Improving Motivation for Physical Activity in Arthritis Clinical Trial (IMPAACT)

Arthritis is a major cause of disability. Of the nearly 70 million persons in the US with arthritis and/or chronic joint symptoms, nearly 8 million are disabled because of their arthritis. Knee osteoarthritis (OA) and rheumatoid arthritis (RA) are two of its most common and disabling forms. Despite evidence that physical activity can improve functional and health related quality of life (HRQOL) outcomes and lower health care costs, the proportion of the US population engaging in the recommended amount of physical activity is low and even lower among those with arthritis. Moreover, there is considerable variation in clinicians' promotion of physical activity for arthritis clients. Care providers infrequently ask clients about their physical activity behavior and report feeling unprepared to promote physical activity. This application studies the effects of a behavioral intervention aimed at promoting physical activity, including lifestyle physical activity, on arthritis-specific and generic HRQOL outcomes. The proposed physical activity management program (PAM) is based on a chronic care model in which allied health professionals promote patient self-management activities outside of traditional physician office encounters. The program is an individualized counseling and referral intervention, conducted by physical activity managers, directed by a theory-based comprehensive assessment of individual patient barriers and strengths related to physical activity performance. The specific aims of this randomized, controlled trial of 480 clients with RA and knee OA followed for 24 months are to test the effectiveness of physical activity management combined with physician physical activity promotion (PAM group) compared to physician physical activity promotion only (control group) in improving arthritis-specific and generic HRQOL, observed measures of function, and objectively measured and self-reported physical activity levels. In addition, exploratory analyses will be done to assess whether the improvements in HRQOL and physical activity performance associated with the PAM program are mediated by increases in physical activity levels and theory-based motivational variables, respectively. This study is intended to generate feasible methods by which health care providers and health care systems can increase physical activity levels in clients with arthritis and to result in widely applicable strategies for health behavior change.

Research Design Overview This five-year study is a randomized controlled trial of 480 clients with arthritis conducted at three sites on the Northwestern University Feinberg School of Medicine campus, the Rehabilitation Institute of Chicago Arthritis Center (RIC-AC), the Northwestern Medical Faculty Foundation Rheumatology Division Ambulatory Center (NMFF-Rheum), and the NMFF General Internal Medicine Ambulatory Center (NMFF-GIM). After a 6 month startup period, two hundred forty clients with RA (120 from RIC-AC and 120 from NMFF-Rheum) and 240 clients with knee OA (120 from RIC-AC, 60 from NMFF-Rheum, and 60 from NMFF-GIM) will be recruited over a 24 month period, and each subject will be followed for 24 months. Randomization will be stratified by diagnosis, site, and disease-specific physical functional status. Half the clients will be randomized to the PAM intervention and half to the control intervention. Independent variables include demographic characteristics, arthritis severity and comorbid conditions, and other IMCHB model variables: motivation (self-regulation efficacy), cognitive appraisal (perceived competency), and affective response (mental health). Outcomes measured by blinded interviewers/assessors include self-reported and objective measures of physical activity, self-reported measures of disease specific and generic health-related quality of life, and observed measures of physical function. All independent variables and outcomes are measured at baseline and nearly all independent variables and all outcomes are measured at 3, 6, 12, 18, and 24 months of follow-up. The primary analysis tests the effectiveness of PAM compared to control groups in improving disease-specific health related quality of life (HRQOL). Secondary analyses test 1) the effectiveness of PAM compared to control groups in improving generic HRQOL, 2) the effectiveness of PAM compared to control groups in improving observed measures of function and cardiovascular fitness, 3) the effectiveness of PAM compared to control groups in increasing physical activity performance, 4) whether PAM's effectiveness in improving HRQOL is mediated by it's ability to increase physical activity and 5) whether PAM's effectiveness in improving physical activity performance is mediated by changes in theory-based motivation variables. Description and justification of the study population Given that there are more than 100 different forms of arthritis (defined here as any chronic condition affecting the joints and surrounding tissues), we have chosen clients with two well-defined and pathophysiologically distinct diseases to study, knee osteoarthritis (OA) and