Shared Decision Making, Decision Support and Breast Conservation Therapy

Shared Decision Making, Decision Support and

Breast Conserving Therapy

Shared Decision Making, Decision Support and

Breast Conserving Therapy

Matthew S. Katz, MD

Radiation Oncology Associates, PA

Saints Medical Center

Lowell, MA

April 27, 2009

Matthew S. Katz, MD

Radiation Oncology Associates, PA

Saints Medical Center

Lowell, MA

April 27, 2009

Shared decision making (SDM) and informed consent Background Patient-related factors Role of health care professional SDM in Breast Conserving Therapy (BCT)

Decision Aids and Decision Support Definition Goals Role in SDM

Shared decision making (SDM) and informed consent Background Patient-related factors Role of health care professional SDM in Breast Conserving Therapy (BCT)

Decision Aids and Decision Support Definition Goals Role in SDM

OverviewOverview

Decision support tools Nomograms IBTR! Effect on BCT utilization Relevance to multidisciplinary setting

Conclusions

Decision support tools Nomograms IBTR! Effect on BCT utilization Relevance to multidisciplinary setting

Conclusions

OverviewOverview

What I learned in medical school

What I learned in medical school

Doctor-Patient Relationship

Doctor-Patient Relationship

Docere = to teach

Obligated to transmit increasingly complex information to facilitate decision making

Increasingly collaborative rather than paternalistic

Docere = to teach

Obligated to transmit increasingly complex information to facilitate decision making

Increasingly collaborative rather than paternalistic

Challenges in Making Treatment Recommendations

Challenges in Making Treatment Recommendations

Do patients understand their own needs when making treatment decisions?

Each patient has different levels of need for information, support and autonomy

Patient’s needs/expectations change over time

Do patients understand their own needs when making treatment decisions?

Each patient has different levels of need for information, support and autonomy

Patient’s needs/expectations change over time

Shared Decision MakingShared Decision Making

More actively engages patients in becomes partners in decision-making process

Opportunity for clinicians to educate and reassure

Risk of increasing patient anxiety

More actively engages patients in becomes partners in decision-making process

Opportunity for clinicians to educate and reassure

Risk of increasing patient anxiety

Key Elements of SDMKey Elements of SDM

Patient knowledge

Explicit encouragement of patient participation

Appreciation of the patient's ability to play an active role in decision

Awareness of choice

Time

Patient knowledge

Explicit encouragement of patient participation

Appreciation of the patient's ability to play an active role in decision

Awareness of choice

Time

Fraenkel & McGraw, J Gen Intern Med. 2007

What factors contribute topatient’s decision process?What factors contribute to

patient’s decision process?

At time of decision Knowledge Deficit Preferred decisional role Uncertainty Anxiety

Evaluation of Decision Satisfaction Decisional Regret Provider Trust

At time of decision Knowledge Deficit Preferred decisional role Uncertainty Anxiety

Evaluation of Decision Satisfaction Decisional Regret Provider Trust

Tools to Assess SDMTools to Assess SDM

Breast Cancer Information Test

State-Trait Anxiety Inventory

CES-D Scale

Decisional Conflict Scale:

Knowledge, Values, Certainty, Support

Decisional Preferences Scale

Others

Breast Cancer Information Test

State-Trait Anxiety Inventory

CES-D Scale

Decisional Conflict Scale:

Knowledge, Values, Certainty, Support

Decisional Preferences Scale

Others

Information NeedsInformation Needs

Essential for informed consent Transmission of information Comprehension

Whether information is considered relevant by patient and clinician can vary

Patients express desire to know more, whether negative or positive

Essential for informed consent Transmission of information Comprehension

Whether information is considered relevant by patient and clinician can vary

Patients express desire to know more, whether negative or positive

Breast Conserving TherapyBreast Conserving Therapy

Since early to mid-1980s, lumpectomy + radiation therapy have been considered equally effective as mastectomy

Majority of women prefer BCT to mastectomy when offered both options

Since early to mid-1980s, lumpectomy + radiation therapy have been considered equally effective as mastectomy

Majority of women prefer BCT to mastectomy when offered both options

How informed are decisions about BCT?

