Mon Oct 5 14:52:04 UTC 2015

Thanks to Jason for providing this paper. We gave the authors access to DXplain to do their study.  I believe it reinforces the idea that when the doctor has made up his/her mind, it is hard to change it, but there are many ways to deal with getting them to re-consider. Ideally, a DDSS would be able to parse the clinical note and come up with a differential. If a high importance disease had a high probability and was not listed in the note, the doctor could then be alerted. This would get around the reality that most patients present straight-forward problems and the DSS should not be forced upon clinicians for every visit.

Edward P Hoffer MD, FACC, FACP
Associate professor of Medicine, PT, Harvard

From: Jason Maude [mailto:Jason.Maude at ISABELHEALTHCARE.COM]
Sent: Monday, October 05, 2015 7:42 AM
Subject: Re: [IMPROVEDX] disappointed

There was a very interesting study published in Jan this year (attached) which showed how providing a GP or family doctor a differential before he/she started thinking improved diagnostic accuracy.

This is a great idea but the issue then is where does the differential come from? The patient must be the ideal person to do this with the tools now available. The Isabel symptom checker is also freely available to patients and includes a triage tool which helps them decide where to present and a pre filled form which the patient can print off and hand to his provider. This form captures the symptoms the patient entered together with the diagnoses suggested which the patient can check to highlight which he/she is concerned about.

I think there is great value in not only the patient providing this information before the consultation but also the provider having a record of how the patient described his/her own symptoms. We have seen cases recently where the patient's correct diagnosis appeared in the list based on how they entered their symptoms but did not based on the providers interpretation of what they thought they key symptoms were!


Jason Maude
Founder and CEO Isabel Healthcare
Tel: +44 1428 644886
Tel: +1 703 879 1890

From: David Hallbert <david.hallbert at GMAIL.COM<mailto:david.hallbert at GMAIL.COM>>
Reply-To: Society to Improve Diagnosis in Medicine <IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG<mailto:IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG>>, David Hallbert <david.hallbert at GMAIL.COM<mailto:david.hallbert at GMAIL.COM>>
Date: Thursday, 1 October 2015 10:23
Subject: Re: [IMPROVEDX] disappointed

John, are you familiar with PKC--now AskMD?  It's based on Larry Weed's work and has patients (not MD's) go through lists of questions about their symptoms and then organizes the symptoms based on diagnosis that have those symptoms and ranks them in terms of severity (so not to miss urgent, serious dx) and then on the number of sxs that go with each diagnosis.  I used to sit in from of a screen with a patient and have to go through a lengthy "differential diagnosis" list that was far more detailed than most physicians can possibly conjure.  Within each dx was a short discussion in patients' language about the dx, the specifics of the symptoms with that diagnosis and also what dx were missing were this what the patient had.  It was an amazing experience and both patient and MD left the discussion feeling the condition had been much more completely explored.  Having a computer generate such a list and gather information cut out so much bias and potential diagnostic error has been available for years and now is available free to patients on Sharecare.  I've looked at Art Papier's site as well and those two programs at least offer a much better opportunity at having a much broader view of diagnostic possibilities for experienced as well as inexperienced diagnosticians to have to consider.

David M. Hallbert, MD
Martins Point Healthcare, Bangor
650 Evergreen Woods
700 Mt Hope Ave
Bangor ME
david.hallbert at martinspoint.org<mailto:david.hallbert at martinspoint.org>

On Sun, Sep 27, 2015 at 1:48 AM, John Brush <jebrush at me.com<mailto:jebrush at me.com>> wrote:
I agree with Art Papier about problem lists. I think that better use of the problem list is one of 2 things that could have immediate impact on improving diagnosis in medicine.
Both Art and I trained at University of Vermont under Larry Weed. We were taught that a problem list should be dynamic. Problem statements can be provisional, temporary, active, permanent, inactive, or resolved. And the problem statement can change form day to day, becoming more specific as the diagnostic process proceeds. For example, "shortness of breath" can become "congestive heart failure" and then become "acute systolic heart failure" as new information is added and the diagnosis is re-evaluated through iterative hypothesis testing.
I think we need to be much more attentive to the logic, syntax and semantics of problems statements so that they are as clear and unambiguous as possible. And having a complete list of evolving provisional problem statements can sometimes cause the practitioner to recognize a constellation of problems that unify into a single diagnosis.
By the way, the other thing that will improve diagnosis is using likelihood ratios as a tool to teach trainees how to place the proper weight on clinical cues.

