Direct from consumers: A survey of 1,048 customers of three DTC personal
genomic testing companies about motivations, attitudes, and responses to testing
Molecular and Clinical Genetics Panel of the Medical Devices AdvisoryCommittee, March 8-9, 2011
Dave Kaufman, Ph.D.Genetics and Public Policy Center
Johns Hopkins University, Washington, DC
Work funded by NHGRI Grant 1R1HG004865-01
Several valid concerns about DTC Testing
• Validity & utility• Risks and benefits are not well understood• Misleading or unwarranted claims • May lead to purchase for the wrong reasons• Consumers may be unable to interpret data• What will people do with it?
Empirical data being collected
DTC consumers’ perspectives Survey of customers of 3 companies23andMe ($429-$499), Navigenics ($999), deCODEme ($800-$2000)
• Randomly sampled & contacted by companies• 2-8 months after receiving results• Online survey, 20 minutes, $10• January - May 2010
• 1,163 of 3,167 invited responded (37%)• 1,048 (90%) qualified
Aims of the survey
• Who is purchasing the tests?• Why?• What are their overall reactions?• Interpretation and understanding• What are they doing with the information?
Limitations
• Cross-sectional survey: no follow-up over time• No data on individuals’ specific genes or risks• Data from all three companies analyzed in
aggregate
DTC Customers (%)
US 2006-08 population 18+
(%)Age
18-34 21 3135-54 34 3855-74 42 23
75+ 4 8
Women 46 51
Household Income$0-59,999 19 58
$60,000-84,999 14 16$85,000-124,999 23 13
$125,000+ 45 15
Education Attained0-12 Years 3 45
1-4 years College 46 46Postgraduate 54 9
Race/Ethnic Group White non-Hispanic 87 65
Asian/PI (non-Hispanic) 5 4Black non-Hispanic 1 12
Hispanic 3 15AI/AN (non-Hispanic) 0.5 1
Other non-Hispanic 3.5 2
Demographics
REASONS FOR TESTING
42% were interested in specific health conditions the company tests for
Rx Response
Eye Diseases
Carrier Status for Rare Disorder
Mental Health
GI Diseases
Autoimmune Diseases
Neurological Diseases
Cardiovascular Diseases
Cancer
0 5 10 15 20 25 30 35
12
12
14
14
16
20
23
24
29
Percent interested in disease type
33% were interested in part because first-degree relative(s) “have been diagnosed with one of the medical conditions included in the service”
How important were the following when you were deciding whether or not to use the company service?
Doctor recommended it
Friend or family recommended
Wanted to try something new
Participating in genetic research
Want to know diseases less likely to develop
Learning how to improve my health
Making sure privacy protected
Access to info not avail. via healthcare
To know conditions family is at risk for
Learn about my ancestral roots*
Want to know diseases more likely to develop
Satisfy my curiosity
-100 -50 0 50 100
Not very important
Not at all
Somewhat Important
Very Important
94
91
90
88
81
79
77
76
72
68
38
7
REACTIONS TO TESTING
39 49 10
Satisfaction
“How satisfied were you with the [23andMe/deCODEme/Navigenics] service?”
VerySatisfied
Satisfied
Very Dissatisfied
(0.5)
Dissatisfied(1.5)
Neither
Men
Women
White
non-H
Black n
on-H
Hispan
icAsia
nOth
er
0 - Some Colle
ge B.A.
Postgrad
uate
$0-59K
$60-84K
$85-124K
$125K+0
10
20
30
40
50
60
70
80
90
10089
86 88 89
8184 85 85
88 88 87 89 89 89
Satisfaction across demographic groups(% who were satisfied or very satisfied)
Overall reactions to the services
I satisfied my curiosity
I feel there is nothing I cando to change my health risks
I have relief from uncertaintyabout some aspects ofmy health
Agree 89%Disagree 11%
Agree 8%Disagree 92%
Agree 49%Disagree 51%
Clarity and value of the reports
Agree 88%Disagree 12%
Agree 38%Disagree 62%
Agree 84%Disagree 16%
92%54%
% who weresatisfied
Reports were easy tounderstand
Felt value of the data was worth the cost
Conclusions were too vague
INTERPRETING RESULTS
What participants felt they learned
Yes 58%
No 42%
Did you learn anything that you could use to
improve your health that you did not already know?
96%
76%
% who weresatisfied
Interpreting the dataParticipants were shown two hypothetical test results,using the exact format their company utilizes in its reports,and asked to interpret them
Interpreting the dataMary: 25% risk of diabetesPopulation risk: 30%
Mike: 11% risk of colorectal cancerPopulation risk : 6%
“Based on the information above, which answer best describes how Mary’s/Mike’s risk of [disease] compares to the average person’s risk?” Mary MikeMore likely to get disease 5 94Less likely to get disease 90 3Has the same risk 4 2I don’t know 2 1
Mary’s diabetes - % with incorrect answer in greenMike’s Colorectal Cancer - % with incorrect in
02468
1012141618
11
7 76
17
8
4 3 3
11
What do people do based on the data?
28% discussed results with a healthcare prof.Including: 20% with primary health care providers1% contacted a genetic counselor19% with other health care professionals
9% followed up with additional laboratory tests
16% changed medication or supplement regimens based on data
10% changed dietary supplements3% changed OTC medicines4% changed Rx, in consult with doctor0.4% changed Rx, without consulting doctor
Among those changing one or more regimens, 54% shared results w/ h.c. provider
Behavioral changes as a result -at one point in time!
34% more careful about diet0.1% less careful65% the same
14% exercise more31% “more determined to exercise”nobody exercises less or is less determined
Conclusions• Demographics of early DTC customers skewed• Curiosity about what might ail you highly
valued• Satisfaction high, though not among those
who found reports difficult
Conclusions• Some of what companies try to communicate
is understood• Whether information given is accurate or valid
is a different question• Measureable issues with interpretation
suggest reports can be improved• Behavioral changes observed
– could be positive or negative, transient or lasting– Are not the same as outcomes
• Longitudinal studies w/richer data needed
Thanks to
Juli MurphyKim KregerJoan ScottGail Javitt
Kathy HudsonStephanie Devaney
Robert GreenSara Katsanis
Survey Participants
Amy DuRossBrenna Sweeney
Joanna MountainAmy Keifer
Jeffrey GulcherJónheiður Ísleifsdóttir
Knowledge Networks
NIH
NavigenicsGPPC/JHU
23andMe
deCODEme
DukeBoston U.
Work funded by NHGRI Grant 1R1HG004865-01