Society of Teachers of Family Medicine Conference on Practice Improvement November 23, 2013 Michael J Barry, President SHARED DECISION MAKING THE PINNACLE OF PATIENT-CENTERED CARE

FOUNDATION MISSION The mission of the Foundation is to inform and amplify the patient’s voice in health care decisions 2

PRINCIPLES THAT GUIDE OUR WORK Supported and encouraged to participate in their health care decisions Fully informed with accurate, unbiased and understandable information Respected by having their goals and concerns honored 3 We believe patients should be:

The Foundation has a licensing agreement with Health Dialog Provides royalties and contract funding to develop and maintain decision support materials Strict conflict-of-interest policy Staff and Medical Editors are prohibited from financial support from the drug and device industries THE FOUNDATION AND HEALTH DIALOG

IS INFORMED CONSENT “REAL?” In a survey of consecutive patients scheduled for an elective coronary revascularization procedure at Yale New Haven Hospital in : 75% believed PCI would help prevent an MI. 71% believed PCI would help them live longer. Holmboe ES. JGIM. 2000; 15:3625

IS INFORMED CONSENT “REAL?” Subsequently, a meta- analysis in 2009 (61 studies and 25,388 patients) concluded: “Sequential innovations in catheter-based treatment for non-acute coronary artery disease showed no evidence of an effect on death or myocardial infarction when compared to medical therapy. Trikalinos TA. Lancet. 2009; 373:911. Stergiopoulos K. Arch Intern Med 2012; 172:312 Pursani S. Circ Cardiocvasc Inter 2012;5:476 Thomas S. Can J Cardiol 2013; 29:472 6

IS INFORMED CONSENT “REAL?” In a survey of consecutive patients consented for an elective coronary angiogram and possible percutaneous coronary intervention at Baystate Medical Center in : 88% believed PCI would help prevent an MI. 76% believed PCI would help them live longer. Rothberg MB. Annals Intern Med. 2010; 153:307. 7

DECISIONS STUDY Conducted by University of Michigan Nationwide random-digit dial telephone survey Probability sample of 2,575 English speaking American age 40+ Reported a discussion of 1 of 9 medical decisions with a health care provider within the past 2 years Response rate of 51% The DECISIONS Study. Medical Decision Making.2010; 30 supplement I. 8

DECISIONS SURVEY: DECISIONS ADDRESSED Surgery Back surgery Knee/hip replacement Cataract extraction Cancer screening Prostate Colorectal Breast Medications Hypertension Hyperlipidemia Depression 9

HOW MUCH DID PATIENTS KNOW? Clinical experts identified 4-5 facts a person should know, for example, common side effects of medications or surgery Respondents were asked the knowledge questions related to their decision For 8 out of 10 decisions, fewer than half of respondents could get more than one knowledge question right 10

ARE PATIENTS INFORMED AND INVOLVED? Question Percent Who Answered Correctly How many people … get pain relief from joint replacement surgery28 … experience a surgical complication (e.g. wound infection)46 … will have replacement last at least 20 years15 How long most people require to return to normal activity39 11 The Decisions Study. Medical Decision Making 2010; 30 supplement 1

TOP THREE GOALS AND CONCERNS FOR BREAST CANCER DECISIONS Condition: GoalPatProv p Keep your breast?71% Live as long as possible?96% Look natural without clothes80% Avoid using prosthesis0% KR Sepucha et al/Pt Education and Counseling 73(2008)

TOP THREE GOALS AND CONCERNS FOR BREAST CANCER DECISIONS Condition: GoalPatProv p Keep your breast?7%71% P<0.01 Live as long as possible?59%96% P=0.01 Look natural without clothes33%80% P=0.05 Avoid using prosthesis33%0% P<0.01 KR Sepucha et al/Pt Education and Counseling 73(2008)

THE SILENT MISDIAGNOSIS “Many doctors aspire to excellence in diagnosing disease. Far fewer, unfortunately, aspire to the same standards of excellence in diagnosing what patients want.” 14 Mulley A, Trimble C, Elwyn G. Patients' preferences matter: stop the silent misdiagnosis. 367 London: King's Fund; 2012

U.S. CORONARY BYPASS RATES 15

FORCES SUSTAINING UNWANTED PRACTICE VARIATION Poor Decision Quality Unwanted Practice Variation Patients: Making Decisions in the Face of Avoidable Ignorance Clinicians: Less than optimal “Diagnosis” of Patients’ Preferences 16

WHAT IS GOOD MEDICAL CARE? It is not just about doing things right It is also about doing the right thing Preference-sensitive care: For many and perhaps most medical problems, there is more than one reasonable option 17

SHARED DECISION MAKING MODEL Key characteristics: At least two participants (clinician & patient) are involved Both parties share information Both parties take steps to build a consensus about the preferred treatment An agreement is reached on the treatment to implement Charles C. Soc Sci Med. 1997; 44:68118

PATIENT DECISION AIDS CAN HELP! Tools designed to help people participate in decision-making Provide information on the options Help patients clarify and communicate the values they associate with different features of the options The International Patient Decision Aid Standards Collaboration 19

