Shared Decision Making in Clinical Practice: What do Clinicians Need

1
Shared Decision Making in Clinical Practice
What do Clinicians Need?

• Blair Brooks, MD
• Charles Brackett MD
• Nan Cochran MD
• Dartmouth Hitchcock Medical Center
• WRJVAH
• With Support from the Foundation for Informed
  Medical Decision Making

2
(No Transcript)
3
Goals of ISDM workshop

• Demonstrate a brief Workshop format to teach SDM
  to primary care trainees and clinicians
• Focus workshop on key and manageable SDM skills
• Introduce educational tools and clinical DA tools
  to support SDM
• Get feedback

4
Why are we teaching SDM?

• Implementing Decision Aids at Dartmouth
• Systematic distribution
• Physician and patient request
• Physician DA prescription rate is lower than
  expected. Why?
• SDM skills
• Clinical tools

5
Why we chose this educational model

• Work within time constraints
• Engage clinicians
• Simplify to key SDM communication concepts
• not disease specific
• Offer tools to reinforce educational session
• Focus on residents (RCT) and include practicing
  MDs
• Outcomes to be measured
• Patient clinician communication
• Using real and simulated patients
• Assess long term (6 month) outcomes

6
Goals of Clinician SDM Workshop

• Why Shared Decision Making in clinical care?
• Introduce skills clinicians need for SDM
• Model tools that support SDM in practice

7
Usual Care - PSA Screening

• ..\DA excerpt movies\PSA movies\trigger
  tapes\PSA_1pp.mov

8
How much variation in this room?

• How many of you order PSA tests routinely on
• all male 50-75 yo patients?
• 75
• 50
• 25
• Never?

9
Are we consistent?

• How many of you
• Provide epidemiologically accurate information
• Elicit patients values
• Check for understanding of knowledge/values
• every time?
• Should we be doing this?

10
Why SDM?The 3 Categories of Care

• Effective care
• Evidence-based care that all with need should
  receive
• Supply-sensitive care
• Visits, hospitalizations, ICU admissions, etc
  where utilization is associated with supply of
  resources
• Preference-sensitive care
• Treatment choices with multiple options with
  different inherent benefits/risks patient values
  important to optimize decision
• Provider opinion often determines which treatment
  is used

11
Variation in medical practice
Preference sensitive care
Effective care
Should there be this much variation?
Supply sensitive care
12
Which surgery rate is right? Impact of improved
decision quality on surgery rates BPH
13
Preference sensitive care warrants shared
decision making
14
What is Shared Decision Making?

• Involves decisions that are
• Shared by doctors and patients
• Informed by the best evidence available about
  alternative treatments
• Weighted according to the specific needs,
  preferences and values of the patient.
• Adapted from Légaré et al. 2006

15
Original Model of Shared Decision Making

• Patient shares values and preferences
• may be based on past experience with medical
  choices
• may depend on current social situation

• Clinician shares medical expertise
• diagnosis
• treatment choices
• probabilities of outcomes

Together they arrive at an informed, shared
decision
16
Do patients want to participate in decision
making?

• Population survey
• 2765 English speaking adults
• asked their preferences re seeking information,
  discussing options, making the final decision
• Findings
• 96 prefer to be offered choices and asked their
  opinion
• 52 prefer to leave final decisions to MD
• 44 prefer to rely on MDs for medical knowledge
  rather than seeking information themselves
• Levinson,W. et al 2005

17
How well do physicians do SDM?

• Informed decision making in orthopedic surgery
• N 133 patients gt 60 yo facing surgical decision
• 92 Nature of decision discussed
• 14 Patients preferred role discussed
• 59 Risks/Benefits, alternatives discussed
• 12 Patient understanding assessed
• Braddock C. et 2008

18
Barriers to Shared Decision Making in Primary
Care

• Clinician
• Autonomy challenged
• Belief that patients dont want to participate
• Too many competing clinical agendas to address
• Inadequate SDM skills
• Evidence difficult to access, interpret,
  communicate
• Values clarification inexperience
• Practice
• Lack of time, reimbursement for SDM
• Patient
• Literacy, numeracy challenges

19
Key SDM Communication Skills

• Identify a preference sensitive decision
• Present Risks/Benefits Clearly
• 2 out of every 100 patients develop an infection
  after this surgery
• Clarify Patient Values
• What is most important to you in making this
  decision?
• Address Decisional Conflict
• Are you sure about the right choice for you? If
  not, how can I help you?