John E. Brush, Jr., M.D., FACC
Professor of Medicine
Eastern Virginia Medical School
Sentara Cardiology Specialists
844 Kempsville Road, Suite 204
Norfolk, VA 23502
Cell: 757-477-1990<tel:757-477-1990>
jebrush at me.com<mailto:jebrush at me.com>

On Sep 25, 2015, at 4:42 AM, Art Papier MD <apapier at LOGICALIMAGES.COM<mailto:apapier at LOGICALIMAGES.COM>> wrote:

This area of what and how an assessment is made and charted is an area
that the SIDM community should work with AMDIS (CMIO)/AMIA communtiy to
issue a recommendation as soon as possible.  Too often clinicians
expediently write a thin assessment and expediently put in an ICD
tentative diagnosis that erroroneously becomes THE diagnosis.  In my
clinic yesterday afternoon we saw a patient that had a single dystrophic
nail and distal finger edema.  The problem had been going on for years.
The patient had previosuly been treated for onychomycosis.  During the
visit i noticed she was constantly manipulating the finger as we talked.
We had a discussion that her problem might be due to this unconcious
habit.  She agreed.  The plan that was recommended was to do a fungal
culture, and for her to keep the finger bandaged and to not manipulate or
rub or touch that distal finger, then to come back in a couple months.  We
use Epic, Our residents chart and i review the notes at the end of clinic
or within 24 hours.  Yesterday late i reviewed the note for the patient.
The assessment was one word "onychomycosis".  I sent the resident a note
through the EHR system that the assessment was not simply onychomycosis
and asked her to change the note to reflect what we discussed.  The
resident used an ICD 110.1  which means "onychomycosis".  How often do
physicians use the expedient ICD?  I know i typically do not enforce
preceision with it, (it takes too much time) We desperately need a code
that says NYD (or equivalent)  so that we do not expediently place one of
the diagnoses that is in the differential, but is NOT with certainty the
diagnosis.  Furthermore a good assessment will explain the "basis" for the
diagnosis.  I rarely see these in a note, even when the diagnosis is
tentative. Since many patients are not seeing the same physician in
follow-up in some care situations, having an accurate note becomes even
more important in preventing diagnostic error.   Larry Weed has written
about what a good note is and the idea of having a "basis" for a
diagnosis.  The ICD and CPT billing system has a systemic flaw in what it
encourages,  which is imprecision and expedience.  We desperately need for
our community to come together with the informatics community and the
ONC/HHS to create a coding system that encourages to note in the EHR that
we are not sure yet when we do not have a basis for the diagnosis.

PS For those that watched the IOM presentation.  A question was read from
the podium alledgedly emailed in by Larry Weed about the role of computers
and decision support systems in fixing diagnostic error.  I spoke with
Larry this week.  Interestingly he did not watch the IOM presentation and
was quite puzzled and a bit upset that someone would email in and
represent themselves as him.  He asked "Why would someone do that?" That's
a good question!

Best Art Papier MD
Asssociate Professor of Dermatology and Medical Informatics University of
Rochester, CEO VisualDx

I'm seconding Brian Jackson's suggestion to move towards attaching levels
of certainty to each diagnosis.  I'm especially fond of the "NYD" label
Sam Campbell had on his list of the 10 best things to have happened in the
field of emergency medicine in Canada.   NYD = not yet diagnoses.  This
would alert the next person in the chain to keep thinking !   If we only
had an ICD-10 code for that .....

