Printer Friendly

The marriage of medicine and computers.

Physicians of the late 20th Century are dealing with unprecedented change. The way we practice was unenvisioned even the short time ago when we trained for our profession. We left residency with the mistaken impression that we could "get our arms around" the information necessary to maintain an academically rigorous practice. The tools we were given--stethoscope, reflex hammer, etc.--were all that were necessary for making a diagnosis, and a pen, paper, and a typewriter were all that were necessary for communicating to our charts and our colleagues. Our patients stayed with us for a lifetime, making their clinical information something we could carry mainly in our heads. Our source of clinical education materials was the few academic journals that arrived monthly for our leisurely review.

We were left on our own to develop "guidelines" of practice. Mostly these were clinical algorithms derived in an empirical, iterative way during our training. You probably couldn't find two of us to agree on all the steps of a diagnostic work-up, let alone reach consensus on treatment modalities. Just the thought of unanimity of practice would have been properly viewed as a Communist plot at best and at worst a failure to appreciate that "medicine is an art--not a science" subject to reproducible rules and standards of practice.

Relationships with our patients were sacrosanct. No third party would dare pierce this sacred bond. Whatever we and the patient mutually decided was the rule. The issue of a "payer" was at best clinically irrelevant and at worst an insidious invasion of an ancient relationship. We went into practice for a lifetime, confident that clinical medicine in our office or hospital would go on unchanged. We would never consider a change in specialty or, worse, a regression to the mundane world of administration. Management, and the information necessary to do it well, were the province of others, typically with less training, status, and intellectual prowess. We delegated to them the control of the "trivial," while we maintained for ourselves the realm of medicine--with all its mystery, power, and characteristic aloofness.

We in medical management, better than most, should appreciate that those days are well behind us. We had the insight to see the change coming over the past decade--and we did something about it. We joined ACPE and continued our education. We expanded our horizons. We learned how to talk to surgeons! We developed, per force, expertise in areas never covered in medical school, or college economics for that matter.

The entire skill set of medical management training has virtually no precedent in past training, either academic or real-life. In most of these areas, however, we see some "skill transferability." We can use our training in psychological management to deal with current dilemmas of employee relations issues; we use our rudimentary skills in biostatistics to analyze current marketing trends etc. The area where these "transferable" skills leave us most lacking, I would submit, is the realm of informatics.

I sometimes think of "informatics" as I used to think of syphilis--"if you understand it, you understand medicine." (Trust me--there is an analogy here!) Medical informatics encompasses the discipline that rolls data, individual facts, into information about patients, the flow of that information, its analysis, the means of recording it, communicating it, translating it from written text to numerical symbols. It also includes ways of categorizing, using, and accessing medical reference information and medical knowledge, ways of processing and monitoring diagnostic results, and ways of applying the existing body of knowledge in systematic ways to help in the process of diagnosis, treatment selection, and treatment monitoring. It is a synthesis of medicine, computer science, linguistics, statistics, pattern recognition, and other scientific domains.

The data we describe are not necessarily about single patients, but patients in populations, in demographic categories, as part of family units. (You get the picture starting to sound like "syphilis?") The technocrats of traditional informatics have their own language (of course) and their own gurus with arcane backgrounds and interests. They communicate in their own journals using daunting jargon serving the dual purpose of keeping outsiders out and finding common ground among insiders.

The informatics concept comes to medicine after an incredibly successful start, having changed the nature of heavy industry, pure science, and other distant disciplines. The players in medical informatics are a strange lot indeed. Their backgrounds are incredibly diverse. First, as expected, we find engineers, specifically electrical engineers, who are drawn to the field through their expertise in computers, telecommunications, etc. Next, we find pure computer scientists who are drawn for obvious reasons-the tool for this new discipline is the computer. Next we find systems (data) analysts. These are the folks who take our data and organize them in meaningful ways. They used to work for grocery store chains, helping them sort their different brands of corn flakes, but the future in medicine looked brighter! Finally, last but not least, are people such as you and I--physicians who, for a variety of reasons, have come to the conclusion that it's time to organize our clinical thoughts, to come to some consensus about the definitions of terms and concepts we use, to develop some standards regarding medical management, etc.

We have the desire to use technology that we see around us to solve problems in our clinical or administrative practices. Many of us started as amateur home users of computers, playing games, writing documents, or tracking expenses on these increasingly complex machines. Then again, as the field becomes more established, quite a few physicians are getting formal training in computer science, in management information systems, and in medical informatics. There are now graduate, doctoral, and postdoctoral programs in medical informatics at various major universities.