20
PSA Screening The data dump

• ..\DA excerpt movies\PSA movies\trigger
  tapes\PSA_2pp.mov

21
Risk communication

• the open two way exchange of information and
  opinion about risk, leading to better
  understanding and better decisions about clinical
  management.
• BMJ 2002

22
Why do clinicians need Risk Communication skills?

• Unfamiliar discipline to clinicians
• Quantitative risks rarely discussed with
  patients
• Research difficult to translate into
  understandable language
• Patients desire greater role, more information
• Patients who receive more information
• more satisfied and adherent
• Patients tend to overestimate benefit and
  underestimate risk without numbers

23
Professional organizations recommend SDM Are we
ready?

• 4/07 ACP guidelines on screening mammography for
  women aged 40-49
• Physicians should inform women about the benefits
  and harms
• Women should make an informed decision that
  includes their preferences and risk profile
• Are physicians up to the task?
• Are patients?
• How and when will this happen?

24
Transparent Risk Communication
25
Positive mammogram - What are the chances I
have breast cancer, doctor?

• The probability that a woman has breast cancer is
  1.
• If she has breast cancer, the probability is 90
  that her mammogram will be positive.
• If she does not have breast cancer, the
  probability of a positive result is 9.
• Hoffrage and Gigerenzer 1998.

26
Same exercise with the data presented as natural
frequencies

• 10 out of every 1000 women have breast cancer
• Of these 10 women with breast cancer, 9 will have
  a positive mammogram.
• Of the 990 women who do not have breast cancer,
  89 will have a positive mammogram.
• A woman with a positive mammogram comes to see
  you and asks What are the chances I have breast
  cancer, doctor?
• If we took 100 women your age with an abnormal
  MG, 9 will actually have breast cancer and 91
  will not
• Hoffrage and Gigerenzer 1998.

27
(No Transcript)
28

• Gigerenzer, G and Edwards From Innumeracy to
  Insight BMJ 2003

29
Communicating EffectivenessWhich statement is
better?

• Mammography reduces the risk of dying from breast
  cancer in the next 10 years by 25
• Mammography reduces the risk of dying from breast
  cancer in the next 10 years from 4/1000 to 3/1000
• If 1000 women have mammography, one will be saved
  from dying from breast cancer in the next 10 years

30
Relative vs Absolute Risk

• If efficacy expressed in relative terms and
  baseline risk is small, pts systematically
  overestimate treatment benefit
• Expressions that magnify benefit are more
  compelling to both clinicians and patients
• Gigerenzer, 2008 Malenka et al. 1993.

31
Mixing Absolute Risk with Relative Risk

• Typically RR reported for benefits and AR for
  harms
• USPSTF Guide uses RR to describe benefits of
  sigmoidoscopy, and AR to describe harms
• BMJ, Lancet, JAMA 1 in 3 articles had mismatched
  framing (2004-6)
• Sedrakyan, A., Shih, C. 2007

32
Survival rates vs. Mortality rates

• Rudy Giuliani ..my chance of surviving prostate
  cancer in the US? 82. My chance...in England?
  44 under socialized medicine (year 2000 5yr
  survival rates)
• Prostate Cancer Mortality rates
• 26 per 100,000 in US
• 27 per 100,000 in England

