Transcript Background and introduction: PowerPoint slides

CVDecide Implementation
January 2013
Richard Thomson
Professor of Epidemiology and Public Health
Decision Making and Organisation of Care Research Programme
Institute of Health and Society
Newcastle upon Tyne Medical School
UK Policy: UK Government
Shared decision making
will become the norm:
“No decision about me
without me”
MAGIC
making good decisions
Thanks for the decision aid…
I prefer this option Doctor
Models of clinical
decision making in the
consultation
SDM is an approach where clinicians and
patients make decisions together using the best
available evidence.
(Elwyn et al. BMJ 2010)
Paternalistic
Shared
Decision
Making
Informed Choice
Patient well informed (Knowledge)
Knows what’s important to them
(Values elicited)
Decision consistent with values
Examples of preference –
sensitive decisions
• Breast conserving therapy or mastectomy for
early breast cancer
• Repeat c-section or trial of labour after previous
c-section
• Watchful waiting or surgery for benign prostatic
hypertrophy
• Statins or diet and exercise to reduce CVD risk
Spectrum of SDM to SMS
“Shall I have a
prostate
operation?”
SKILLS
TOOLS
“Shall I have a
knee
replacement?”
“Shall I take a
statin tablet for
the rest of my
life?”
“Should I use
insulin or an
alternative?”
“I would like
to lose
weight”
“I would like to
eat/smoke/drink
less”
SDM – evidence
Cochrane Review of Patient Decision Aids(O’Connor et al
2011):
Improve knowledge
More accurate risk perceptions
Feeling better informed and clear about values
More active involvement
Fewer undecided after PDA
More patients achieving decisions that were informed and consistent with their
values
Reduced rates of: major elective invasive surgery in favour of conservative
options; PSA screening; menopausal hormones
Improves adherence to medication (Joosten, 2008)
Better outcomes in long term care
Are patients involved?
Patients who would like more invelvement in decisions about
their care (source: NHS Inpatient Surveys 2002 - 2011)
100
90
80
Percentage
70
60
50
45
46
47
47
48
49
48
48
48
48
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
40
30
20
10
0
Year
Why is decision
support needed?
• To allow high quality preference-based shared decisions
to be made
Why?
• Give patients the treatment they need and no less, and
the treatment they want and no more. (Al Mulley)
• No decisions in the face of avoidable ignorance
• Reduce un-warrranted variation
Aims of CVDecide
project
• to produce an interactive tool to assist
cardiovascular risk communication that will
form part of GPs’ existing electronic
desktops
• to extend the potential of the tool beyond
assessment of baseline risk by introducing
estimates of risk and benefit of preventive
interventions
Development
• Software development and incorporation within
EMIS.
• Incorporation of predictive equations
– Framingham and QRISK equations for cardiovascular
risk prediction
– Evidence-based predictive models for the effectiveness
of interventions, including lifestyle changes
– Adverse effects based on robust data.
• A period of iterative development with clinicians
and patients to assess acceptability and usability.
Implementation pilot
• Service based usability testing and process
evaluation in a sub-set of practices.
• Make CVDecide available for further learning
and testing.
• Work on requirements for roll out
• Link to Health Check programme
Results
• Three practices, six practitioners, 24 patients
• Consultation times 20-30 mins; tool open from 1- 19
mins (mean 4.7) per patient and 1.6 -12 mins by clinician
• Patients better prepared to decide after clinic (Deliberate
scale)
• Mean (SD) change in score for perceived behavioural
control was10.7 (4.5).
• Greater intention towards beneficial change and greater
perceived behavioural control regarding lifestyle factors
• Increased accuracy of risk perception
Results
• All patients would recommend this consultation
to a friend
–
–
–
–
–
–
“visual impact (M61)
“it’s explained a lot better than normally”(F73)
“it brings a smile to your face as well”(F43)
“the calculation on the computer told me”(M73),
“opened your mind up”(F50)
“for them [practitioner] to actually flag it up on a
screen gives you the ability to discuss what is up
there with the other person as well” (F43)
– “It highlights the fact it can happen to anybody, your
100 faces”(F53)
Results
– “It highlights the fact it can happen to anybody, your
100 faces”(F53)
– “In my case it was 29% of naughty red faces and if I
stopped smoking it would reduce by 12%” (M64).
– “I’d rather do that [loose weight] than go on statins”
but then acknowledged that if weight loss was not
successful he would, “consider the doctor’s opinion of
going onto statins.”(M66)
Results:HCPs
• Reported as quick, clever, visual and that patients
seemed to like it
• “not going to work [in 10 minute consultation]”
• “by doing it themselves, with diet and exercise they
could, maybe half their risk for some”
• “not persuaded, but quite keen to try [lifestyle changes]”
• “I don’t like medications – that phrase comes up very
frequently, and the other thing is that people sometimes
underestimate what they can do by lifestyle”
Improving the tool
•
•
•
•
•
•
Generally very well received
Easy to use after limited practice
Capacity to print out
Write back to record (align with SOP)
Access to QRISK
Consistency of risk communication within the
service
• Issues of who should use which components
(e.g. related to prescribing)
• Use in EMIS web
Conclusions:
Update
• Fully developed and integrated tool for use
in EMIS
• Further amendments have been made
– QRISK or Framingham
– Writes back to clinical record (including local
SOP READ codes)
– Print out options
– Better recorded action plan
Conclusions:
Roll out
• Three phase roll out, with review at end of each
stage
• Random selection and offer
• Support
– Web site (http://www.ncl.ac.uk/ihs/research/dmoc/cvdecide/index.htm )
– Training (includes optional 2-3 hr advanced skills)
– Technical (see web site)
• Evaluation
– Patient and clinician interviews
– Log data