community

Review and Special Articles Evaluating Primary Care Behavioral Counseling Interventions An Evidence-Based Approach Evelyn P. Whitlock, MD, MPH, C. Tracy Orleans, PhD, Nola Pender, PhD, RN, FAAN, Janet Allan, RN, PhD, CS Overview: Risky behaviors are a leading cause of preventable morbidity and mortality, yet behavioral counseling interventions to address them are underutilized in healthcare settings. Re- search on such interventions has grown steadily, but the systematic review of this research is complicated by wide variations in the organization, content, and delivery of behavioral interventions and the lack of a consistent language and framework to describe these differences. The Counseling and Behavioral Interventions Work Group of the United States Preventive Services Task Force (USPSTF) was convened to address adapting existing USPSTF methods to issues and challenges raised by behavioral counseling intervention topical reviews. The systematic review of behavioral counseling interventions seeks to establish whether such interventions addressing individual behaviors improve health outcomes. Few studies directly address this question, so evidence addressing whether changing individual behavior improves health outcomes and whether behavioral counseling interventions in clinical settings help people change those behaviors must be linked. To illustrate this process, we present two separate analytic frameworks derived from screening topic tools that we developed to guide USPSTF behavioral topic reviews. No simple empirically validated model captures the broad range of intervention compo- nents across risk behaviors, but the Five A’s construct—assess, advise, agree, assist, and arrange—adapted from tobacco cessation interventions in clinical care provides a workable framework to report behavioral counseling intervention review findings. We illustrate the use of this framework with general findings from recent behavioral counseling intervention studies. Readers are referred to the USPSTF (www.ahrq.gov/clinic/prevenix.htm or 1-800-358-9295) for systematic evidence reviews and USPSTF recommendations based on these reviews for specific behaviors. Medical Subject Headings (MeSH): behavioral medicine, counseling, health behavior, health promotion, patient education, preventive health services, primary health care (Am J Prev Med 2002;22(4):267–284) © 2002 American Journal of Preventive Medicine Introduction In 1998, the Agency for Healthcare Research andQuality (AHRQ) reconvened the U.S. PreventiveServices Task Force (USPSTF) to update its recom- mendations for clinical preventive services. This Task Force represents primary care disciplines (nursing, pediatrics, family practice, internal medicine, and ob- stetrics/gynecology), preventive medicine, and behav- ioral medicine. Two evidence-based practice centers (EPCs)—Oregon Health & Science University and the Research Triangle Institute/University of North Caro- lina—were contracted to prepare systematic evidence reviews that the Task Force uses in developing its recommendations for preventive care. Although the USPSTF evidence-based methods are widely applicable throughout medicine, to date they have been used primarily to assess services, such as preventive screen- ing, rather than those requiring behavioral counsel- ing.1,2 The current Task Force recognized a twofold need: (1) to expand its evidence-based approach to better assess behavioral counseling interventions, and (2) to formulate practical communication strategies for From the Oregon Health & Science University Evidence-Based Prac- tice Center, Kaiser Permanente/CHR (Whitlock), Portland, Oregon; U.S. Preventive Services Task Force (Orleans), Princeton, New Jersey; U.S. Preventive Services Task Force (Pender), Ann Arbor, Michigan; and U.S. Preventive Services Task Force (Allan), San Antonio, Texas Address correspondence and reprint requests to: Evelyn P. Whit- lock, MD, MPH, Kaiser Permanente/CHR, 3800 North Interstate Avenue, Portland, Oregon 97227-1098. E-mail: Evelyn.whitlock@ kpchr.org. The full text of this article is available via AJPM Online at www.ajpm-online.net. 267Am J Prev Med 2002;22(4) 0749-3797/02/$–see front matter © 2002 American Journal of Preventive Medicine • Published by Elsevier Science Inc. PII S0749-3797(02)00415-4 describing services that are effective in changing behavior. The Counseling and Behavioral Interventions Work Group of the USPSTF adapted the USPSTF generic screening analytic framework, which guides systematic reviews, to address behavioral topics more specifically, and it has promoted a consistent organizational con- struct for describing behavioral counseling interven- tions. Clinicians are referred to current products of the USPSTF (www.ahrq.gov/clinic/or 1-800-358-9295) for systematic evidence reviews of specific behavioral coun- seling topics and related USPSTF evidence-based rec- ommendations and clinical considerations beyond the scope of this paper. This paper has three purposes: 1. To promote a broader appreciation of the importance of behavioral counseling in- terventions in clinical care and the con- text for their delivery. 