rheumatoid arthritis (RA). Knee OA is a prototype of a chronic degenerative arthropathy that has a high prevalence and is associated with a great amount of disability. RA is a prototype of the most common chronic inflammatory arthropathy, again associated with a high rate of disability. We expect that if the PAM program is effective in improving the physical activity and health status levels of these two distinctly different groups of arthritis clients, then it should be effective for many other, if not all, other arthritis clients. Case definitions. Classification criteria developed by the American College of Rheumatology (ACR) will be used for clients with both knee OA and RA. The criteria utilizing knee x-rays will be used to increase the specificity of the knee OA criteria. These are standard case definitions that are used routinely in the rheumatology literature and have been shown to have high sensitivity and specificity.108, 109 Selection criteria. Clients fulfilling the ACR criteria for knee OA or RA will be eligible for this trial if the meet the following criteria: 1) age 18 or older, 2) able to ambulate at least household distances (50ft), and 3) able to speak and understand English, 4) no plans for total joint replacement in the following 12 months, 5) no contraindication to physical activity due to a comorbid conditions, and 6) no plans to relocate away from the Chicago land area for at least 24 months. Methods for recruiting, enrolling, and sustaining participation. The Rehabilitation Institute of Chicago Arthritis Center (RIC-AC) and the Rheumatology and General Internal Medicine Divisions of the Northwestern Medical Faculty Foundation (NMFF-Rheum and NMFF-GIM), the practice of fulltime Northwestern University Feinberg School of Medicine faculty members, will be the sites where clients will be recruited for this trial. While on the same campus, these RIC and NMFF are physically separated from one another and have separate administrative structures. Both have more than adequate numbers of eligible clients to enroll in this study. Based on administrative data, more than 500 individual clients with RA and 700 individual clients with knee OA were seen at the RIC-AC in the past twelve months. This is an average of more than 40 RA and nearly 60 knee OA clients a month, a number which is likely to grow as the practice has been experiencing an approximate 15% growth in clients per year. More than triple these numbers of RA clients are seen at the NMFF-Rheum practice annually. Com-parable numbers of knee OA clients are seen at RIC-AC and NMFF-Rheum. To assure ready access to knee OA subjects, NMFF-GIM will also be a site for those subjects. Over 700 unique clients with explicit ICD-9 CM visit codes for knee osteoarthritis or rheumatoid arthritis were identified for NMFF-Rheum from EPICcare electronic medical records just for the months May-July 2004; approximately 200 additional clients with osteoarthritis and 25 with rheumatoid arthritis were seen at the NMFF-GIM over the same period. Several hundred clients with RA and knee OA are currently enrolled in an IRB-approved research registry which should facilitate recruitment into this study. Enrollment and sustained participation in the proposed study will be enhanced because of the proven willingness of clients in these academic practices to participate in ongoing observational studies. Dr. Leena Sharma has been following a cohort of 300 subjects with knee osteoarthritis for the past 5 years. Drs. Pamela Semanik and George Liang (an NMFF rheumatology colleague) have been able to recruit 185 and 311 RA clients respectively for their studies within one year. Based on this and other similar experiences, we expect attrition or lost to follow-up rates of less than 10% per one year and less than 20% over two years. Practice newsletters and the presence of familiar and friendly research assistants will enhance and sustain interest in ongoing research initiatives at both sites. Study interviews will be conducted at the convenience of the subject, either face to face or on the telephone using computerized telephone interview software. Institutional Review Board approval has already been obtained for collecting medical record and health status data on RA clients at the NMFF-Rheum site and physical activity and health status data on both knee OA and RA clients at the RIC-AC. Description of the consent process. Eligible clients will be informed of this study by the participating physicians of the RIC-AC and the NMFF-Rheum and NMFF-GIM practices by mail and at the time of a scheduled clinic visit. Mailings of study announcements to participants listed in the IRB-approved RA and knee OA client registries at RIC-AC, NMFF-Rheum, and NMFF-GIM. At the inception of the study, participating physicians will request retrospective ICD-9 coded visit data from RIC-AC, NMFF-Rheum, and NMFF-GIM for all clients coded with RA or adult knee OA (diagnosis codes 714.0 or 715.96, excluding clients with v codes for aftercare of a previous joint replacement) seen at the 3 sites during the previous 12 months (Month 1). These visit data contain client names, addresses and phone numbers. By the end