A Discrete Choice Experiment Reveals Physician Bias in Management of Desmoid Tumors Simianu VV, Cizik AM, Flum DR, Mann GN, Davidson DJ University of Washington.

A Discrete Choice ExperimentReveals Physician Bias in Management of Desmoid Tumors

Simianu VV, Cizik AM, Flum DR, Mann GN, Davidson DJ

University of WashingtonDepartment of Surgery, Department of Orthopaedics and Sports Medicine

Desmoid Tumors

• Which approach is best? Resection: morbid, 70% recurrence-free survival Systemic therapy: still morbid, 70% symptom improvement Observational results: no progression in 65%

• When available evidence is controversial… Individualized decision

Patient characteristics

Physician preference

??

Research Question

• Can we quantify physician preference in treating desmoid tumors?

Patient characteristics

Physician preference

X%

Y%

• Discrete Choice Experiments: measure preferences in healthcare decisions simultaneously assess multiple attributes of therapy mimicking complex clinical decisions

• 3 Components: Treatment attributes Scenarios with paired choices Conditional model analytics

Methods: DCE

• Abdominal desmoid: 37yo female, 6cm mass• Extremity desmoid: 29yo male, 6cm mass• Attributes:

Outcome: probability of 2-year PFS Morbidity: immediate and/or long-term

• Vary attribute levels, size and symptoms

Methods: DCE

DC#1Attribute Treatment A Treatment Boutcome low high

morbidity low highWhich treatment do you prefer? A B

• 2 weeks later: unblinded treatment names Resection (high probability PFS, high morbidity) Systemic (med probability PFS, med morbidity) Observation (low probability PFS, low morbidity)

Methods: DCE

DC#1Attribute Observe Resectionoutcome low high

morbidity low highWhich treatment do you prefer? Observe ResectionDC#2

Attribute Systemic therapy Resectionoutcome medium low

morbidity medium lowWhich treatment do you prefer? Systemic therapy Observe

• Surgical Specialty (n=7)• Nonsurgical Specialty (n=6)

• How often do you treat desmoid tumors?

Participants

<1 1 to 5 5 to 10 >100

1

2

3

4

5

Number of desmoid cases seen per year

Num

ber o

f par

ticip

ants

Results

Lower Morbidity Higher MorbidityLower PFS Higher PFS

Results

Lower Morbidity Higher MorbidityLower PFS Higher PFS

Abdominal Blind

Unblind

Extremity Blind

Unblind

Results: Non-surgeons

Lower Morbidity Higher morbidityLower PFS Higher PFS

Abdominal Blind

Unblind

Extremity Blind

Unblind

Results: Surgeons

Lower Morbidity Higher MorbidityLower PFS Higher PFS

Abdominal Blind

Unblind

Extremity Blind

Unblind

• Assumptions of treatment attributes

• Simplification of clinical decision

• Generalizability to other physicians

Limitations

• Outcome is preferred attribute

• Similar in both surgeons and non-surgeons For both tumor locations

• DCE is new and complementary methodology to study preferences

Conclusions

FUNDING T32DK070555

COLLABORATION Surgery, Orthopaedics, OncologyCERTAIN CHOICES

MENTORSHIP Amy M. CizikDarin J. DavidsonDavid R. FlumGary N. Mann

Karen D. HorvathBrandt K. OelschlagerCarlos A. Pellegrini

Acknowledgment

University of WashingtonDepartments of Surgery and Orthopaedics and Sports Medicine

Preferences in Soft-Tissue Sarcoma:http://washington.co1.qualtrics.com/SE/?SID=SV_a9mQKpmkVtNTHYp

Acknowledgment

University of WashingtonDepartments of Surgery and Orthopaedics and Sports Medicine