Pre-visit Assessment of Patient Agendas & Their Discussion in a Clinic Visit: An RCT Academy of Health - 2009 Betty Chewning PhD1, Betsy Sleath PhD2, Carolyn Bell MD, Kevin McKown MD, Morris Weinberger PhD, Brenda and Bob Devellis PhD, Dave Kreling PhD, Rich Van Koningsveld MS, Dale Wilson, MA Funded by National Institute of Aging, NIH AG-18362 1University of Wisconsin School of Pharmacy & Sonderegger Research Center 2University of North Carolina, School of Pharmacy & Sheps Center Significance Unvoiced patient (Pt) agendas & incomplete attention to agendas has been documented widely Particularly for psychosocial issues (Greene et al, 1989) Older adults may be less active (Kaplan et al, 1995). Goal: To increase MD attention to Pt health priorities in way that cues Pt & MD to discuss in midst of increasing time pressures Goals Today: Present results of a computer assisted intervention (CAI) to increase discussion of Pt priorities in rheumatology clinic visits Discuss implications for future research and health care organizations Study Design Longitudinal, RCT (baseline, 6-mo, 12-mo) After baseline Pts randomized Within MD for computer assessment at 6-mo & 12-mo visits (C) Control group had CAI lifestyle assessment (E) Experimental group had CAI priority assessment of top 3 health priorities to improve 30 days Pts and MD’s received summary printouts for both Study Design (C) Printout listed lifestyle status with respect to: Veggies, fruit, seatbelt, tobacco, alcohol use. (E) Printout listed top 3 priorities for improvement (E) Pt Agendas – AIMS2 (Meenan et al, 1990) Computer asked: In the next month what 3 areas of your health would you most like to improve? (subscales within AIMS2) Mobility Hand finger function Arm function Walking/bending Pain Family support Social Activity Mood Tension Self Care Work Household tasks Fatigue * - Added as non-AIMS2 Sample 6 rheumatology clinics in Wisconsin & North Carolina 18 rheumatologists 440 patients >44 years of age, RA diagnosis 91 % approached agreed to participate 73 % retained one year for observations Data Collection Baseline, 6-mo, 12-mo data collected Audiotaped & coded all encounters Topics discussed at visit Pts surveyed post-visit at clinic visits Recalled pre-visit health priorities AIMS 2 health status for each domain Collected Pt computer assessments Pre-visit health priorities Inter-Rater Reliability of Discussion Events 7+ coders in two locations; bi-weekly calls Inter-Rater Reliability of Tape Coding High (Kappa) .76-.95 Were the patient’s top priorities for improvement (AIMS2 items) discussed in the visit? Sample Demographics Women: 76% White: 87% Mean Age: 62 Education: 42% H S or less Mean Income: <$40,000. Health Concerns Discussed Baseline Pain discussed often (98% visits) Walk/ bend (50%); Hand/finger function (21%); Work (21%); Household tasks (20%); Fatigue (20%) Mobility (7.1%); Arm (6%); Mood (4.5%) Social activities (3%); Tension (2.1%) Self Care & Support discussed rarely (< 1% visits) Compared 3 Printout Groups For Each AIMS2 Priority Area Group 1: (E1) Printout says Pt wants to improve X target priority domain in next 30 days. Group 2: (E2) Printout says Pt wants to improve other health domains, but NOT the X target priority domain. Group 3: (Control) Printout summarizes lifestyle issues, NOT health domain priorities Pt wants to improve % Taped Visits Discussed Mood Example: (E) Baseline (226): No cue = 5.8% visits (C) Baseline (214) No cue = 3.3% visits (E) 12-Mo (155)=19% (C) 12-Mo (166)=3% Cued Priority = 56% visits No cue = 3% visits: (9/16 visits) (5/166 