of project month 2, all clients who were identified from the registry as having relevant prior visit codes will be mailed a study solicitation letter, with a cover letter signed by their attending clinic physician, with a consent form and stamped return envelope for signed consent or refusal to participate. This mailing will provide study investigator phone numbers for participants to obtain further information. For those willing to participate, study coordinators will be able to screen for eligibility, including information on prior or immanently scheduled knee replacement surgery, as well as the date of clients' next regularly scheduled clinic appointment (if known). Consent materials will describe the need for six no-cost study visits with transportation and meal reimbursement policies. Within two weeks of the initial mailing, study personnel will initiate telephone contact with IRB-approved registry clients who have not responded to the mailing to further explain the study and determine interest in study participation. If interested, a second mailing will be made to these clients as needed. Medical records for all registry clients who return signed informed consent will then be reviewed to determine eligibility. Subsequent clinic schedules will be examined for consented, eligible clients and patient reported MD appointment times on a continuous basis to determine each client's next upcoming scheduled physician visit. Consented clients will be contacted again by phone within about six weeks of a regularly scheduled clinic visit. Based on the date of this forthcoming physician appointment, clients will be scheduled for a baseline study visit (with maximum flexibility to arrange weekday or weekend study visits at the patient's convenience) approximately three weeks before the "index" or baseline physician appointment. This entire process will be repeated, at study months 8 and 16, to assist in meeting enrollment targets In-Person Enrollment. Clients not reachable by phone or mail, who prefer being enrolled in person, or who are not in the registry but are identified by their physicians as potential subjects at the time of an initial clinic visit, will be directly approached for participation by study coordinators at the time of a clinic visit. Written information from these physicians will be given to eligible clients to read in the waiting room upon arrival at the clinic and a short video about the study will be available for viewing before or after the visit. During the clinical visit the physician will discuss the opportunity for the client to participate in this study of physical activity and health status and factors that are associated with good and poor health status and will answer any questions that arise. If the client agrees to participate, a research assistant will further explain the study in detail, obtain written informed consent, and schedule an in person and telephone interview to collect study data. If a client consents, and is deemed eligible after a subsequent medical record review, the date of their next regularly scheduled follow-up physician visit will be used to determine the timing of their initial baseline study assessment visit. Non-eligible clients will be thanked for their time and interest, and added to the registry for future study notification, if desired. The Pre-Randomization (Baseline) Study Assessment Visit (See Figure 2) At the baseline study assessment visit, a study coordinator will fit all participating clients with the Actigraph accelerometer, provide instructions and packaging for mailed return, perform the observed physical function test battery, and schedule the first telephone interview one week later (about two weeks prior to the index clinic visit) at clients' convenience. This will enable telephone interview data (including the HRQOL and physical activity data) to refer the same week that the Actigraph data are being recorded. Clients will also be advised of prospective dates for five subsequent study assessment visits, the first two at three month intervals, the final three at six month intervals, which will be entirely unrelated to any subsequent MD visits. Randomization and the Index Physician Visit (See Figure 2) Randomization will be stratified by diagnosis (RA vs. knee OA), by site (RIC-AC vs. NMFF-Rheum for subjects with RA, RA-AC vs. NMFF-Rheum vs. NMFF-GIM for subjects with knee OA), and by level of disease- specific physical function (above or below specified cutpoints for the HAQ Disability Scale for RA subjects and the WOMAC-Physical Function Scale for knee OA subjects; the cutpoints set as the median values from RA and knee OA subjects who participated in previous clinical studies of RA and knee OA at RIC and NMFF). Consecutive clients within each strata will be randomly assigned to either PAM or control intervention. Randomization will occur to allow study personnel to notify each participant about their group assignment well before the index physician visit at which all subjects will receive MD physical activity counseling (full description is below). If selected into the PAM group, clients will be requested to allow additional time after their existing