33
Lead time bias
34
Overdiagnosis bias
35
The Power of Framing Effects

• McNeil found pts, students and MDs more likely to
  choose surgery over radiation treatment for lung
  cancer patients when outcome framed as
  probability surviving vs. probability of dying.
• McNeil BJ. NEJM 1982

36
Gain Framing

• Emphasizes benefits of performing a behavior
• You will live longer if you quit smoking
• More effective when promoting prevention

37
Loss Framing

• Emphasize costs of NOT performing a behavior
• You are more likely to die earlier if you
  dont get screened for colon cancer.
• More effective when promoting screening

38
Goal Balanced Framing

• If we look at 100 women like you who have this
  surgery, 97 will survive and 3 will die
• Cumbersome to do for all Risks and Benefits
• Use this approach for the major message

39
Summary Recommended strategies for
communicating risk

• A. Natural Frequencies
• B. ARR
• C. Survival rates
• D. Balanced framing
• Which one does not belong in this list?
• Non transparent risk communication is more likely
  to mislead or manipulate/bias.

40
Summary PointsHow to Do Transparent RC

• Use natural frequencies
• Out of every 10 patients who take Prozac, 3-5
  experience a sexual problem.
• Use absolute risksMammography screening reduces
  the risk of dying from breast cancer by about 1
  in 1,000 from about 4 in 1,000 to 3 in 1,000.
• Use mortality rates There are 26 prostate
  cancer deaths per 1000,000 American men vs. 27
  per 1000,000 in Britain.
• Use balanced framingIf we look at 100 women
  like you who have this surgery, 97 will survive
  and 3 will die
• Use graphics, pictures

41
Helping Patients Understand Risk Information

• Risk of what? Over what time frame?
• How big is the risk?
• Does the risk information apply to you?
• How big is the change in risk?
• Does the change in risk reasonably apply to you?
• How does this risk compare to other risks?
  (context)
• If you get annual mammography screening starting
  at age 40, you will get about the same benefit as
  driving 300 fewer miles every year.
• Gigerenzer, - Reckoning risks
• Schwartz, Lisa et al. 2007

42
Next patient
64 y.o. man s/p MI with PTCA/DES 3.5 years ago,
no symptoms of CAD since. Is on beta blocker,
statin, aspirin and plavix. Complains of easy
bruising.
43
Over 2years, MI/stroke/CV death 6.8 vs.
7.3 (plavix vs placebo) RR 0.93,
p0.22 Hospitalization for ischemic event
16.7 vs. 17.9 RR 0.92, p.04 Severe
bleeding 1.7 vs. 1.3 RR 1.2, p.01 Bhatt
et al NEJM 2006
44

• If 1000 patients took plavix for 2 years, 5
  fewer would have a heart attack, stroke or
  cardiovascular death and 995 would have no
  benefit compared to 1000 patients taking aspirin
  alone.
• In addition, of these 1000 patients taking
  plavix, 4 more patients would have a serious
  gastrointestinal bleed.

45
Using a Drug Facts Box to Communicate Drug
Benefits and Harms

• Improved consumer knowledge of benefits and side
  effects
• Most control participants overestimated benefit,
  65 by a factor of 10 or more

Schwartz, Woloshin, Welch. Ann Int Med 2009
46
Presenting Risk Information

• Similar profile Probability of outcome in people
  like me who experience the problem
• Format Use quantitative, qualitative and
  graphic formats (100 faces) to enhance
  understanding
• Framing Risk message Positive message
  reduces bias

47
Key SDM Communication Skills

• Identify a preference sensitive decision
• Present risks/benefits Clearly
• 2 out of every 100 patients develop an infection
  after this surgery
• Clarify Patient Values
• What is most important to you in making this
  decision?
• Address Decisional Conflict
• Are you sure about the right choice for you? If
  not, how can I help you?

48
Asking Patients Values

• What is most important to you in making this
  decision?
• What clinical situations/decisions are patients
  values important?
• What is challenging about asking this question?

