ANNOUNCES
THE MEDICAL SPREADSHEET
In the very near future, your medical consultation will probably be modulated
by a piece of software that will be as indispensable to the doctor as the
number spreadsheet is to the accountant. We have finally caught up with Dan
Bricklin (of Visicalc fame, the forerunner of the accounting spreadsheet that
launched the Apple II microcomputer) and Carolus Linneas (Swedish scientist who
founded the Linnean classification system that gave us species names such as
Homo sapiens in the 1750s). A new class of software called the PLUM Medical
Spreadsheet® (patent pending) was demonstrated at the Royal Australian College
of General Practioner (RACGP) Computer Conference, Melbourne Convention Centre,
August 7-9, 1997 and the subsequent international Asia Pacific Association for
Medical Informatics (APAMI) 97 Conference in the Darling Harbour Convention
Centre, Sydney, Australia on August 10 -13, 1997. PLUM is a combination of the
power of the iterative problem solving method, characteristic of the
spreadsheet, and a medical belief system constructed along the lines of the
Linnean biological system. The PLUM name is an acronym for Presentation Links
Unity and Management, which is the health model used in the medical
spreadsheet.
The typical accounting spreadsheet
is a number cruncher and gives the accountant quick answers to "What
if?" type queries. The results of this "What if?" analysis are
placed in the spreadsheet cells; this sets up the conditions for the next round
of calculations with no manual transcription. The accountant's electronic
spreadsheet is prodigious for tasks that require repetitive work with a hand
held calculator. Hitherto, there is no such equivalent spreadsheet in either
the medical or legal domains. With PLUM, during a client encounter, there is
the capability of :
|
allowing data entry and recording |
|
performing "What if?" calculations pertaining to client diagnosis and management, with the results placed in cells for the next round of evaluation and |
|
features such as scrollable
worksheets which can be saved. |
is used in a
real or simulated patient or client encounter environment. During a patient
encounter, the clinician needs to record the clinical details and constantly
makes evaluations of clinical problems, resulting in treatment in the event of
a diagnosis. In the event that no diagnosis is made, further investigations are
instigated. PLUM facilitates the recording of clinical data while at the same
time using the same worksheet to provide computer evaluation of clinical status
with resultant guidance regarding problem analysis, investigations and
treatment. The motivation of PLUM is to introduce the spreadsheet metaphor into
clinical medicine. There were at least two main barriers that needed to be
surmounted to attain such an invention. The first was coming up with a
replacement for the number system used in conventional spreadsheets. PLUM uses
a word-based coding and medical belief system modelled on the Linnean
classification. In this system a medical condition such as gout is a medical
species with its own phyllum, class, order, family and genus. Surmounting the
second barrier was finding a suitable homogenous clinical data model of patient
health status at the local encounter and the global level.
The traditional model of the
encounter is called SOAP for Subjective (symptoms), Objective ( physical signs
), Assessment (what the doctor thinks of the whole consultation) and Plan of
management. In the old medical model, the patient global status is a list of active/inactive
problems. The homogenous model used by PLUM is the Graduated Discrete
Definition Model which allows the clinical encounter data to be viewed in the
same framework as the patient global status. This new model overcomes all the
shortcomings of the traditional medical record where the encounter is
disjointed from the patient global status. PLUM provides the clinician with a
spreadsheet tool for use in patient care to effect efficient diagnosis,
management and data recording. There is also a paper equivalent to the medical
spreadsheet which forms the basis of an effective manual/hybrid medical record
system. The input and output of the medical spreadsheet process are based on
the cells of the spreadsheet during a real or virtual patient encounter. As the
input and output arising from the patient evaluation are all cell based, this
provides the powerful paradigm of iterative and hypothetical type problem
solving. Applying the same spreadsheet metaphor, the pages (or worksheets) of
this medical spreadsheet can be scrolled back and forth and get saved for
future reference. This clinical data model used in the medical spreadsheet
supports a logical framework for the recording of the clinical encounter and
displays cumulative patient data in the same format as the clinical encounter.
