Status Review:
* Larry: The project to improve the sphygmochron prototype
application has 2 program bugs and 3-4 features. The next phase
is to write tests to support the maintenance of it. We need a
version control system. We need to decide if we want to post it
on a facility like Sourceforge.org and place it under version
control. El: Does this need satisfy FDA regulations? Bob: We need
a simple quality system, would take 3/4 of a page double spaced.
I'll write it. Germaine: We have Institutional Review Board (IRB)
approval for using the Sphygmochrom as part of the Halberg Center's
activities. El: We need to assure that we are complying with FDA
regulations We need to add a member to the team who has this experience
for future issues.
* Bob: Questioned the sphygmochron technique of assuming a circadian
cycle based on daily values. He wants to do this before continuing
with his presentation of the data analysis topic.
* Mary Jo: Got her piezo film boards working. Began working on
getting signals and measuring them consistently. Now working on
connecting the lab oscilloscope connected to computers for data
recording and analysis. She is comparing the analysis from the
midarm and the forearm. Also, she is determining the number of
measurements needed for calibration. Also, do it with different
times of the day with the same person and different people near
the same time. Also, measure the same person near the same time
under different activity, e.g. after walking and sitting. Germaine
said that heart rate is determined by measurements integrated
over time, while blood pressure is determined from the envelope
of many measurements.
* Alex: Received all the boards and parts. Dave is working on
increasing lab access, e.g. the number of people who can open
the lab.
* Gerry: Has been working on standardizing vocabulary.
* Germaine: Had a meeting of day - night ratio vs. circadian,
validated that people who have CHAD and are normal-tensive would
benefit from treatment. Has been reading a book by Dr. Kane at
the UMN, about chronic illness, similar to the work of Lawrence,
because medicine works well when there is an acute issue, then
solves it and return to normal. The incidence of chronic illness
is increasing: hypertension, obesity. This will not be solved
by short term treatment, but needs lifetime follow up and care.
The observation is made that current medicine is not set up to
treat these people, because different specialists must see it,
different support systems, will to change medicine to make it
more a team activity instead of just a specialist. Also, needs
Larry's work on the prototype modifications.
Coding Noise - Silent Threat to Our Health
Overview: Coding Noise
* Cognitive Steps in Clinical information
* Patient Information in clinical context
- Clinicians' mind: selection filter noise.
- Chart Note: Recording Noise
- Code billing form: Coding Noise
* Selection filter noise
* Recording Noise
- Coding Noise
-- Diagnosis
-- Flaws in Coding / classification
*Cognitive Steps of Clinical Information
* Clinician's Mind
- Select: Keep < 10% of information
- Noise: selection filter
* Patient information in clinical context
- Fine-grained information - many details - "pennies"
-- Lose information, misinformation is introduced, then the gaming
aspect as a biased game, as what you have to record as a clinic
to get reimbursed, because their is nothing in the standard vocabulary
to identify this.
-- For example, you can use your claims data to do clinical analysis.
Selection Filter Noise
* Clinicians are taught to select the
- Pertinent Positives (things that might cause the symptoms)
- Pertinent Negatives (things to rule out causes)
- and discard the rest.
* Not two clinicians use the same selection filter
- Even if they trained together.
* Fortunately, many different filters will read to the same clinical
diagnostic assessment!
Q:Bob: Are the references on diagnosis useful? Gerry: Some clinicians
believe that if you don't know what is going on in the first few
minutes, you should refer the patient to another clinician, because
you don't know the patterns, possibly based on your training,
e.g. Legionella - Legionnaires disease.
Recording Noise
* Valid findings
* Correct Clinical Diagnostic Assessment
* Recorded on the wrong patient,
or more often
* Not recorded at all.
- Your pen writes normal physical exam, even when you saw something
different! You usually write the normal symptoms more than the
abnormal symptoms. Often written in the wrong place. Notes may
not reflect the observations that were used to make the diagnosis.
The chart contains some meaningful information, but often isn't
reliable or complete. A lot of the old chart data is out of date,
has a short shelf life
Regarding the digital patient record, they are using templates to select and then use free text to add, that which isn't a selectable choice.
Coding Noise
* Miscoding: Assigning the wrong code
- CSF otorrhea (spinal fluid leaking out through the ear) vs CSF
rhinorrhea (in Medline), (eg. spinal fluid leaking out through
the nose.)
-- since physicians are not getting paid for coding, there is
a disincentive to correct it.
* Newly Recognized Diagnostic Entities
- HIV /Legionella
* "The System" will NOT allow the exact diagnosis to
be recorded! Often requiring them to move it to ICD 9 or 10 for
standardized diagnosis, e.g. Tracking patients with West Nile
Fever, they have to be diagnosed as miscellaneous fever.
* No Codes Exist for Diagnostic Entities
- Diagnosis/Code/Classification
- Flaws in Coding/Classification.
(Also the coding doesn't carry the which edition of the coding
vocabulary they used when assigning it.
Diagnosis/Code/Classification
* Diagnosis: Free Text used by the doctor
* Code: a thing substituted for something else
* Codes are identifiers
* Classifications are organizers
- Sets of pigeons in which to place other things.
ICD was originally developed as a pathology system, not a clinical system. E.g., 10-11 types of clitis media (ear infection).
Physician diagnostic notes are intended to treat the current patient, but it is being used to analysis public health.
Flaws in Coding / Classification
* Diagnostic Entity: Now way to capture, in permanent codes, the
exact diagnosis which the physician records.
* Classification set: We can't tell what individual codes mean
because they contain no identifiers.
One classification fits all: In our statistics derived from medical
records, we must make a single classification, ICD-9-CM, serve
all purposes.
* EMR: The electronic medical record typically sacrificed completeness
and accuracy for convenience in data input. (Often the note is
intended to remind the physician of what they were thinking when
they wrote it).
* Society: Increasingly interfering without placing complete and
truthful information in our medical records. The coded data is
used for reimbursement; the free text note is used for diagnosis.
Q: Have quality improvement techniques been applied to clinical healthcare? A: Yes, with funded demonstration projects, but often when the funding runs out, the technique is abandoned. Now it is used to achieve mass customization. However, it doesn't really work because you can't standardize the inputs. In 5-10 years, people will not go into primary care because they can't make enough to payoff their loans.
Next Sensor Team meeting: Sunday, July 2 at 9:00 a.m. in ECE 2-110 (lab)
Next Phoenix Project meeting: Sunday, July 9 at 2:30 p.m. in Mayo 748
This page is maintained by Ellis S Nolley. It was last updated on 26 June 2006.
The author(s) provide this information as a public service, and agree to place any novel and useful inventions disclosed herein into the public domain. They are not aware that this material infringes on the patent, copyright, trademark or trade secret rights of others. However, there is a possibility that such infringement may exist without their knowledge. The user assumes all responsibility for determining if this information infringes on the intellectual property rights of others before applying it to products or services.
Copyright (C) 2006 Ellis S. Nolley. Copying and distribution of this page is permitted in any medium, provided this notice is preserved.
Back to the Meeting Archive Page