physician appointment for an initial, no cost, expense-reimbursed visit with the physical activity manager. The description of the PAM program is below. If spending additional time after the index physician visit concludes is inconvenient for the client, a separate PAM study visit for experimental group clients will be scheduled as soon as possible thereafter. Other than potential interaction with physicians about referrals or other experimental group activities, the PAM study assessment visits will remain completely independent of study clients' medical care, which will proceed in a usual and customary manner. This is analogous to chronic disease management in other settings. Physician physical activity promotion to be given to all study subjects The physician physical activity promotion intervention, which is to be given at the index physician visit (Figure 2) is designed to be accomplished in less than 5 minutes, and includes a) determining whether the client is meeting the CDC physical activity recommendations, b) encouraging the client to work toward or maintain this level of physical activity, and 3) eliciting and discussing the client's questions and/or concerns about physical activity. There are two reasons for giving this intervention to both groups. First, it is standard practice in some, but not all, practices participating in this trial and second, it is postulated that PAM will only be successful if MD's support its goals. The Physical Activity Management (PAM) Program (See also Appendix) The physical activity management program draws on principles of Cox's model described earlier. Because the potential for positive patient health outcomes increases as the provider intervention/interaction is better fitted to the uniqueness of the patient (background, cognitive, affective, and motivational manifestations), the Physical Activity Management program will be conducted using a standard intervention protocol guided by the results from the ACTA assessment. The ACTA interview and self-score form provide opportunities for the client to reflect upon his/her physical activity and to identify the supports and barriers to participation. The client works with the physical activity manager to identify and select goals that s/he personally finds motivating. Together, the client and physical activity manager negotiate the strategies that will be needed to help the client achieve their self-identified goals. This may involve the physical activity manager: providing information about physical activity, community resources that are available to the client making referrals as necessary to other care providers helping the client to identify day-to-day activities or ways the client can incorporate more activity in their life allowing the client the ultimate decisional control. The intervention was revised following patient feedback and clinician experience with the pilot programs. The physical activity management program now consists of 6 components, in part based on work by Cimprich110: a verbal explanation of the purpose of the physical activity management program; identification of barriers and supports to physical activity and physical activity goals using the ACTA; guidance in developing an action plan, selecting ways to be more active and a plan for what to do if the client has difficulty increasing their activity; establishing an agreement in writing to increase their physical activity each week; a strategy of re-cording progress; and a verbal explanation of the plan for future meetings with the physical activity manager. The IMCHB recognizes the role of choice and self-determination in health behavior, the ability of the clinician's interaction style and intervention approach to support or discourage health behavior, and the dynamic impact of the patient's singularity profile on health outcomes 111. To optimize this feature, the client will negotiate with the physical activity manager a contract that outlines goals, commitment to increase activity, what to do if things go wrong, use of a pedometer, three monthly calendars on which to record daily steps and which includes a list of activities the client can try to increase physical activity, and a copy of an activity pyramid with lifestyle physical activity suggestions. The client is seen or contacted for physical activity management at base-line, 3, 6, 12, 18, and 24 months at which time the ACTA will be repeated. The client's physician(s) will receive a physical activity management summary report at these times. The client and physical activity manager will negotiate other contacts as may be indicated by the client's needs. 