How informed are decisions about BCT?

Population based sample of 1844 women Only 48% knew that survival is equal between

mastectomy and BCT

Only 16% knew that BCT may have higher local recurrence rate than mastectomy

Lower knowledge with:

male surgeon

lack of treatment options

less Internet or health pamphlet use

Population based sample of 1844 women Only 48% knew that survival is equal between

mastectomy and BCT

Only 16% knew that BCT may have higher local recurrence rate than mastectomy

Lower knowledge with:

male surgeon

lack of treatment options

less Internet or health pamphlet use

Fagerlin et al, Patient Ed Counseling 2006

What decisional role docancer patients want?

What decisional role docancer patients want?

Preferred Role Incidence

Active 20 - 39%

Collaborative 28 - 64%

Passive 8 - 52%

True autonomy in decision-making is rare (0.5-1%)

Preferred Role Incidence

Active 20 - 39%

Collaborative 28 - 64%

Passive 8 - 52%

True autonomy in decision-making is rare (0.5-1%)

Deber et al, Health Expectations 2007

Desired vs. Actual Role in Breast Cancer

Desired vs. Actual Role in Breast Cancer

Survey of 145 women with breast CA 1 week after surgery or neoadjuvant chemotherapy

Only 41% felt they had a choice in their treatment

63% had desired decisional role 30% preferred SDM 78% preferring active or passive

Survey of 145 women with breast CA 1 week after surgery or neoadjuvant chemotherapy

Only 41% felt they had a choice in their treatment

63% had desired decisional role 30% preferred SDM 78% preferring active or passive

Vogel et al, Psychooncology 2008

Preferred Role and Psychological Distress

Preferred Role and Psychological Distress

Passive preference associated with depression

SDM preference patients had highest anxiety levels

Passive preference associated with depression

SDM preference patients had highest anxiety levels

Vogel et al, Psychooncology 2008

Do breast cancer patients’desired role change?

Do breast cancer patients’desired role change?

Longitudinal study of 205 breast cancer patients

Desired decision-making role asessed at baseline

Active, Collaborative or Passive

Subsequently asked again ~ 3 years later

Longitudinal study of 205 breast cancer patients

Desired decision-making role asessed at baseline

Active, Collaborative or Passive

Subsequently asked again ~ 3 years later

Hack et al, Psychooncology 2006

Baseline Preferred Role

Active Collaborative Passive Total

Role Preference at 3 years

A 33 18 21 36%

C 19 22 43 42%

P 3 2 41 22%

27% 21% 52%


Desired role changes over timeDesired role changes over time

• “Active” patients had better QoL, physical, emotional metrics and less fatigue through treatment

• Patients that had been passive expressed ‘role regret’ more than decisional regret

• “Active” patients had better QoL, physical, emotional metrics and less fatigue through treatment

• Patients that had been passive expressed ‘role regret’ more than decisional regret