visits) Not Cued P = 13% visits: (21/134 visits) Analyzed Probability That Topic Discussed For 3 Groups Predicted probability separately that each of AIMS topics was discussed separately at visit (Logistic Regression) Tracked experimental & control groups Divided experimental groups by whether topic was cued as a pre-visit priority or not at the 6-mo (T1) or 12-mo (T2) visit Model probabilities adjusted for clinic differences & Pt AIMS dimension score at time of visit; baseline score referent Cases insufficient to model 3 psycho-social domains discussed rarely and also pain discussed almost all the time Graph Key: + <.10; *< .05; ** <.01 Probability Mood Discussed by 3 Groups at Baseline, 6-mo, 12-mo 30 Percent Probability Mood Discussed 40 22.9 19.6 20 EXP Mood Priority EXP Mood Not Priority Control Not Cued 8.8 10 0 4.5 1.6 3.3 T0 T6 10.5 4 T12 Hand-Finger Arm Function 60 60 40 35.4 * 30 23 21.9 21 15.5 10 Percent Percent 44.5 ** 40 20 51.8 ** 50 50 33 ** 30 21.9 20 17.6 10 6.8 6 0 0 TO T1 TO T2 80 30 5 13.1 12.7 12.7 50 Percent Percent 19.3+ T2 40 50.1 46.4 43.1 47.9 46.9 40.8 38.4 30 20 6.7 6.2 7.1 T1 60 22.4 ** 10 2.7 73.1 ** 70 20 15 8 Walk Bend Mobility 25 6.2 5.7 10 0 0 TO T1 T2 EXP not Priority EXP Priority Control Not Cued TO T1 T2 Household Tasks Fatigue 40 37.2 ** 50 41.8 + 30 20 20.2 17.8 15.9 15.8 10 20.1 12.1 17.9 20.6 14.1 0 T1 T2 TO T1 T2 Work Social 80 30 72.9 * 70 25 17.3 * 15 10 5.6 3 2.8 TO EXP not Priority 7.2 5.2 3.4 T1 EXP Priority Percent 60 21.7 + 20 Percent 21.3 19.7 8.2 TO 0 21.7 20 10 0 5 33.4 ** 30 Percent Percent 40 50 40 30 20 27.8 20.6 21.5 19.8 19.2 10 18.9 17 0 T2 Control Not Cued TO T1 T2 Found Overall Trend Probability to discuss most AIMS domains higher at 6-mo or 12-mo if printout listed domain as Pt priority For 6 of 9 domains it was true for both 6-mo and 12-mo visit In general Experimental group Pts without priority listed on their printout trended toward similar probabilities as the control group Limitations Study conducted with RA Pts only Study conducted with specialists; need primary care provider study Two states Small sample Pts: Dividing E groups limited sample for infrequent priorities In Summary Some Pt health topics discussed frequently; some discussed rarely regardless of whether a Pt priority Insufficient to depend on CAI cue alone to raise topics CAI assessment & printout increased probability discussing Pt’s health priorities esp Mood (Note: tobacco also) Easily transferrable and inexpensive Feasible to implement prior to visit Accepted and found useful by patients and physicians Next Research/ Analysis Analyzing intervention & discussion of priorities in relation to: Outcomes (pain, HQOL, mood, self efficacy, helplessness) Pt characteristics Encounter characteristics RCT - larger Pt sample with mood issues Plan to use a depression screener for eligibility THANK YOU to all of our patients, physicians and colleagues who collaborated on this project. Questions/ Suggestions! Other Domains Besides Mood If Pt reported domain was priority, probability it would be discussed was significantly higher (predicted by regression with covariates clinic, AIMs-2 scale score same time, baseline discussion score is referent) Arm/Shoulder (6 mo P<.001; 12 mo P< .002) Fatigue (no AIMS) (6 mo P<.014; 12 mo P< .006) Hand/Finger (6 mo P<.002; 12 mo P< .04) Social Activities (6 mo P<.032; 12 mo P< .03) Mobility (6 mo P<.01; 12 mo P< .11) Walk/Bend (6 mo P<.007) Household (12 mo P< .063) Work (12 mo P<.045) Pain (discussed usually) Self care, support, tension (very rare priority/discussion)