2. To describe the generic analytic frame- works developed to guide the systematic review of behavioral counseling topics for the third USPSTF. 3. To detail the practical organizational con- struct (the Five A’s) adopted by the USPSTF to describe intervention research more consistently in order to foster its application in clinical settings. Background Healthy People 20103 sets two major goals for the United States: (1) to increase quality and years of healthy life, and (2) to eliminate health disparities among different segments of the population. The next decade offers unprecedented opportunities for healthcare systems and providers to address these goals by promoting healthy lifestyles among the diverse populations they serve and by adopting policies that will institutionalize preventive services. Changing the health behaviors of Americans has the greatest potential of any current approach for decreas- ing morbidity and mortality and for improving the quality of life across diverse populations.4 In their landmark paper, McGinnis and Foege5 linked 50% of the mortality in the United States from the ten leading causes of death to lifestyle-related behaviors, such as tobacco use, poor dietary habits and inactivity, alco- hol misuse, illicit drug use, and risky sexual practices. These behaviors remain problematic in today’s soci- ety despite having been previously targeted for im- provement.6 Thus, the U.S. Department of Health and Human Services has designated five lifestyle factors as Healthy People 2010 3 health indicators by which to track progress in improving the health of the nation over the next decade (Table 1). Improv- ing health behaviors is an important approach to health disparities, because those who are economi- cally and/or socially disadvantaged, including those in low-income ethnic/racial minority groups, dispro- portionately bear the prevalence of risky health behaviors and the burden of preventable morbidity and mortality.7 The unabated impact of health-damaging behaviors among Americans makes it imperative that healthcare providers and healthcare systems seriously consider these behavioral issues and accept the challenge of routinely providing quality behavioral counseling inter- ventions where proven effective. The 1996 edition of the Guide to Clinical Preventive Services by the USPSTF concluded: “Effective interventions that address per- sonal health practices … [for] … primary prevention … hold greater promise for improving overall health than many secondary preventive mea- sures, such as routine screening for early disease. Therefore, clinician counseling that leads to improved personal health practices may be more valuable than conventional clinical activities, such as diagnostic testing.”1 Nevertheless, rates of behavioral counseling intervention by pediatricians, nurse practitio- ners, obstetrician/gynecologists, internists, and family physicians for the priority behaviors dis- cussed above still fall far below national targets.3,8,9 In fact, gaps in the delivery of clinical preventive services are greater for behavioral counseling than for screen- ing or chemoprophylaxis.10 This stems in part from the relative paucity of good research evidence to support the behavioral counseling intervention recommenda- tions in the last Guide to Clinical Preventive Services.1 The quality and quantity of good research evidence for the effectiveness of behavioral counseling inter- ventions are increasing. Brief interventions inte- grated into routine primary care can effectively ad- dress the most common and important risk behaviors.11–22 The strongest evidence for the effi- cacy of primary care behavior-change interventions comes from tobacco-cessation research11,12,14,15,19 and, to a lesser extent, problem drinking.11,16 –19,21,22 Accumulating evidence also shows the effectiveness of similar interventions for other behaviors.11,19,20 These interventions often provide more than brief clinician advice. Effective interventions typically in- volve behavioral counseling techniques and use of other resources to assist patients in undertaking advised behavior changes.12,19 For example, interven- tion adjuncts to brief clinician advice may involve a broader set of healthcare team members (e.g., nurses, other office staff, health educators, and pharmacists), a number of complementary communication channels (e.g., telephone counseling,22,23 video or computer- assisted interventions,24–26 self-help guides,27 and tai- lored mailings28), and multiple contacts with the patient.12,14,19,29 See related Commentary on page 320. 