Subj: What if an "old DOS" program "drag-raced" new ones?
PREAMBLE:
The following four pages are posted to answer objections and questions
about a free student/physician noteware couched in DOS and proferred
to this list in the recent past. In brief, the pages point out that the free-
ware is very fast and includes estimates that it might help educators with
evidence-based teaching and their graduates go ont to practice evidence-
based medicine [1].
Because of the proven capabilities for recording, storing, retrieving and
analyzing any fact, value or observation now put on paper, those that
master the freeware can expect:1. A fifty percent reduction in diagnos-
tic error because the software automatically retrieves old patient data,
that when reviewed, helps correct future mistakes. 2. An eighty percent
reduction in typing strokes to operate a computer or transcribe dictated
medical records. 3. A fifty percent cut in administrative overhead and
office billing expenses because the software can automatically encode
any historical item, physical finding, diagnosis or chargable procedure
that is put in a chart*. 4. On-line analysis functions that can audit any
clinical event, quality parameter or care outcome that is put in memory
by code. These audit trails can also be made to tell how students re-
spond to training, how they make judgments using the principles of
evidence-based medicine and grade tests of their learnings.
CRITIQUE FROM A MEMBER OF ONE LISTSERV: "Your applications
are not user-friendly".
ANSWER: I agree. My software's 81 main functions carried in 200 screens
are not friendly when first opened. However, most program seem difficult
until they are operated tens to hundreds of cycles. After reading the
built-in
directions and a little practice, the free apps get friendlier, easier and
faster
with every transaction.
The cozy look and feel of some applications may be because they do not
do much whereas "off-putting" programs ask that an operator work a little
to realize its benefits.
CRITIQUE: "I am sorry...but most of the computer-savvy students that I
teach will not waste their time on anything slower then machines with 200
milliherz clock speed...or application sets dating from before 1997. To go
back to your old DOS programs... would be too painful for me to suggest".
ANS: You and I can challenge the young savants to an informational
"drag-race". They get to load the biggest Cray computer or any of the
world's fastest parallel arrays with their choice of the newest softwares.
You and I will run the 1984 DOS/ FoxPro freeware in any old 386 or 486
personal unit. We will beat the "with-it folks" every time in the making,
messaging, retrieval and analysis of text-based medical records. Each
of tens of thousands of patient entries and hundreds of thousands of
factual recordings can be processed in seconds to a few minutes.
Why and how is this kind of speed possible from something "old"?
First, Windows and Apple-like programs are wonderful for beginners, for
the chronically-forgetful and those who are not "power-users". Power
users are regarded as those operating a computer through most of a
work day doing repetitive, stereotypical jobs. But, it is not well-appre-
ciated that the graphical user interfaces of WIN and MacIntosh have
very unfriendly speed and productivity limitations that can never be
overcome even after thousands of repetitive work cycles (below).
Second, most software offerings now sold for money or promoted by
institutions and quasi-governmental agencies are not "smart", not
"pro-active" and not very good at making lists. Each of these charges
and what they mean will be addressed in turn.
As to "list-making". Why is it so important to those in healthcare?
List-making is the central, most critical activity pursued by students as
they learn. For example, nearly all training involves the assembly and
memorization of fact lists about body structure, lists of the normal and
abnormal features of health and disease as well as yet more lists of the
tests and procedures required to appraise patients, diagnose conditions
and select treatments.
The table of contents in the front of any text book provide "lists" of the
knowledge that is made available. Then, the indices in the back stand
as yet more "lists" of what should have been learned.
Graduated physicians are list-makers too, but in a different way. Each
patient encounter generates physical findings, historical details, test
results, diagnoses and therapies for a specific person. Whether hand-
written, dictated or computer-generated, the case notes and treatment
records that evolve from care are really selected abstracts and replica-
tions of data elements from medical training lists.
The proferred freeware is a powerful list-maker that retrieves data from
old training and formulates them to make new electronic medical
records.
Yet another problem with most "medical softwares". Few attempt to
group or batch the data developed from patient care as people present
themselves to doctors individually, as groups of out-patients and as
lengthy rosters of hospitalzed cases. There are many side benefits
and processing efficiencies that can be gained from applications that
provide this kind of information management.
On the definitions of the terms "smart'" and "proactive": when used in the
context of information processing, these lables imply the existance of
automatic functions that look up old patient identities from system
memory to affirm or correct new entries, that there are automatic trans-
fers of previously-registered data from memory into new records to save
repetitive typing, and that a program seems to have almost human
"look-ahead" capabilities as it automatically sequences from completed
entry fields to the next logical work step of the recording job at hand.
One must concede that the earliest disc operating systems were slowed
by antique chips and made unfriendly because of the arcane commands
that had to be typed character by character at the system's infamous C>
prompt. But, these problems seem to be all that the new-age computer-
niks remember. They overlook or conveniently forget that later DOS offer-
ings have been so arranged that both simple as well as complex work
steps can be driven with one to a few touch-entered key presses. They
also fail to acknowledge that many of the custom programs have been
designed so that the DOS command line's C> prompt is never seen
after the applications are booted into view.