Satisfaction andProvider Trust after BCS

Satisfaction andProvider Trust after BCS

U Michigan survey of 714 breast cancer patients after BCS

Patients answered several months to several years after initial plans for BCS

Rated satisfaction, decisional conflict, decisional regret and provider trust

U Michigan survey of 714 breast cancer patients after BCS

Patients answered several months to several years after initial plans for BCS

Rated satisfaction, decisional conflict, decisional regret and provider trust

Walgee et al, Cancer 2008

Surgical ResultsSurgical Results

Re-Excision

One 43%

Two 12%

Required Mastectomy 11%

Infection 13%

Seroma 14%

Breast Asymmetry

Minimal 36%

Moderate 33%

Large 30%

Re-Excision

One 43%

Two 12%

Required Mastectomy 11%

Infection 13%

Seroma 14%

Breast Asymmetry

Minimal 36%

Moderate 33%

Large 30%


Decisional Conflict and Regret

Decisional Conflict and Regret

Endpoint %

Certain of decision 55

Prepared to make decision 61

Felt effective in the

decision-making process 87

Regretted decision 13

Endpoint %

Certain of decision 55

Prepared to make decision 61

Felt effective in the

decision-making process 87

Regretted decision 13


Patient SatisfactionPatient Satisfaction

No correlation with #re-excisions, complications or need for mastectomy

Breast asymmetry is associated with less satisfaction, less certainty in decision for BCS and more decisional regret

Women receiving RT more likely to feel have decisional conflict

No correlation with #re-excisions, complications or need for mastectomy

Breast asymmetry is associated with less satisfaction, less certainty in decision for BCS and more decisional regret

Women receiving RT more likely to feel have decisional conflict


Provider TrustProvider Trust

Lower MD trust with asymmetry, postop complications, need for re-excision

Trend toward less MD trust with higher education and Caucasian ethnicity

Lower MD trust with asymmetry, postop complications, need for re-excision

Trend toward less MD trust with higher education and Caucasian ethnicity


Differences by AgeDifferences by Age

Older women tend to make more immediate decisions Limited cognitive resources Greater knowledge/experience More likely to take passive role

Often have lower QoL, physical and emotional functioning scores

Older women still benefit from SDM

Older women tend to make more immediate decisions Limited cognitive resources Greater knowledge/experience More likely to take passive role

Often have lower QoL, physical and emotional functioning scores

Older women still benefit from SDMMeyer & Talbot, Psych Aging 2008Hack et al, Psychooncology 2006Liang et al, JCO 2002

What Barriers to Doctors see to SDM?

What Barriers to Doctors see to SDM?

Doctor-Related:

Insufficient information @ 1st Visit 29%

Insufficient time with the patient 28%

Patient-Related:

Misconceptions about disease 27%

Indecision 24%

Anxiety 22%

Lack of understand of information 20%

Doctor-Related:

Insufficient information @ 1st Visit 29%

Insufficient time with the patient 28%

Patient-Related:

Misconceptions about disease 27%

Indecision 24%

Anxiety 22%

Lack of understand of information 20%

Shepard et al, JCO 2008

Can clinicians interpret the patients’ desired role?Can clinicians interpret

the patients’ desired role?

101 candidates for BCS evaluated at baseline and f/u interview

Surgeons of these patients also interviewed

Assessed patient preference, patient and MD’s perception of the decision-making process

Average consultation = 21 minutes

101 candidates for BCS evaluated at baseline and f/u interview

Surgeons of these patients also interviewed

Assessed patient preference, patient and MD’s perception of the decision-making process

Average consultation = 21 minutes

Jann et al, JCO 2004

85% of women wanted shared or active role

Among patients desiring shared role: 50% felt more active than preferred 16% didn’t feel involved enough

Perception of patient and doctor correlated in only 38%

85% of women wanted shared or active role

Among patients desiring shared role: 50% felt more active than preferred 16% didn’t feel involved enough