268 American Journal of Preventive Medicine, Volume 22, Number 4 Rationale for Behavioral Counseling Interventions in Clinical Care Healthcare providers and their staff play a unique and important role in motivating and assisting patients’ healthy behavior changes. Patients report that primary care clinicians are expected sources of preventive health information and recommendations for pa- tients.30 For instance, in a recent survey, the vast majority (92% to 98%) of adult members of health maintenance organizations indicated that they ex- pected advice and help from the healthcare system in key behaviors, such as diet, exercise, and substance use.31 Similarly, healthcare providers generally accept32 and value their role in motivating health promotion and disease prevention.33,34 Healthcare systems are natural settings for interven- tions to improve health behaviors for many individuals, because repeated contacts typically occur over a num- ber of years. Interventions to help patients change unhealthy behaviors, like treatments for patients with chronic disease, often require repetition for modest effects over time. Continuity of care offers opportuni- ties to sustain individual motivation, assess progress, provide feedback, and adjust behavior change plans.35 In fact, most clinicians have multiple opportunities to intervene with patients on matters related to health- behavior change: patients aged �15 years average 2.4 visits per person annually to office-based physicians, and those aged �15 years average 1.6 to 6.3 visits per year, with visit frequency increasing with age.36 More- over, 93% of children and youth and 84% of adults aged�18 years have a specific source of ongoing health care.3 Not surprisingly, people with a usual source of health care are more likely than those without to receive a variety of clinical preventive services.3 The healthcare setting is not the only setting for approaches to support healthy behaviors. The Guide to Community Preventive Services features evidence-based recommendations from the Task Force on Community Preventive Services for population-based interventions. Those recommendations include policy or environ- mental changes or individual and group interventions outside the clinical setting intended to change risky behaviors; reduce specific diseases, injuries and impair- ments; and address environmental and ecosystem chal- lenges.37 These preventive policies and approaches complement the individually focused interventions that the USPSTF addresses. Objectives and Scope of Behavioral Counseling Interventions Behavioral counseling interventions in clinical care are those activities delivered by primary care clinicians and related healthcare staff to assist patients in adopting, changing, or maintaining behaviors proven to affect Table 1. Healthy People 20103 leading health indicatorsa Health indicator 1997 Baseline 2010 Goals Tobacco use (%) Cigarette smoking, adults 24 12 American Indian/Alaskan Native 34 12 Family income, poor level 34 12 Current tobacco use by youth (past 30 days) 43 21 Smoking cessation attempts Adults 43 75 Pregnant women 12 30 Adolescents (grades 9–12) 73 84 Overweight and obesity (%) Proportion of adults at healthy weightb 42 60 Mexican Americans 30 60 Lower income (�130% poverty threshold) 29 15 Obesityc in adults (�20 years) 23 15 Overweight/obesity in children and teens (6–19 years)d 11 5 Physical activity (%) No leisure-time physical activity (�18 years) 40 20 American Indians/Alaskan Native, African American, or Hispanic 46–54 20 Moderate physical activitye Adults (�18 years) 15 30 Adolescents (grades 9–12) 20 30 Substance abuse Proportion of adults exceeding low-risk drinking guidelines (%)f Females 72 50 Males 74 50 Alcohol-related auto deaths 6.1/100,000 4/100,000 American Indian or Alaska Native 19.2/100,000 4/100,000 Persons aged 15–24 years 11–7/100,000 4/100,000 High school seniors never using alcohol (%) 19 29 Binge drinking (%) Adolescents (12–17 years) 8.3 3 High school seniors 32 11 College students 39 20 Adults 16 6 Youth (12–17) using marijuana in the last 30 days (%) 9.4 0.7 High school seniors never using illicit drugs (%) 46 56 Responsible sexual behavior (%) Unmarried females (18–44 years) whose partners used condoms 23 50 Teens abstain from sex or use condoms 85 95 a Other leading health indicators include mental health, injury and vio- lence, environmental quality, immunizations, and access to health care. b 18.5�BMI�25. c BMI of �30. d �95th percentile of gender- and age-specific BMI from year 2000 U.S. growth charts. e Moderate activity of 30 minutes a day, �5 days a week. f Males �14 drinks/week or �4 drinks/occasion; females �7 drinks/ week or �3 drinks/occasion. Source: From Healthy People 20103 Adapted from public domain docu- ment; also available online at: http://www.health.gov/healthypeople. BMI, body mass index. Am J Prev Med 2002;22(4) 269 health outcomes and health status. Common health- promoting behaviors include smoking cessation, healthy diet, regular physical activity, appropriate alco- hol use, and responsible use of contraceptives. Behavioral counseling