Computerized Patient Records Development
within
a
Rural
Health Care System
As an Internist practicing in a small rural
southern town from the early 1980's I struggled with the need for a legible
medical record. Typical of many physicians, my handwriting was awful. I decided
early on to bite the bullet and pay the price to have my office medical records
typed by a medical transcriptionist. Watching my partners' experience with some
of the early vintage computer based office management systems, I stayed with a
paper pegboard system of office management for the first several years of my
practice, but ultimately realized that computers, specifically PC's, were on
the scene to stay. By 1985, I had purchased my first PC and trained myself to
utilize its basic functions.
Because of the need to manage a large mailing
list stemming from a medical education organization I had founded, I had to learn some basic database
management skills. By early 1986, I had completed a first attempt at creating a
database driven computerized medical records software program for my own office
patients. Hence, rather than having the medical transcriptionist type office
notes into a word processor, I switched to a database model early on. With some
tweaks, my partner liked the program, and we added his records to the system a
year or so later. A couple of docs in the same office building liked it and we
added theirs to the pool. By the early 1990's I had made the program available to virtually all the physicians practicing
in the county.
By that point in time, I had begun typing the
progress notes of my own hospitalized patients. After doing this for a few
months, I noticed that a list of notes on a particular patient would accumulate
during a hospitalization. The idea of simply copying and pasting the notes into
chronological order to create the essence of a discharge summary emerged. As
these began to show up in the hospital's paper patient records, other
physicians noticed and this spawned the idea of having all the physicians
dictate their own daily inpatient progress notes as contrasted with the
traditional handwriting of notes. This would of course require some changes in
the medical records department since notes would have to have some priority in
order for physicians to utilize daily typed notes to follow the patients'
progress during the hospitalization. In addition, this would impose
considerably on the transcriptionists' performance because word processors were
simply not designed (at that point in the history of their development) as
organizational writing tools. I began work on a database model for inpatient
medical records that would allow the transcriptionists to type notes directly
into a relational database file. This would render the flexibility of having
the software organize the data from the notes into a variety of traditional
records (History and Physicals, Discharge Summaries, Consult Notes, Operative Notes, etc.).
By early 1995, I presented to the hospital's
board of directors a database program that would embrace these basic tenets of
design for their consideration. I was given the authority to make the necessary
medical records departmental changes to accommodate the program design and two physicians
(in addition to myself) were appointed to help work on its implementation.
First, I trained one medical transcriptionist using my own patients' records.
With this success behind us, she trained the rest of the transcriptionists and
we set up a call schedule for weekend coverage by the transcriptionists. I must
admit, at this point, that the software's implementation was successful more
because of the skill and willingness to work with us than upon the software's
design. They gave me excellent feedback and over a period of about six weeks,
together we tweaked the software's design to better fit their workflow and to
fit the physicians' needs. By the end of the summer of 1995, we had all staff
physicians (about 20 at the time) using the software. Physicians do not
generally embrace new technology with much enthusiasm. Their initial response
to anything that alters their established work routine is that of rejection.
Because of this, I will admit that my efforts were focused primarily on the
medical transcriptionists with the intent of keeping the software's functions
as transparent to the physicians as possible. They went on about their work
routine with the only change having been that of dictating (a task they already
utilized daily) their progress notes rather than writing them. This actually
saved them a small amount of time, but was seen as no great benefit by most.
The real jackpot from their perspective came when they realized that the
software actually assumed their task of creating a discharge summary. This,
more than any other aspect, helped pave the way to the software's acceptance.
None of our physicians have dictated a discharge summary since March of 1995.
There were other unexpected benefits with the system's implementation. For
example, one of the most vexing problems facing hospital administration is that
of convincing physicians to keep up their inpatient records such as discharge
summaries. The national average of time (in days) from discharge to billing for
hospital inpatient stays is 61 days. Ours has been less than 30 days since the
software was implemented. Hence, time in accounts receivable is markedly
improved using this approach because the problem of physicians' (un)willingness
to keep records current is taken out of the loop.
In similar fashion, we are able to bring data
from Emergency Department visits forward to populate components of the initial
History & Physical Exam, further saving the physicians time on
"paperwork". Today, we have more than half a million patient records spanning
nearly two decades in the system. Down time has been less than twelve hours
during the entire seventeen years of use.
Medical transcriptionists now connect to the
data files via VPN (virtual private network) connections and many work from
home. Today (2002) if one should walk into our hospital's medical record
department, he would find no paper charts except for those that have not yet
been shreded since a recent hospitalization. (Paper charts are still used in
the day to day care during a hospital stay. Once discharged, the patient's
paper record is destroyed and the digital record becomes the permanent record.)
All records on stored on the digital media of an electronic medical record
system combining the permanent nature inherent to optical discs and the rapid
access inherent to magnetic media.
Most of the physicians' offices now have
fiber connections to the hospital's network, although we began with dial in
modem connections. A patient seeing a surgeon for consultation in his office
can (with appropriate password permission) view the primary care physician's
office notes to see why the patient was referred. He may also view the records
of prior hospitalizations and/or surgeries directly from his office. The
records are legible, organized, and seamlessly available instantly upon need.
Emergency Department physicians may view records of patients' office visits to
local physicians, allowing access to histories of allergies, prior medical
problems, and current medications being taken. The system allows record level
password protection and event tracking allowing for HIPAA compliance for
access..
A. Robert Sheppard, M.D.