Perception of patient and doctor correlated in only 38%

Jann et al, JCO 2004

PerceivedDecision-Making Process

PerceivedDecision-Making Process

Type of Decision Patient MD

Patient Only 6% 17%

Patient with MD Input 56% 25%

SDM 30% 56%

MD 8% 8%

Patient satisfaction correlated to perceived rather than preferred role

Similar study showed clinicians have difficulty assessing patient’s desired role

Type of Decision Patient MD

Patient Only 6% 17%

Patient with MD Input 56% 25%

SDM 30% 56%

MD 8% 8%

Patient satisfaction correlated to perceived rather than preferred role

Similar study showed clinicians have difficulty assessing patient’s desired role

Jann et al, JCO 2004Hudak et al, Med Decis Making 2008

SDM and Breast Conserving Therapy

SDM and Breast Conserving Therapy

Early stage patients have several different treatment options

Mastectomy +/- reconstruction

BCS +/- RT

Whole breast RT

APBI External Beam, MammoSite, Interstitial

Intraoperative

Early stage patients have several different treatment options

Mastectomy +/- reconstruction

BCS +/- RT

Whole breast RT

APBI External Beam, MammoSite, Interstitial

Intraoperative

Decisional Aids and BCSDecisional Aids and BCS

Walgee et al, JCO 2007

Increased BCS with DAIncreased BCS with DA


DA Enhances Patient Knowledge

DA Enhances Patient Knowledge


Decision Board from JAMADecision Board from JAMA

Whelan et al, JAMA 2004

May improve: Knowledge Patient satisfaction Improve physical/emotional function

May reduce: Decisional conflict Sensation of pain

No increased anxiety/depression using a DA

May improve: Knowledge Patient satisfaction Improve physical/emotional function

May reduce: Decisional conflict Sensation of pain

No increased anxiety/depression using a DA


Decision Aids (DA)Decision Aids (DA)

62% of patients felt a DA improved communication with the doctor

98% would recommend using the DA to other patients

DA use also favored by physicians

62% of patients felt a DA improved communication with the doctor

98% would recommend using the DA to other patients

DA use also favored by physicians


Decision SupportDecision Support

Computer based information resource to provide users with support for making decisions

Can help to standardize approaches to diagnosis, workup and treatment

Can also be used to individualize the process to the needs of different users

Computer based information resource to provide users with support for making decisions

Can help to standardize approaches to diagnosis, workup and treatment

Can also be used to individualize the process to the needs of different users

Goals inDecision Support

Goals inDecision Support

Improve patient knowledge base

Individualize information to each patient

Reduce clinicians’ knowledge deficit

Reduce effect of physician bias

Reduce patient distress and decisional regret

Improve patient knowledge base

Individualize information to each patient

Reduce clinicians’ knowledge deficit

Reduce effect of physician bias

Reduce patient distress and decisional regret

Advantages of Decision Support Tools*

Advantages of Decision Support Tools*

Help set tone and put patients at ease

Ensure essential information is transmitted

May be better than clinician: Reduce embarrassment Allow learning at comfortable pace Make it more effective time use when meeting

with clinician May more effectively engage in SDM

Help set tone and put patients at ease

Ensure essential information is transmitted

May be better than clinician: Reduce embarrassment Allow learning at comfortable pace Make it more effective time use when meeting

with clinician May more effectively engage in SDM

Green et al, Am J Med Genetics 2001* For genetic counseling

Decision Support and SDM in Breast CancerDecision Support and SDM in Breast Cancer

Multicenter phase III trial randomized 246 women <60 with breast cancer to: Standardized written materials Computer support at home

Both given before seeing MD

Patients surveyed 2 and 5 months later

Multicenter phase III trial randomized 246 women <60 with breast cancer to: Standardized written materials Computer support at home

Both given before seeing MD

Patients surveyed 2 and 5 months later

Gustafson et al, J Gen Intern Med 2001

Decision Support and SDM in Breast CancerDecision Support and SDM in Breast Cancer

At two months: Increased patient’s competence to deal with new

information Improved patient comfort in level of participation

in decision-making Increased confidence in MD

At five months: Enhanced patient’s information competence No difference in participation metrics

At two months: Increased patient’s competence to deal with new

information Improved patient comfort in level of participation

in decision-making Increased confidence in MD

At five months: Enhanced patient’s information competence No difference in participation metrics


Computer support was particularly helpful for: Less educated Non-Caucasian Underinsured

Computer support was particularly helpful for: Less educated Non-Caucasian Underinsured


Predictive Models in Breast Cancer

Predictive Models in Breast Cancer

Statistical methods of determining the most likely factors that predict a certain outcome

