 11%, however, reported that their doctors had talked with them about smoking during that time, and only about 3% participated in the most efficacious programs, which were individualized and interactive clinics, counseling, and classes [19].
In population-based trials for treating just smoking, we were able to proactively recruit 80% of a representative population of 5180 smokers [20] and 85% from an HMO population of 4653 smokers [21], [22]. Two principles produced such high participation rates: proactively reaching out to smokers rather than passively waiting for them to reach out to us for help; and letting smokers know that the program was tailored to all smokers in each stage of change and not just those ready to take action.
The first goal for population impacts on multiple behaviors is to generate high recruitment rates. Leaders of MHHP concluded that they should have focused just on a single behavior, such as smoking, rather than trying to get entire populations to participate in multiple behavior change [19]. In our first population-based trial on multiple behaviors, we were able to recruit 83.6% of a special population of parents of ninth graders who were participating in prevention programs at school [23]. At 24-month follow-up, we were able to demonstrate significant effects for each of the three targeted behaviors – smoking, diet, and sun exposure – treated at home by stage-based expert systems generated by interactive technologies [23].
The current study applied these stage-based tailored communications at home to a population of patients of primary care practices. This study also expanded the multiple behavior intervention from three to four behaviors. The objectives were to recruit a majority of these patients and to significantly reduce each of the four targeted cancer behavior risk factors: smoking, high-fat diet, sun exposure, and relapse from regular mammography screening. Almost all of the studies to date have been designed to intervene on multiple behaviors in a population rather than multiple behaviors within an individual [7]. The present study applied the same approach despite the fact that research to date indicated a low probability of success for significant improvements on four targeted behaviors within a primary care population.
Section snippets
Subjects
A large health insurance organization provided a list of patients from primary care practices. The practices were participating in a contemporaneous study testing the effects of an educational outreach intervention to increase office practice adoption of cancer prevention activities. Practices were eligible if at least one practice physician: (1) was enrolled as a provider for the collaborating health insurance organization; (2) identified his/her specialty as Family Medicine, Internal
Sample characteristics
Table 1 presents the means and standard deviations for the demographic and smoking history variables for each of the two groups. The full sample had a mean age of 44.7 years (SD = 12.7), 30.1% were male, 68.0% were married, 96.7% were non-Hispanic Caucasian, and the mean education was 14.5 years (SD = 3.2). Of the 1211 smokers (22.4% of the full sample), 30.9% were in precontemplation, 45.6% were in contemplation, and 23.5% were in preparation. Of the 3834 subjects at risk for sun exposure
Discussion
This study demonstrated in a patient population that stage-matched expert system interventions produced significant reductions in smoking, high-fat diets, high-risk sun exposure, and relapse from mammography screening when compared to the assessment only group. This study also demonstrated that proactively providing stage-matched programs for multiple health behavior changes generated high recruitment rates comparable to recruitment rates found in previous research targeting the single risk of
Acknowledgments
Grants CA 50087 and CA 27821 from the National Cancer Institute supported this research. Authors wish to acknowledge the contributions of study participants, Sally Cottrill, Gabriel Reed, and Kathryn Meier for assistance with materials development, recruitment, and follow-up in this large research project. Acknowledgement is given to Guy Natelli for expert systems programming and to Don DiCristofaro for SMS programming. The authors also greatly appreciate the collaboration and assistance
References (57)
Addressing multiple behavioral health risks in primary care: broadening the focus on health behavior change research and practice
Am. J. Prev. Med.
(2004)- et al.
Addressing multiple behavioral health risks in primary care: a synthesis of current knowledge and stakeholder dialogue sessions
Am. J. Prev. Med.
(2004) - et al.
Prevalence of multiple chronic disease risk factors: 2001 National Health Interview survey
Am. J. Prev. Med.
(2004) - et al.
Meeting recommendations for multiple health lifestyle factors: prevalence, clustering, and predictors among adolescent, adult, and senior health plan members
Am. J. Prev. Med.
(2004) - et al.
Physician screening for multiple behavior health risk factors
Am. J. Prev. Med.
(2004) - et al.
Translating what we have learned into practice: principles and hypotheses for interventions addressing multiple behaviors in primary care
Am. J. Prev. Med.
(2004) Toward a public policy agenda for addressing multiple health risk behaviors in primary care
Am. J. Prev. Med.
(2004)- et al.
Multiple behavioral risk factor interventions in primary care: summary of research evidence
Am. J. Prev. Med.
(2004) - et al.
Interactive behavior change technology: a partial solution to the competing demands of primary care
Am. J. Prev. Med.
(2004) - et al.
Beyond efficacy testing redux
Am. J. Prev. Med.
(2004)
An expert system intervention for smoking cessation
Patient Educ. Couns.
Evaluating a population-based recruitment approach and a stage-based expert system intervention for smoking cessation
Addict. Behav.
Counselor and stimulus control enhancements of a stage matched expert system for smokers in a managed care setting
Prev. Med.
An expert system intervention for smoking cessation
Addict. Behav.
Relapse situations and self-efficacy: an integrative model
Addict. Behav.
Randomized controlled community trial of the efficacy of a multicomponent stage-matched intervention to increase sun protection among beachgoers
Prev. Med.
The effect of a stage-matched and tailored intervention on repeat mammography
Am. J. Prev. Med.
Dietary applications of the stages of change model
J. Am. Diet. Assoc.
A comparison of four self-report smoking cessation outcome measures
Addict. Behav.
Program for regression analysis of repeated measures using generalized estimating equations
Comput. Methods Programs Biomed.
Evaluating primary care behavioral counseling interventions: an evidence-based approach
Am. J. Prev. Med.
Emerging research: a view from one research center
Am. J. Health Promot.
Effects of community-wide education on cardiovascular disease risk factors: the Stanford Five-City Project
J. Am. Med. Assoc.
Community education for cardiovascular disease prevention: risk factor changes in the Minnesota Heart Health Project
Am. J. Public Health
The Pawtucket Heart Health Program: community changes in cardiovascular risk factors and projected disease risk
Am. J. Public Health
Evaluating community-based collaborative mechanisms: implications for practitioners
Health Promot. Pract.
Cited by (282)
Assessing efficacy of a web-based smoking cessation tool - QuitAdvisorMD: Protocol for a practice-based, clustered, randomized control trial
2024, Contemporary Clinical Trials CommunicationsEcoSanté Lifestyle Intervention: Encourage Reflections on the Connections between Health and Environment
2023, ACM Transactions on Computer-Human InteractionHelping Skills for Counselors and Health Professionals: Building Culturally Competent Relationships
2023, Helping Skills for Counselors and Health Professionals: Building Culturally Competent RelationshipsNeuromarkers associated with two levels of physical activity in persons with chronic conditions
2023, Journal of Applied PhysiologyUnderstanding Health Determinants: Explanatory Theories for Social Epidemiology
2023, Understanding Health Determinants: Explanatory Theories for Social Epidemiology