interventions occur all or in part during routine primary care and may involve both visit-based and outside intervention components. For instance, assessment of behavioral health risks may occur at the time of enrollment in a health plan or at the time of a clinical visit. Behavioral counseling may take place in routine primary care visits and/or through telephone contacts or personalized mailings of self-help guides or materials. Referral to more intensive clinics in the community also may be included. While the USPSTF primarily evaluates interventions that in- volve clinicians as part of routine primary care, USPSTF liaisons assigned to a particular behavioral topic define the scope of clinical intervention approaches reviewed for any given topic, such as problem drinking or physical activity. Behavioral counseling interventions differ from screening interventions in several important ways that affect the ease and likelihood of their being delivered. Behavioral counseling interventions address complex behaviors that are integral to daily living; they vary in intensity and scope from patient to patient; they re- quire repeated action by both patient and clinician, modified over time, to achieve health improvement; and they are strongly influenced by multiple contexts (family, peers, worksite, school, and community). Fur- ther, “counseling” is a broadly used but imprecise term that covers a wide array of preventive and therapeutic activities, from mental health or marital therapy to the provision of health education and behavior change support. Thus, we have chosen to use the term “behav- ioral counseling interventions” to describe the range of personal counseling and related behavior-change inter- ventions that are effectively employed in primary care to help patients change health-related behaviors. As with its use in other contexts, “counseling” here de- notes a cooperative mode of work demanding active participation from both patient and clinician that aims to facilitate the patient’s independent initiative and ability to cope.38 Engaging patients actively in the self-management practices needed to change and main- tain healthy behaviors is a central component of effec- tive behavioral counseling interventions. Theories and Models of Behavior Change Behavior change theories and models from the social and behavioral sciences explain the biological, cogni- tive, behavioral, and psychosocial/environmental de- terminants of health-related behaviors. Thus, they also define interventions to produce changes in knowledge, attitudes, motivations, self-confidence, skills, and social supports required for behavior change and mainte- nance.39 The application of relevant theoretical models to behavioral counseling interventions is an important contribution to strengthening health research in this area.40 A literature review of 1174 articles evaluating health behavior, education, and promotion interven- tions published between 1992 and 1994 found that 44.8% of these were explicitly theory based.41 Six theories and models addressing determinants of health- behavior change at the intrapersonal, interpersonal, and environmental levels (Table 2) and two cross- theoretical key constructs/theories were most com- monly cited in this research. Promising, if not substan- tial, empirical evidence supports the validity of all eight theories in predicting or changing health behavior.41 In addition to those listed in Table 2, self-efficacy and social network/support were the other two most com- monly cited constructs in the current literature. Self- efficacy is an individual’s level of confidence in his or her own skills and persistence to accomplish a desired goal and predicts future behavior across a wide variety of lifestyle risk factors.42 Social networks are a person- centered web of social relationships.43 These relation- ships provide social support that can assist the individ- ual through “stress buffering” and other mechanisms.43 These theories focus on diverse, interacting levels of influence on an individual’s behavior. On the intraper- sonal level, multiple internal factors influence an indi- vidual’s behavioral choices and actions, and there is considerable variability in these factors among individ- uals with the same objective health behavior. For exam- ple, in the stages-of-change/transtheoretical model (Table 2), behavioral change is thought of as an ongoing process with multiple stages that often in- cludes relapse and recycling into renewed efforts to change.44 On the interpersonal level, individual behav- ioral choices occur in a context that includes the influence of social and environmental conditions in the family and larger community.41,45 Behavioral influences operate within a broadly con- ceptualized ecologic paradigm emphasizing that a dy- namic interaction between functional levels—intraper- sonal, interpersonal, and the physical environment— continues over an individual’s lifetim