Statistical relationship doesn’t necessarily equal cause/effect relationship

Statistical methods of determining the most likely factors that predict a certain outcome

Statistical relationship doesn’t necessarily equal cause/effect relationship

Different Statistical ModelsDifferent Statistical Models

Risk group stratification

Nomograms

Artificial Neural Networks (ANN)

Classification and Regression Tree (CART)

Formulae

Risk group stratification

Nomograms

Artificial Neural Networks (ANN)

Classification and Regression Tree (CART)

Formulae

Not all statistical models give consistent results

Not all statistical models give consistent results

Mitchell et al, J Urol 2005

Individualized Decision Support

Individualized Decision Support

Provide clinician with evidence-based risk estimates

More helpful to cancer patients than standardized information

Interface can help optimize transmitting information by

Age Ethnicity, Language Education Psychological state Gender

Provide clinician with evidence-based risk estimates

More helpful to cancer patients than standardized information

Interface can help optimize transmitting information by

Age Ethnicity, Language Education Psychological state Gender

IBTR!IBTR!

Similar tool to Adjuvant! for determining risk of ipsilateral breast tumor recurrence after BCS

Included RCTs, meta-analyses and created composite estimate of 10-yr risk of LRR

Similar tool to Adjuvant! for determining risk of ipsilateral breast tumor recurrence after BCS

Included RCTs, meta-analyses and created composite estimate of 10-yr risk of LRR

IBTR!IBTR!

Includes data from RCTs, EBCTG

meta-analyses and single institution data

Variables included to calculate LR were: Age Tumor size Margin status LVI Grade Use of chemotherapy Use of hormonal therapy

Includes data from RCTs, EBCTG

meta-analyses and single institution data

Variables included to calculate LR were: Age Tumor size Margin status LVI Grade Use of chemotherapy Use of hormonal therapy

Sanghani et al, Am J Clin Oncol 2007

IBTR! ResultsIBTR! Results

MGH ValidationMGH Validation

Not enough women without RT to assess IBTR! estimates of no RT

Used IBTR! to estimate LRR in 1138 eligible women undergoing BCS and RT

Divided women into five risk groups: LRR <3% 3-5% 5.1-10% 10.1-20% >20%

Not enough women without RT to assess IBTR! estimates of no RT

Used IBTR! to estimate LRR in 1138 eligible women undergoing BCS and RT

Divided women into five risk groups: LRR <3% 3-5% 5.1-10% 10.1-20% >20% Sanghani et al, ASTRO 2008

Women with IBTR! estimates of <10% were accurate within 1%

Less accurate for Group 4 (4%) and Group 5 (7%) but still within 95% CI

Similar study from Vancouver demonstrated IBTR! reliable for low risk women, but overestimated LRR in <41, +margins or no RT

Women with IBTR! estimates of <10% were accurate within 1%

Less accurate for Group 4 (4%) and Group 5 (7%) but still within 95% CI

Similar study from Vancouver demonstrated IBTR! reliable for low risk women, but overestimated LRR in <41, +margins or no RT

Sanghani et al, ASTRO 2008Truong et al, ASTRO 2008

CaveatsCaveats

May not reflect more recently published research

Differences in patient population

May not include all the relevant variables

ER/PR/Her2-neu status Biomarkers

IBTR! limited to 10-Yr rates no comparison to mastectomy

May not reflect more recently published research

Differences in patient population

May not include all the relevant variables

ER/PR/Her2-neu status Biomarkers

IBTR! limited to 10-Yr rates no comparison to mastectomy

Will Decision SupportIncrease BCT Utilization?

Will Decision SupportIncrease BCT Utilization?

