Yes indeed Allen, you are absolutely right
You raise an important issue, about communicating uncertainty – especially in shared decision-making
I think most less-numerate people can cope with aleatory uncertainty – most people get the idea that we can’t say which way a coin
will land or what number will come up when we roll a die (although gambler’s fallacy suggests that people find it harder to realise that even after a sequence of 6,6,6,6 the next roll is as likely to be a 6 as any other number. But then we’re at the beginning
of
Rosencrantz and Guildenstern are dead!). In NICE decision aids we routinely say ‘it is not possible to know what will happen to any individual person’ . I think
Communicating epistemic uncertainty is more of a challenge. The IPDAS review by
Trevena et al summarises things by saying
- Many people exhibit “ambiguity aversion,” avoiding decision making and showing affective responses
to situations described as having epistemic uncertainty.
- Little consensus exists regarding how best to communicate these concepts.
That was 2012. I do know that there is active ongoing research in this area.
As we know, GRADE gives an overall measure of quality, which is perhaps worth exploring in communication with people facing the decision,
especially if we link that to
explanations of those terms:
- High: Further research is very unlikely to change our confidence in the estimate of effect
- Moderate: Further research is likely to have an important impact on our confidence in the estimate
of effect and may change the estimate
- Low: Further research is very likely to have an important impact on our confidence in the estimate
of effect and is likely to change the estimate
- Very low: Any estimate of effect is very uncertain
Best wishes
Andy
Andy Hutchinson
Medicines Education Technical Adviser
National Institute for Health and Care Excellence
Tel: 07824 604962
From: Evidence based health (EBH) [mailto:[log in to unmask]]
On Behalf Of Shaughnessy, Allen
Sent: 14 June 2018 13:47
To: [log in to unmask]
Subject: Re: Number needed to treat
Getting into the weeds a little more: Adding to Neal's and Andy's good points, I also think about the the 95% CI around an NNT, especially when the CI around the absolute difference is large (in this study, it was a five-fold diffference, 95%
CI 5·1–24·9). The 95% CI would then be 4.1 - 19.0 for this NNT of 6.7. This range gives me an understanding of the precision of the estimate (not very good) and an NNT of 19 gives me a different feel than an NNT of 4.1. Similar calculations can be done for
the natural frequencies.
Allen F. Shaughnessy, Pharm.D., M.Med.Ed.
Director, Master Teacher Fellowship
Professor of Family Medicine
Tufts University School of Medicine
195 Canal Street
Malden, MA 02148
781-338-0507
On Thu, Jun 14, 2018 at 8:33 AM, Neal Maskrey <[log in to unmask]> wrote:
On days 14–42, and on average:
· With hormonal therapy alone, 43 infants in 100
had spasms and 57 did not
· With hormonal therapy plus
vigabatrin, 28 infants in 100 had spasms and 72 did not
Thanks Andy! The key words being "on average’ because of course it isn’t possible to say in advance whether an individual child will be better with one treatment or the other, or be helped with neither, or whether one, or both treatments
may make things worse or even harm that individual. We do seem to be seeing a collective recognition that there is a huge difference between decision making for an individual and the evidence for a population. Both are needed of course, but the population-based
evidence can only inform individual decision making, and does not dictate a policy to be applied to everyone, everywhere, all of the time. Modern decision making in medicine is steadily progressing towards consultations being a skilled, nuanced negotiation
involving but not dominated by information - which is itself necessarily balanced by its validity and relevance (themselves involving some careful judgements).
Great example Rohini. Thank you for posting.
Visiting Professor of Evidence-informed decision making, Keele University
On days 14–42, and on average:
· With
hormonal therapy alone, 43 infants in 100 had spasms and 57 did not
· With
hormonal therapy plus vigabatrin, 28 infants in 100 had spasms and 72 did not
The information transmitted in this electronic communication, including all contents enclosed and/or attached, is intended only for the person or entity to whom it is addressed and
may contain material that is confidential, privileged or otherwise protected from disclosure. Any review, re-transmission, copy, disclosure, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other
than the intended recipient is prohibited. If you received this information in error, please contact the sender and properly destroy this communication and all copies thereof.
To unsubscribe from the EVIDENCE-BASED-HEALTH list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=EVIDENCE-BASED-HEALTH&A=1
__________________________
Delivered via MessageLabs
__________________________
