Historical Perspectives and trends of
Medical
Informatics
regarding Countries at Different Stage of Developement
version 2.0, 10 June 2010,
- Introduction:
- A reflexion about where the current projects fit in an
historical perspective, raising questions about possible need of
different approaches when introducing informatics in new regions for
the first time ?
- Early adopter of informatics in developed countries:
- Prehistoric situation:
- Accurate descrition of symptoms were written in medical
textbook at last since the 17th century.
- Around 1900, careful handwritten records of individual
patients did become usual practice in some hospitals. At that
time the undestanding of infections did make sucessful surgery
possible,
although no antibiotics were easily available before 1950.
- In 1968 a visionary mile stone was the publication of "The
problem Oriented Medical Record" by L Weed, a professor of Family
Practice in an US university. The idea was simply to keep an
updated one page overview of the all the problems of the patient.
Every issue is explicitely identified as far as currently already
understood by the doctor and the related decision are noted explicitly.
He had already a vision of what
could be achieved using computers. Since he had actually no
computer at hand, but in order to manage the follow-up of patients with
multiple issues, he did require his assistants to keep updated
summaries of the current situation, rewriting the full page of paper
every time some issue did change, what we can get today on a screen in
miliseconds.
- Before 1970 computers were nearly not available in healthcare.
- +- 1970: Epidemiology:
- Since about a century the WHO did epidemiological surveys and
did develop the ICD, International Classification of Diseases.
The
manual processing of large collections of statistical information was
very laborious and epidemiologist were the first to take advantage of
the
potential of informatics.
- The focus was on classifications of deseases in larges
populations. However it was nearly not useful nor used for
individual patient care.
- +- 1975: Administration:
- The next step was to use computers in hospitals in order to
facilitate labor intensive tasks in the administration, as the
production of invoices.
- +- 1978: Lab:
- The first large scale medical application was the handling of
laboratory data inside the lab, for example facilitating
10.000 elementary transactions a day. The set of available lab
tests was growing and the workload become easier to manage with
acomputer, rather than with
pencil and paper.
- +- 1980: Communication:
- The first use of electronic communication was to make lab
reports rapidly available in the wards of the same hospital.
Indeed
wired communications accross several hundred meters become available at
affordable price.
Large labs still distributing paper reports did begin to send via
private postal services
which were faster than the official postal system. Reports
printed up
to 20:00 h were distributed before 07:00 h the next morning to a list
of frequent destinations at regional level.
- 1995-2005 : "Personal Computer":
- PC did become largely available and some doctors did begin to
keep their notes on a PC, i.e. a "Personal Computer".
- Such a "personal" computer was not particularly
intended to share information with anybody, except some "Email like"
echange of messages, e.g. lab reports in encrypted Emails as "HL7"
minded messages, although nearly every lab system has his own way to do
it.
- The market was and is still today very fragmented in hundred
of "medical
record" softwares, generally focusing on only one specific target group
of users, GP, or nurses, or cardiologist, or dermatologist,
etc...
- Collaboration limiting factors:
- Since all these systems were proprietary software there was
absolutely no incentive for interoperability. Official
standardization commissions did exist for many years but did not much
achieve up to now.
- In general doctors are individualistic and universities did
not pay attention to collaboration work. Indeed in universities
most
professors are
naturally the best in their own specialized domains.
- Economic model issues. In countries where doctors are
independent workers, they are paid in function of the number
consultations.
Maybe an excellent incentive for their involvment in the care of their
own patients and for the quality of their work, but actually to some
extend a conflict of interest with the notion of collaborative
work.
- Today in 2010 in Europe it seems that less than 1/4 of
doctors are really working with electonic support for medical
purposes. The other doctors still use papers to some
extend.
The claim that in some countries 100% of doctors own a computer will
not say they are really using it for medical purposes. The case
in
France where the declarations to the Social Security must be elctronic.
- The most important concern related to "personal computer" is
that most medical record systems currently in use, provide not much
more than "narrative approaches". In fact more or less an
electronic simulation of the traditional paper record. The
structure remains rougly the same as before i.e:
- Essentially a chronologic directory of the narrative
reports about every consultations, i.e. contacts with the patient, one
at a time. This means that the only way to retrieve a specific
information is to read everything. For example:
- If one would like to review all the current
medications of the patient, itt is difficult because prescriptions
may have been given during different previous contacts, by the same
doctor as well even worse if having prescribed by other doctors.
Hopefully the patient will be able to tell exactly what he is currently
taking.
- When records are strutured they are generally sorted on the
origin of the information, e.g. lab, Xrays,
history of admissions, vaccinations, etc... This means that the only
way to retrieve the new information from the latest weeks is to review
all chapters.
- 1990-2000: Discharge diagnoses codes:
- In some countries the Social Security did begin to require
the reporting of ICD codes for every discharge from hospital, as a
mandatory condition for the funding of the hospital. However this
has
no added value for individual patient care. Moreover from the
point of view of Public Health management, the information was limited
to hospitals and nearly no information was
yet available for all healthcare activities outside hospitals, in
primary
care as
well independent specialists.
- 1990-2005: Trend toward concentration:
- There is a trend to group hospital activities in larger
centers. Small hospitals were converted in elderly care
institutions for the growing number
of aged population. A few independent doctors begin to create
group
practices of 3 to 8 colleagues, providing services 365 days a year and
sharing a common premise with a secretary.