In the rush to be "up", "top". "not wet" and "savvy", the same enthusiasts
consistantly overlook the serious rate limitations of Win and Apple-like
programs. Moused cursor-crawls are always slower to execute then the
key-driven commands of "old DOS" programs. Why?
Consider this. The control of a cursor through a field of icons or action
points of a work screen demands the complete dedication of an opera-
tor's brain-eye axis and the total coordination of the arm, hand and fingers
before any transaction can be initiated with a "click". Computed produc-
tivity of any kind is always slowed and workers bound by moused pro-
grams can never improve execution speed even after thousands of re-
petitive operations. In summary, it does not require rocket science or
a randomly controlled trial to intuit that a key-press is always faster
then a mouse crawl.
Think more on this. Both simple as well as complex transactions can
be ordinated with finger taps that can be almost instantly and often in-
stinctively performed. Operators of properly-arrayed "command-line"
programs can keep their eyes on their jobs. At worst, even hunt and
peck typists need only quick glances at a keyboard to find what button
to hit. It is then immaterial how much computing power or clock speed
is put under a mouse...it will always be slower then a key-press.
These limitations have been noticed even by Microsoft. Some of Win-
dow's "Help" screens will tell how to make mice move with key-driven
commands. This haunts back to the fact that "friendly Windows" are
really nothing more than "old DOS" functions with pictures in front.
QUERY: "Would your software be used in class...to record valuable inform-
ation that might be appropriate for retention and later use?".
ANS: My notion is that lap top units would be used in class and connected
to a network either wired or wireless. In this mode, notes could be taken,
electronic syallabi accessed as needed, and electronic quiz answers
posted back to the network for a professor's grading.
QUERY: "I have downloaded your software and by the time a student tries to
take notes in classs by typing data in the right space of the right screen,
the
lecturer will have moved two concepts beyond".
ANS: This is a pertinent worry. It is likely that lectures abstracted from
the
network or copied from disc hand-outs would be brought up on a student's
screen and followed line-by-line with a cursor as the discourse goes on.
Marginal comments, amplifications and "watch this" alerts would be add-
ed by the student from time to time in free text. Following the lectures, a
student's study time would include transferring specific disease, procedural
and labor-saving transaction codes into the various compartments of the
applications. This kind of review and repeat brain-to-hand exercise should
meet Sir William Osler's turn-of-the-century directive for student study:
"...take a pencil in hand...make notes... and then study them to excise
the misconceptions and mistakes lodged in the Gall and Spurzheim
centers of the brain".
QUERY:"Your software sounds GREAT. What I would not give for the
ability to retreive information as easily as you describe".
ANS: You can have any or all of the following retrievals for free within
seconds to a few minutes: the entry of three or four seach letters will
get you many retrievals from the libraries and dedicated information
caches of the applications. Any electronic case recording, whether it
names a patient or a titles a lecture is recoverable by entering a few
letters of that name or title. My personal files hold about 150,000
patient notes and entry of a few search letters screens up to fifty cases
holding those search letters in less then a tenth of a second. The
freeware's integrated relational database enables searches for five
coded facts, values or observations at a time. It delivers the specific
counts or totals of each search term by looking through 750,000 files
in five minutes. Please...note that this recovery time is with an older
machine. One is certain that Crays, Pentia and other new micro-
processors would perform a little faster.
QUERY:"With all the problems I saw in your applications, get a
programmer to revamp it".
ANS: It would be great to translate the old applications into Health
Level Seven, the Health Plan Employer Data and Information Set,
JAVA, Linux or any of the latest languages de jure [2,3]. But, after
fifteen years of Medicare paycuts and hospital closures from mangled
care, I have run out of "spare" money. My electronic hobby of thirty
years has been subsidized with the surplus funds from private practice.
While the $300,000 costs for fifteen years of programming and equip-
ment have been more than repaid, I will now have to stand aside.
Those of you with access to deep pockets and lush grants are wel-
come to try the revampment.
In closing, the listserve has yet to return names of softwares that are
in use or recommended to students that might help them learn training
lists and then go on to be helpful in medical practice. The one discussed
above is still available through the courtesy of Doctor David Solomon
at http://www.med-ed-online.org/rsoftware.htm.
To restate, some 200 screens carry applications that might help with
learning and case care. The software utility integrates five major inform-
ation processing functions: the making and storage of electronic records,
their
retrieval, the automated messaging of data by vocal and facsimile techniques,
supports for decision-making and the on-line analysis of any coded fact regis-
tered in a system's memory.
Karl Thord Dockray, MD, DABR, ABNM
1808 19th Street
Lubbock, Texas 79401 USA
[log in to unmask]
09 22 1999
REFERENCES
1. Dockray, KT: Differing Ideas About Software for Medicine and Management,
Proceedings from the American Association of Management/International
Association of Management, 17th Annual Conference, Aug 1999. pp 165-170.
2. Schneider, EC et al: Enhancing Performance Measurement, JAMA, 282,
# 12, September 22/29, pp 1184-1190.
3. McDonald, CJ: Quality Measures and Electronic Medical Systems, JAMA,
282, #12, September 22/29, pp 1181-1182.
*. The US version of the freeware can manage the old Read codes that
come in capital letters. It would need some revision to accept those
codes using both upper and lower case characters.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