Use of decision support doesn’t guarantee favoring BCT

Uncertainty in best RT modality may be impediment to choosing BCT

Use of decision support doesn’t guarantee favoring BCT

Uncertainty in best RT modality may be impediment to choosing BCT

Prospective cohort of 125 BCS candidates seen at Dartmouth-Hitchcock Medical Center

Assessed for surgical choice, decisional conflict, values, knowledge, and decision role preference

Prospective cohort of 125 BCS candidates seen at Dartmouth-Hitchcock Medical Center

Assessed for surgical choice, decisional conflict, values, knowledge, and decision role preference

Collins et al, JCO 2009

>90% of women: Estimated 10-yr local failure after BCT Knew mastectomy and BCT had equal survival Had time to make a decision without an adverse

impact on survival

>90% of women: Estimated 10-yr local failure after BCT Knew mastectomy and BCT had equal survival Had time to make a decision without an adverse

impact on survival


Treatment Choice Before DA After DA After Consult

Mastectomy 22% 31% 35%

Breast Conserving Surgery

34% 31% 65%

Uncertain 43% 38% --


What factors influence decision for surgery?

What factors influence decision for surgery?


Treatment Options in BCT:More is Less*

Treatment Options in BCT:More is Less*

More RT options may lead to: Sense of Knowledge Deficit Uncertainty Anxiety Distress Regret Dissatisfaction

Increasing the complexity of decision-making for BCT may increase mastectomy rates

More RT options may lead to: Sense of Knowledge Deficit Uncertainty Anxiety Distress Regret Dissatisfaction

Increasing the complexity of decision-making for BCT may increase mastectomy rates

* Barry Schwartz, The Paradox of Choice

SDM and Multidisciplinary Breast Cancer Care

SDM and Multidisciplinary Breast Cancer Care

Pros: Enhances specialist communication Increases patient confidence in “team” approach Improves patient’s efficiency in obtaining expert

opinions in a single visit

Cons: Information overload (? enough depth) May affect patient’s ability to assert a more active

decisional role Medico legal aspects

33% of MDs feels discussion environment is suboptimal 85% may disagree with MDM decision but 71% don’t voice it

Pros: Enhances specialist communication Increases patient confidence in “team” approach Improves patient’s efficiency in obtaining expert

opinions in a single visit

Cons: Information overload (? enough depth) May affect patient’s ability to assert a more active

decisional role Medico legal aspects

33% of MDs feels discussion environment is suboptimal 85% may disagree with MDM decision but 71% don’t voice it

Decision Support before Consultation?

Decision Support before Consultation?

Provide information resource to patients before meeting with clinician

Assess patient’s desired decisional role more accurately

Cognitive/psychological support to “prime” patient for SDM

Reduce decisional regret, improve patient’s cancer experience

Provide information resource to patients before meeting with clinician

Assess patient’s desired decisional role more accurately

Cognitive/psychological support to “prime” patient for SDM

Reduce decisional regret, improve patient’s cancer experience

ConclusionsConclusions

SDM has become increasingly important

Unmet patient needs when deciding between mastectomy and BCT

Patient’s perceived role is key to satisfaction and provider trust

We don’t always assess patient’s desired decisional role accurately

SDM has become increasingly important

Unmet patient needs when deciding between mastectomy and BCT

Patient’s perceived role is key to satisfaction and provider trust

We don’t always assess patient’s desired decisional role accurately

ConclusionsConclusions

Better doctor-patient communication can reduce distress and decisional regret

Decision support can facilitate SDM for patient-clinician dyad

Effective SDM is a more important goal than BCT rates

Adjuvant! and IBTR! are just the beginning

Better doctor-patient communication can reduce distress and decisional regret

Decision support can facilitate SDM for patient-clinician dyad

Effective SDM is a more important goal than BCT rates

Adjuvant! and IBTR! are just the beginning

AcknowledgementsAcknowledgements

Julie Jones, MD

Alphonse Taghian, MGH

David Wazer, Tufts Medical Ctr.

Julie Jones, MD

Alphonse Taghian, MGH

David Wazer, Tufts Medical Ctr.

Health & Medicine

Shared Decision Making, Decision Support and Breast Conservation Therapy