49
Decisional Conflict

• Uncertainty about which course of action to take
  when the choice among competing actions involves
  risk, loss, regret, or a challenge to personal
  life values.
• Legare et al 06

50
Consequences of Unresolved Decisional Conflict

• 59 times more likely to change mind
• 23 times more likely to delay decision
• 5 times more likely to have regret
• 3 times more likely to fail knowledge test
• 19 more likely to blame practitioner for bad
  outcomes

Sun, Q. MSc thesis. University of Ottawa, 2005.
Gattelari Ward J Med Screen 2004
51
Assessing Decision Conflict SURE

• Do you feel Sure about the best choice for you?
• Do you Understand the benefits and risks of each
  option?
• Are you clear about which benefits and risks
  matter most to you? (Risk/benefit ratio)
• Do you have enough support and advice to make a
  choice (Encourage)

52
Time is a major factor in primary care Can
we do one more thing?

• Time required for a full time clinician to
  deliver all
• highly recommended preventive services
• Time required to deliver all highly
• recommended chronic care services
• Yarnall, AJPH, 2005,2006
• __________________________________________________
  _____________
• Decision aids and other tools
• can help with this challenge.

7.4 hrs/day
10.6 hrs/day
53
Decision aids

• Adjunct to counseling
• Inform re options, benefits, risks
• Specify probabilities of outcomes
• Clarify personal values and norms
• Guide in deliberating and communicating

54
Decision Aids Can Improve Patient Knowledge

• ..\DA excerpt movies\colon\Colon Ca
  Knowledge.WMV.mov

55
Decision Aids Can Help Patients Clarify Values

• ..\DA excerpt movies\PSA movies\values
  excerpts\Bob_-_risk_of_treatment_complications.mpg

56
Effectiveness of DAs Cochrane Review

• 55 RCTs show that Decision Aids
• Improve knowledge
• Improve realistic expectations
• Lower decisional conflict
• Increase patient involvement in decision making
• Decrease number undecided
• Increase agreement between values and choice
• Cochran collaboration 2007

57
Can Decision Aids be successfully used in Primary
Care? DHMC experience

• Feasible, different distribution models employed
• Automatic
• Clinician Prescription
• Preventive medicine DAs (3 years)
• PSA 4288 distributed
• Colon Cancer 617 distributed
• Prescribed DAs (2 years)
• Chronic Condition/Preventive DAs
• 2214 distributed

58
How DAs can be used clinically PSA Clinician
summary report
Patient choice and values agree (concordant)
59
Clinician summary report CRC
Patient values are conflicting and discordant
with choice.
60
DA Summary Tools to help clinicians with SDM
61
SDM focused on Rx of Knee OA

• ..\DA excerpt movies\knee\Knee OA3 .WMV.mov

62
Practice SDM consultation

• Goals
• Engage patient in decision about knee surgery
• Communicate risks/benefits using summary tool
• Elicit patient values
• Assess for decisional conflict

63
Case practice

• You are seeing a 65 yo patient who is overweight
  and has severe osteoarthritis of the knee. She
  has been using Naprosyn and you gave her a
  steroid injection last visit (2 months ago) to
  see if that helped her pain.
• She saw the DA about treatment options last
  month. She is coming back in with the same chief
  complaint .
  
  
• _______________________________________
• Pair up 1 patient, 1 clinician
• Next slide will have relevant data
• Time 5 minutes

64
Summary of Content Tool Knee DA
65
Practice Debrief

• Clinicians How did you introduce patients role
  in decision making?
• Clinicians How did you present risk data?
• Patients did the clinician ask you what you
  cared about most in making this decision?
  (Values)
• Patients did the clinician ask if you were sure
  about your decision? (Decision conflict)

66
Debrief

• How would SDM work in your practice?
• How can you use Decision Aids?
• Summary tools?

67
End of Clinician Workshop

• Thank you for your participation!
• Now put back on your expert hats!
• Feedback?
• Challenges?