- 1980-2010: Electronic archives:
- The price of space in electronic archives did decrease
progressively of a factor 10.000 . Today data compression is no
more a
concern and large images can be archived a very low cost.
- Electronic archives are now more reliable than paper
archives,
because they can easily duplicated at low cost and backup can be saved
in more than one location, making the risks of data destruction much
lower.
- The handling of electronic archives cost less than time of
personnel who was necessary to necessary to handle manual
archives.
- In
tropical countries the maintenance of paper archives can be a concern
for other reasons,
due to organisational problems and climatic conditions, although the
cost of the work force is currently still much lower.
- 2005-2010: Networks:
- Although local networks inside hospitals did exists since
about 1980, large networks accross organization boundaries are just
beginning. Several factors play here a role:
- Lack of understanding and of agreements about
confidentiality rules.
- Fears and lack of trust about power or status issues
between healthcare professionals. Indeed networks require clear
rules
and agreements about who has power to do what. As an example how
far
do doctors agree to share patient record information with nurses, a
question encountering very different opinions.
- Maybe exaggerated anxiety about the security of new
networks.
Although this question is on itself of course critical, there is a
discrepancy between this anxiety and the usual way things are going in
the traditional practice, with a lot of negligences, about which nobody
complaints.
As an example medical information is often provided by ordinary
telephone, without any serious check about who is really at the other
side of the line, like " You are speaking with Dr XXX, please the
latest lab results from patient YYY".
- The availability of fast communication technologies, as
ADSL covering all the territory is relatively recent in the latest 5
years.
- Current challenges:
- Evolution from "narrative records" to "operational support",
"medical methodology":
- Narrative descriptions:
- Many existing patient records contains mostly narrative
descriptions of facts, as well on paper as well in computer systems.
New event are simply written down below the previous ones. This
is of course already very useful but do not take advantage of the
potential of IT.
- Every actor did keep his own chronological story in his own
way.
- Workflow support:
- Informatics networks can now provide a lot of support for
collaborative work. A few example of actors sharing information:
- Between GP and specialists.
- Between GP and nurse going at home.
- Between doctor and pharmacist, no more exclusively a one
way transmission of orders on receipts, i.e. small pieces of paper
often very difficult to read.
- Between basic primary case in a small center of
developing region and an expert in a remote university.
- ...
- In all such situations one should now pay more attention to
provide useful operational information about the current issues and the
current ongoing care activities.
This require more attention for "medical work methodology".
- Decision support:
- Medical methodology is seen as a necessary step to pave
the way for the introduction automated decision support procedures.
- Evolution from "messasage exchanges" toward "on-line networks":
- Although the use of on-line access to shared information
has many advantages, up to now this approach did take off relatively
slowly in medicine. This evolution does not preclude that message
exchanges will continue to represent a significant proportion of the
communication traffic:
- Message exchanges:
- The
sender need to be willing to send some message and need to take some
action. A reason why there is a risk that the partners in charge
of
the same patient would not get complete information in time.
That message become soon a kind of historical picture of the situation
at some
point in the past.
- Most current messages exchanges are based on some
"standard", like "HL7", are not easily
readable directly and require to buy some kind of specific viewer,
included in some proprietary software package.
- On-line access sharing:
- At any time get the current version of the situation.
- Due to the de facto standardization of web
technologies, no need for specific software since every computer has a
web browser.
- The medical sector is relatively behind in comparison with
other domains like travel agencies or banking, already working in real
time for years.
- Persistent misundrstandings between care and epidemiology:
- Many medical records projects are still suffering from lack
of clear definition of the objectives. Both of the following
goals are very valuable on itself, but implies different requirements
which need to be clearly understood:
- Patient care:
- Patient care require precise information about individual
patient, including qualification of the information like the "degree of
belief" of any assertions. At the current stage the use of free
text is still the most easy way of working. For example free text
allows phrases containing simple expressions like "not to be
excluded", "possible", "probably", ... which in practice may have
critical consequencse for the treaments. In primary care we are
dealing with many "problems" which require immediate decisions, while
there is not yet any "academic diagnose".
This source of information reflect a lot or reality but is too confuse
for epidemiological studies.
- Epidemiology:
- Epidemiologists need to group many individual cases in
the most nearby "class" as defined in the ICD from the WHO. They
do not need the details and work with "black or white" information in
order to analyse mean outcomes in large populations.
In practice many doctors are working for their patients, but are not at
all motivated to pay attention to such classifications, unless forced
by financial incentives.
- Early and late adopters of informatics, developing regions:
- Today 40 years later, emerging regions can take advantage of
the current telemedicine opportunities and should not necessarily go
through the same path of trials and errors as it was the case by early
adpters of informatics.
Migration to a more advanced work methodology and technical minded
maturity will always need some efforts, but could now go through a
shorter path?
- In developing countries wireless communication is growing
much
faster than wired networks. Despite the poverty, Africa seems to
have already more
than 100
millions mobile phones.
- Having very few resources, efficiency of healthcare is much
particularly critical. Getting more qualified
professionnals in remote areas is difficult and takes much
time.
Installing telemedicine networks can contribute to solutions of the
problems and is feasible much sooner.
- The brain drain remains a problem in healthcare because the
most qualified people are attracted by an easier life in more affluent
countries,
although if possible they would prefer to remain in their original
environment. Good telecommunication can alleviate the burden of
intellectual isolation.
- .....