PATIENT DECISION AIDS: TOOLS TO FACILITATE SDM 20 Describe a specific condition Present information organized around specific decisions Strive to keep information accessible (charts, graphs) and balanced Encourage patients to interpret information in context of their own goals and concerns Engage viewers with real patient stories Advise patients to make decisions with their physician © Informed Medical Decisions Foundation 2013

COCHRANE REVIEW OF DECISION AIDS In 86 trials in 6 countries of 34 different decisions, use has led to: Greater knowledge More accurate risk perceptions Lower decision conflict Greater participation in decision-making Fewer people remaining undecided Stacey et al. Cochrane Database of Systematic Reviews. 2011, Issue 10. Art. No.: CD

CHOICE OF ELECTIVE SURGERY 22

A CHORUS OF VOICES CALLING FOR SDM

DEMONSTRATION SITE PROGRAM Objective: to demonstrate that the use of patient decision aids and the process of shared decision making can effectively and efficiently become part of day-to-day care

1.Invite patient to participate 2.Present options 3.Provide information on benefits and risks 4.Assist patient in evaluating options based on their goals and concerns 5.Facilitate deliberation and decision making 6.Assist with implementation SIX STEPS TO SDM 25 © Informed Medical Decisions Foundation 2013

KNOWLEDGE SCORES BY DA EXPOSURE LEVEL: AGE GROUP p =.102 p <.001* 1 p =.220 (n= 613) (n= 1954) (n= 1273) Includes all valid demonstration site surveys in Illume database distributed in a primary care setting as of 8/1/12 (unweighted) * All significance tests are independent sample t-tests; * = Difference in means is statistically significant (p ≤ 0.05) 1 DA Exposure Level definition: Low = none of either OR some of both OR (some of one AND none of the other); Medium = Most of both OR (most or all of one AND (none or some of the other)); High = All of both OR (all of one AND most of the other)

KNOWLEDGE SCORES BY DA EXPOSURE LEVEL: EDUCATION LEVEL p =.71 p <.001* 1 p =.016* p =.006* Includes all valid demonstration site surveys in Illume database distributed in a primary care setting as of 8/1/12 (unweighted) * All significance tests are independent sample t-tests; * = Difference in means is statistically significant (p ≤ 0.05) 1 DA Exposure Level definition: Low = none of either OR some of both OR (some of one AND none of the other); Medium = Most of both OR (most or all of one AND (none or some of the other)); High = All of both OR (all of one AND most of the other) (n= 1694) (n= 983) (n= 1099)

IMPORTANCE RATINGS BY DEMOGRAPHIC GROUP Extremely Very Somewhat Not at all n p =.32 p =.031* 625 1,966 1,203 3,794 1, ,141 2,208 1,574 p =.045* Includes all valid demonstration site surveys in Illume database distributed in a primary care setting as of 8/1/12 (unweighted) **Statistically significant (p ≤ 0.05) (Chi square test)

DECISION ROLE PREFERENCES BY DEMOGRAPHIC GROUP You Both equally Your HCP n p <.001* Age Education Gender p <.001* 628 2,013 1,385 4,027 1,721 1,013 1,154 2,265 1,678 p <.001* Includes all valid demonstration site surveys in Illume database distributed in a primary care setting as of 8/1/12 (unweighted) *Statistically significant (p ≤ 0.05) (Chi square test)

PATIENT LEANINGS BEFORE AND AFTER DA: SCREENING (OMNIBUS TESTS) 30 Includes all valid demonstration site surveys in Illume database distributed in a primary care setting as of 8/1/12 (unweighted) *Significant difference (p ≤.05) (McNemar test)

Conducted by Health Dialog 174,120 health plan members, subset of 60,185 with six preference-sensitive conditions indicating need for SDM Randomized to usual-support or enhanced-support groups (~3x the contact) Telephonic health coaching about SDM (pDAs), self-care, and behavior change Wennberg D, et al. NEJM 2010; 363:1245 Veroff D, et al. Health Affairs 2013; 32:285 RCT of Telephonic Health Coaching

After 12 months, for the subset: 22.8% of the enhanced-support group and 7.5% of the usual-support group received a health coach contact Total medical costs per person were 5.3% ($23.27 pmpm) lower (P<0.05) Hospitalizations were 12.5% fewer (P<.001) with 9.9% fewer surgeries Cost about $5.00 pmpm RCT of Telephonic Health Coaching Wennberg D, et al. NEJM 2010; 363:1245 Veroff D, et al. Health Affairs 2013; 32:285

Introduced pDAs for hip/knee arthroplasty candidates in 2009 Reached 28% of eligible knee (N=3510)and 41% of hip patients (N=820) Over 6 months: 38% fewer knee replacements 26% fewer hip replacements 12-21% lower costs Arterburn D, et al. Health Affairs 2012; 31(9) HIP AND KNEE DECISION AIDS AT GROUP HEALTH 33

CMMI Innovation Grant “HVHC: Engaging Patients to Meet the Triple Aim” 16 member systems (~50 million served) will deploy “patient and family activators” Coaching and pDAs for DM, heart failure, back surgery, hip/knee arthroplasty Outcomes: quality and cost THE LARGEST IMPLEMENTATION STUDY? 34

THANK YOU! 35