Notes on Dr. Groopman Talk on Medical Decisions

I attended an interesting talk at UCLA by physicians and authors Jerome Groopman and wife Pamela Hartzband. It was called "When Experts Disagree: The Art of Medical Decision Making." It looks like you can watch a video of a similar talk they did at another school online. The talk is related to their book Your Medical Mind.

Pamela studied at Harvard Medical School and did her residency in endocrinology at UCLA. She is now a physician educator and author. Jerome did college and med school at Columbia, residency at Mass Gen, and fellowship at UCLA in hematology and oncology. He's now a staff writer for The New Yorker and author of 5 books. He's done extensive research on HIV and cancer and has about 400 publications. Nice.

My notes and takeaways from the talk are below.

Thesis: When you understand why the experts disagree, you can make better medical decisions.

Conflicting expert opinions in news (looking at same research study results)
Examples:
Take or not take Vitamin D
Mammogram check protocol

Putting a number on utility
Linear scale from 0-1 (0 = death, 1 = perfect health)
Time trade off (how much time to give up to not have side effect)
Standard gamble (estimate odds you'd take to avoid some outcome vs. chance it might kill you immediately)

All 3 methods flawed cognitively
Can't predict future life
People adapt, med conditions not static

Self-reported quality of life not objective

In britain, blindness is rated 0.5 (reduces life 50%)
In US, much higher (blindness not as bad)

Daniel Kahneman gave address recently to med analysts
Classic way of measuring utility like measuring the ether
Textbooks not giving answers

Sir William Osler: listen to the patient, he'll tell you the answer

They interviewed many patients to come up with mindsets.

2 mindsets of people when diagnosed
Maximalist: be proactive, do everything and more
Minimalist: minimize medications

2 mindsets of medication orientation
Naturalism orientation: prefer those extracted from herb (60% of population)
Tech orientation: prefer those synthesized in lab using latest tech

2 mindsets when taking pills
Believers: confident about cure
Doubters: think treatment worse than disease

There's a medical risk calculator online they mentioned
Baseline risk of disease/bad event from your case might be really low (like 1%) vs. message pharma ad says like "med will reduce risk by 30%" (look at absolute risk and risk difference rather than relative)

Cognitive traps
Availability: effect of stories that stay in mind; overestimating likelihood of similar event; when you hear a story, go back to the #s
Framing: 10% of side effect sounds different from 90% of no side effect; physicians just as susceptible

Drug ads: for every $1000 spent, 24 new rx's written
Use stories and #s
#s framed in most positive way (risk reduced 30% [not 30% of 1%])

Dartmouth study found that drug ads make med seem 10X better than really is

Insurance company ads contribute
"Right outcome" is false hope
Controversies around expert opinions

Expert committee disagreements
Mammograms
PSA screening

Their medical mindsets:

Jerome
Eastern European Jewish tradition
Docs on pedestal
Docs heroes like presidents
Sci/tech honored, anything natural was throwback to village life
He is maximalist, believer, and tech orientation: every med and procedure means better health

Pam
1st child in family
Parents were ahead of curve on exercise
She is minimalist doubter

The gray zone
No one right answer for everyone
Patient can ask what would you do; not best way to decide.
Watch and wait vs. surgery yesterday

End of life
Advance directives: 50% of people change minds and choose differently from what wrote (because can't forecast future in circumstances that aren't experienced AND because can't predict all scenarios); experts have formulas for when medicine at end of life is futile, but they don't work in real world because can't identify those who will survive with good life quality
Surrogate decision making: go back to the mindsets

Medical decisions:

Patient mindset: first understand this
Numbers: look to see how apply to the individual
Stories: Daniel Gilbert research on estimating outcome utility best done by talking to someone who had the outcome

When patient knows his/her medical mind, can communicate this to their doc and better explain risk/benefit

Does take time between physician and patient
People aren't automobiles on assembly line; takes time

Studies available online to anyone; doc's role changing from providing info to interpreting info
Patients like understanding where they fit into #s

Patient doesn't have to match mindset w/ doc; doc needs to know how to talk to any mindset

Healthcare reform and economics have insurance companies rating docs and dinging them when patients not complying
50% of medical studies/guidelines overturned in 5 years; 23% in 1 year

Step back and just acknowledge that often don't have good data
instead, government panel just does sweeping regulation that isn't right

Presenting relative risk is allowed but misguides

Merits and costs of drug ads:
Public service announcement vs. hypnotic suggestion/persuasion
Can better restrict how present info

Quality/outcome approaches have failed
Hard to quantity
Those things that matter aren't quantifiable; what's quantifiable doesn't matter
EHR won't save $80B; not as much duplication of tests as claimed
Adds extra costs and time waste from restricted methods of data entry

Problems with metrics-based/pay-for-performance med: creates incentives to meet metrics and not patient care end goal

Checklists good for surgery, procedures, infection decreases but less for medicine

Insurance companies are businesses, not about saving lives or helping you get "right" outcome

Healthcare costs going up because now we have more people living longer.