112, 113 An autonomy-supportive provider and health care environment are key intervention strategies to assist in initiating and maintaining behavior change. A detailed description of the protocol can be found in the Appendix. Self-determination theory distinguishes between autonomous and controlled behavior regulation. Autonomous behaviors are initiated and sustained to the extent that people experience a true sense of volition and choice, and act because of the personal importance of the behavior. Controlled behaviors are performed because the individual feels pressured, either by external or intrapsychic forces 114.The autonomy supportive physical activity manager encourages the client to maintain decisional control of the process, choosing activities at which they feel they are most likely to succeed but also in assisting in development of a plan for what the client can do if they have difficulty increasing or maintaining their physical activity. By continuing to maintain contact with the client over a 2 year period, to revise plans and goal achievement as necessary, this intervention recognizes the need to begin with primarily volitional activities i.e., consciously chosen and implemented activities until they are integrated into an habitual routine in the clients everyday life. This autonomy-supportive intervention also is informed by the chronic disease management model of Wagner and colleagues by utilizing protocols, placing more clinical responsibility on an allied health professional, emphasizing patient education and increased access to clinical and community expertise, and by giving the physician greater availability of clinical information relevant to the client's physical activity. Follow-up Study Assessment Visits (Figure 2) All study clients will be scheduled for five subsequent study visits at 3, 6, 12, 18, and 24 months after the baseline assessment, at which Actigraph fittings, observed physical function tests, and subsequent telephone interview follow-up within a week will be provided for both experimental and control subjects. Clients who do not continue to be seen at the study clinics for clinical care will be encouraged to remain in the study. Reimbursement for parking and meals will be made at each visit. Randomization will occur to allow study personnel to notify each participant about their group assignment well before the index physician visit at which all subjects will receive MD physical activity counseling (full description is below). If selected into the PAM group, clients will be requested to allow additional time after their existing physician appointment for an initial, no cost, expense-reimbursed visit with the physical activity manager. The description of the PAM program is below. If spending additional time after the index physician visit concludes is inconvenient for the client, a separate PAM study visit for experimental group clients will be scheduled as soon as possible thereafter. Other than potential interaction with physicians about referrals or other experimental group activities, the PAM study assessment visits will remain completely independent of study clients' medical care, which will proceed in a usual and customary manner. This is analogous to chronic disease management in other settings. Physician physical activity promotion to be given to all study subjects The physician physical activity promotion intervention, which is to be given at the index physician visit (Figure 2) is designed to be accomplished in less than 5 minutes, and includes a) determining whether the client is meeting the CDC physical activity recommendations, b) encouraging the client to work toward or maintain this level of physical activity, and 3) eliciting and discussing the client's questions and/or concerns about physical activity. There are two reasons for giving this intervention to both groups. First, it is standard practice in some, but not all, practices participating in this trial and second, it is postulated that PAM will only be successful if MD's support its goals. The Physical Activity Management (PAM) Program (See also Appendix) The physical activity management program draws on principles of Cox's model described earlier. Because the potential for positive patient health outcomes increases as the provider intervention/interaction is better fitted to the uniqueness of the patient (background, cognitive, affective, and motivational manifestations), the Physical Activity Management program will be conducted using a standard intervention protocol guided by the results from the ACTA assessment. The ACTA interview and self-score form provide opportunities for the client to reflect upon his/her physical activity and to identify the supports and barriers to participation. The client works with the physical activity manager to identify and select goals that s/he personally finds motivating. Together, the client and physical activity manager negotiate the strategies that will be needed to help the client achieve their self-identified goals. This may involve the physical activity manager: providing information about physical activity, community resources that are available to the client making referrals as necessary to other care providers helping the client to identify day-to-day activities or ways the client can incorporate more activity in their life allowing the client the ultimate decisional control. The intervention was revised following patient feedback and clinician experience with the pilot programs. The physical activity management program now consists of 6 components, in part based on work by Cimprich110: a verbal explanation of the purpose of the physical activity management program; identification of barriers and supports to