In this Graduated Discrete Definition Model, the classification of all clinical
data uses one main criterion, which is the degree of definition of a clinical
datum in terms of readiness for medical treatment and/or prognostication.
With such a classification
framework, at one end of the scale there would be the not yet defined
clinical data such as a clinical symptom or sign; these clinical data items do
not have the degree of definition necessary for treatment or prognostication.
At the other end of the spectrum, we have the well defined clinical data
such as a diagnosis which has a clear prognosis and treatment. There can be one
or more intermediate categories positioned between the not yet defined and
the well defined categories. The preferred option of PLUM is an
intermediate category called the about to be defined category. To
complete this classification model, there is an additional category to include
all extrinsic patient clinical data comprising treatment and investigations.
The tetrad version of the Graduated Discrete Definition Model is called PLUM.
All clinical data to describe the patient status and patient management is
classified into the four categories of :
|
Presentation - this comprises all not yet defined clinical data such as symptoms and signs |
|
Links - this comprises all about to be defined clinical data such as the abnormal test results and provisional diagnoses made by the doctor; these entities are not specific enough for treatment and/or prognostication but better defined as compared with the not yet defined data |
|
Unity - this comprises all well defined clinical data of the clear diagnosis type where there is specific treatment and/or prognostication |
|
Management - this
comprises laboratory and radiological investigations, drug treatment ,
procedures and process of care. |
represents the four categories of the tetrad version of the
Graduated Discrete Definition Model. The clinical encounter form has four
cells, representing the four categories of PLUM. The Presentation cell is
reserved for not yet defined clinical data. The Links cell is reserved
for about to be defined clinical data. The Unity cell is reserved for well
defined clinical data. The Management cell is for clinical data related to
treatment and investigations. The PLUM model provides the framework for the
spreadsheet. The Graduated Discrete Definition Model, of which PLUM is the
tetrad implementation, is congruent with the underlying logic of the doctoring
process, which is the processing of unclear clinical information to resolution
in the form of a well-defined diagnosis. This is followed up with treatment, or
investigation if resolution is not possible. The “What-if?” query is launched
from the drop down menu, of which there is a choice of over 50 useful types out
of a possible 225 basic permutations. Typical queries of the medical
spreadsheet would be:
|
given symptoms
and signs, show possible diagnoses |
|
given a set of possible
diagnoses, symptoms, signs and laboratory test results, show only diagnoses
that conform to available data |
|
given a list of
diagnoses, show diagnosis that can unite several diagnoses. |
each
computer-assisted evaluation, the worksheet is updated and the worksheet page
number is incremented by one. The PLUM worksheet on the screen is called a page,
it depicts a real or hypothetical image of the patient status and can be saved
and recalled.
Docle Systems stumbled onto the
PLUM Medical Spreadsheet almost by default. It is the originator of the Linnean
Docle medical coding system used by over 2000 medical practitioners
Australia-wide. It was the congruent classification of disease entities, akin
to the building of a belief system framework that allowed the medical belief
system to be viewed using the spreadsheet metaphor. PLUM is relevant to any problem
domain where the problem solver needs to arrive at a diagnosis given a set of
indeterminate data. The knowledge domain that almost mimics the medical domain
is the legal field, and an equivalent theory for legal problem solving in the
spreadsheet form also exists (patent pending). PLUM represents a new class of
software - non-numerical spreadsheets. There is a stark message to doctors -
modern day medical practice is being swamped with information overload arising
from the advances in medical technology. The medical litigation arising from
the failure to cope with this plethora of information is ugly. The Information
Technology solution is to well, go for PLUM or....... choke on SOAP. PLUM is in
advanced beta, a subscription service is now available for those wanting to
immerse themselves in the forefront of medical computing.
Further information can be
obtained from:
Dr Y. Kuang Oon
Docle Systems
29 Darryl Street
Scoresby, Victoria 3179, Australia
Phone: 61 3 97638935
Fax: 61 3 97649788
Email: docle@compuserve.com
http:// www.docle.com.au