On Sep 24, 2015, at 10:32 AM, Jackson, Brian <brian.jackson at aruplab.com<mailto:brian.jackson at aruplab.com>>

We might want to be careful about this one.  From a research perspective,
analyzing time to diagnosis would indeed be interesting and productive.
But if in the process we give regulators, payors and attorneys new ways to
punish delays at an individual physician level, it could amplify premature
closure and overdiagnosis.

A suggestion others have brought up, and I suspect this needs to be pushed
more aggressively, is the concept of explicitly labeling diagnoses with
their level of certainty, e.g. as tentative or working diagnoses where
appropriate.  Sometimes empiric therapy is the best way to proceed.  When
empiric therapy fails, it doesn't necessarily mean the diagnostic process
failed, i.e. that a diagnostic error occurred.

Many of the complications introduced by both medicolegal and quality
improvement efforts come from treating diagnosis as a black and white
situation.  As much as I like the current news coverage of the IOM report,
it's reinforcing that black/white perspective.

--Brian Jackson

From: robert bell
[mailto:0000000296e45ec4-dmarc-request at LIST.IMPROVEDIAGNOSIS.ORG]
Sent: Wednesday, September 23, 2015 3:25 PM
Subject: Re: [IMPROVEDX] disappointed

But our foot is in the door.

Also, as we get more information on Time to Diagnosis, it would seem that
there could be a kind of "standard time to diagnosis," for less common
diseases/conditions that could be adjusted from time to time as we get
more proficient (e.g. for myasthenia gravis), perhaps even adjusted in
some way for the medical sophistication of the medical center in question!

Do we need to talk about standards for what is a delayed diagnosis?  Or
maybe that has already been discussed.

Just publishing a list of the common conditions we THINK are diseases of
delayed diagnosis would be a start.

Rob Bell, M.D.

On Sep 23, 2015, at 11:19 AM, Michael Grossman <Michael.Grossman at MIHS.ORG<mailto:Michael.Grossman at MIHS.ORG>
<mailto:Michael.Grossman at MIHS.ORG>> wrote:

I agree with your assessment. The IOM sometimes seems to fall short of
expectations. Their relatively recent report on Graduate Medical Education
was disappointing.
The fact that multiple news agencies are running with this story may lead
to some misunderstandings , especially on the nature of "delayed
Michael Grossman, MD MACP

-----Original Message-----
From: Robert L Wears, MD, MS, PhD [mailto:wears at UFL.EDU
<mailto:wears at UFL.EDU>]
Sent: Wednesday, September 23, 2015 9:31 AM
Subject: [IMPROVEDX] disappointed

Great that misdiagnosis and related failures are getting attention, but

This is a disappointing effort; a naïve, keyhole view of a complex

I count 82 mentions of 'error' in the first 15 pages (~5.5 per page), 753
in the entire document.

Lots of discussion about biases (78 mentions) but important issues that
challenge the focus on 'error', such as hindsight bias, or outcome bias,
are never mentioned even once.

This restriction to a very narrow framing of the problem is unlikely to
lead to progress.


Robert L Wears, MD, MS, PhD
University of Florida              Imperial College London
wears at ufl.edu<mailto:wears at ufl.edu> <mailto:wears at ufl.edu>
r.wears at imperial.ac.uk<mailto:r.wears at imperial.ac.uk> <mailto:r.wears at imperial.ac.uk>
1-904-244-4405<tel:1-904-244-4405> (ass't)            +44 (0)791 015 2219<tel:%2B44%20%280%29791%20015%202219>
Nothing matters very much, and very few things matter at all.



To unsubscribe from IMPROVEDX: click the following link:

Visit the searchable archives or adjust your subscription at: http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?INDEX

Moderator:David Meyers, Board Member, Society for Improving Diagnosis in Medicine

To learn more about SIDM visit:

The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.

Moderator: David Meyers, Board Member, Society to Improve Diagnosis in Medicine

HTML Version:
URL: <../attachments/20151005/6ce4686a/attachment.html>

More information about the Test mailing list