physical activity and physical activity goals using the ACTA; guidance in developing an action plan, selecting ways to be more active and a plan for what to do if the client has difficulty increasing their activity; establishing an agreement in writing to increase their physical activity each week; a strategy of re-cording progress; and a verbal explanation of the plan for future meetings with the physical activity manager. The IMCHB recognizes the role of choice and self-determination in health behavior, the ability of the clinician's interaction style and intervention approach to support or discourage health behavior, and the dynamic impact of the patient's singularity profile on health outcomes 111. To optimize this feature, the client will negotiate with the physical activity manager a contract that outlines goals, commitment to increase activity, what to do if things go wrong, use of a pedometer, three monthly calendars on which to record daily steps and which includes a list of activities the client can try to increase physical activity, and a copy of an activity pyramid with lifestyle physical activity suggestions. The client is seen or contacted for physical activity management at base-line, 3, 6, 12, 18, and 24 months at which time the ACTA will be repeated. The client's physician(s) will receive a physical activity management summary report at these times. The client and physical activity manager will negotiate other contacts as may be indicated by the client's needs. 112, 113 An autonomy-supportive provider and health care environment are key intervention strategies to assist in initiating and maintaining behavior change. A detailed description of the protocol can be found in the Appendix. Self-determination theory distinguishes between autonomous and controlled behavior regulation. Autonomous behaviors are initiated and sustained to the extent that people experience a true sense of volition and choice, and act because of the personal importance of the behavior. Controlled behaviors are performed because the individual feels pressured, either by external or intrapsychic forces 114.The autonomy supportive physical activity manager encourages the client to maintain decisional control of the process, choosing activities at which they feel they are most likely to succeed but also in assisting in development of a plan for what the client can do if they have difficulty increasing or maintaining their physical activity. By continuing to maintain contact with the client over a 2 year period, to revise plans and goal achievement as necessary, this intervention recognizes the need to begin with primarily volitional activities i.e., consciously chosen and implemented activities until they are integrated into an habitual routine in the clients everyday life. This autonomy-supportive intervention also is informed by the chronic disease management model of Wagner and colleagues by utilizing protocols, placing more clinical responsibility on an allied health professional, emphasizing patient education and increased access to clinical and community expertise, and by giving the physician greater availability of clinical information relevant to the client's physical activity. Follow-up Study Assessment Visits (Figure 2) All study clients will be scheduled for five subsequent study visits at 3, 6, 12, 18, and 24 months after the baseline assessment, at which Actigraph fittings, observed physical function tests, and subsequent telephone interview follow-up within a week will be provided for both experimental and control subjects. Clients who do not continue to be seen at the study clinics for clinical care will be encouraged to remain in the study. Reimbursement for parking and meals will be made at each visit.

Participants not in the PAM group will be in the PC (physician counseling) group and will receive advice on lifestyle physical activity from their rheumatologist or primary care physician.

Lifestyle Physical Activity Promotion as given by Physical Activity Advocates. Advocates will meet with participants at baseline and every 3 months throughout the study to promote increase in lifestyle physical activity reached through goal setting and Motivational Interviewing (MI)

Inclusion Criteria: 18 yrs. or older ACR criteria for KOA or RA Able to speak and read English Able to ambulate at least household distances Exclusion Criteria: Planned total joint replacement surgery in the next 2 years Contraindication to physical activity due to co-morbid condition Unable to perform basic self-care (functional class IV) or walk household distances Planned Relocation away from Chicago area in next two years

Lee J, Dunlop D, Ehrlich-Jones L, Semanik P, Song J, Manheim L, Chang RW. Public health impact of risk factors for physical inactivity in adults with rheumatoid arthritis. Arthritis Care Res (Hoboken). 2012 Apr;64(4):488-93. doi: 10.1002/acr.21582.

Ehrlich-Jones L, Lee J, Semanik P, Cox C, Dunlop D, Chang RW. Relationship between beliefs, motivation, and worries about physical activity and physical activity participation in persons with rheumatoid arthritis. Arthritis Care Res (Hoboken). 2011 Dec;63(12):1700-5. doi: 10.1002/acr.20616.