What is coming is a major opportunity to marry the complex nature of clinical medicine with the equally complex set of computer tools. This "marriage" will bear some rather fascinating offspring. We will begin to think and communicate more analytically about clinical issues in our daily lives--not just researchers, but ordinary physicians. We will gain wholly new and exciting "views" of data on our patients, our practices, our hospitals, our HMOs, etc. Through an analysis of these data will emerge the opportunity to "improve"--the siren song of total quality management (TQM) of late, but always something for which good clinicians strive. A review of recent TQM literature applied to medicine will reveal an increasing number of articles that involve informatics or rely on computer systems in the tracking and analysis of medical data.

Lest we get caught up in a frenzy of optimism, we would be wise to look at some of the constraints in this effort. What are the roadblocks? Why aren't we farther along? A more cynical observer would ask the logical question: "If doctors are so smart, and computers have been around for so long, why does medicine lag behind the grocery industry in this regard?" Peter Spitzer, MD, a computer expert and long-time observer of the medical informatics scene observes that:

* Medicine has been too nonstandardized everybody has practiced differently.

* The field is very complex; while a grocery store, bank, or insurance company may be dealing with dozens of different business processes and transactions, we are dealing with hundreds to thousands.

* Medicine is extremely information intensive, and much of the information is complex--multimedia, time-variant, fuzzy, and nonlinear; mainstream computing has only recently started producing the required tools.

* Information systems haven't had the capacities, throughputs, and price performance needed to run large health care operations.

* Compared to other information-intensive industries, health care has traditionally underinvested in MIS as a percentage of operating budget because there were no incentives to be cost efficient or to monitor quality in an objective way.

* Physicians in private practice have a historical interest in optimizing their own time use, practice style, and incomes--an institutional or management view was lacking.

An answer that occurs to me involves preoccupation. We're so busy trying to do a good job with the tools we have that we haven't found time to invest in the acquisition of these new skills. (I don't know why. We all learned how to use the sigmoidoscope fast enough). Maybe it's that old hang-up with "medicine as an art"-- unmeasurable. We haven't learned the axiom, "You can't improve what you can't measure."

Probably the easiest constraint to discuss (and, ideally, the easiest to overcome) is that of not taking the time to become familiar with the "beast" itself, the computer. This was certainly justified in years past. Dealing with the user interface of even 5 years ago was daunting--so much gibberish on the screen; arcane instructions for the not so easily put off; lack of support from vendors, whose attitude seemed to be, "If you don't understand, you probably shouldn't be playing with the machine in the first place"; and, of course, the enormous cost of getting started. These were the days of the chief financial officer acting as director of management information systems. Finance was the only department that could justify the costs of computers as work tools because, after all, it was in charge of collecting the money. Implicit was the concept that physicians went to it for data and its analysis.

How strange--we, the scientists, should have managed this all along. Things have changed! Perhaps it's because the computer industry woke up to the fact that its potential audience consisted of smart people who didn't appreciate always being made to feel dumb. Certainly, the machines and the programs on them have improved. The memory capabilities have been enhanced on a logarithmic scale. Costs for this enhanced performance have dropped precipitately. And, best of all, the interface with the user (that's you and me) has improved beyond all expectations. Gone are the arcane instructions. (I now judge new program software by the adage, "If you have to read the manual to proceed, it's not a great product.") Most users I know just sit down and noodle around with a new program for a while until, voila!, "I wrote a letter," or "I produced a spreadsheet," or, more recently, "I just created my own database to track clinical issues in my practice, and I did it myself."

Whether you're jumping on the bandwagon now with a new P.C. running '"Windows," or are committed to the Apple world of Macintosh machines, you are seeing a user-friendly environment where amateurs can succeed at real computing tasks. For many, computer systems installed some years ago present a real constraint. Users are now getting demanding; they want the ability to create their own "queries" in their own formats without the bother and expense of hiring a consultant or getting in line behind other users of proprietary software to present their suggested upgrade. Large medical practice mainframe vendors are scrambling to provide the computing power and ease of use available in off-the-shelf systems.

This increasing awareness of computers and of program applications has lead physicians into an area once reserved for statisticians, researchers, etc.--the concept of a database. This is a term that will become as familiar to clinicians in the future as the concept of "sepsis." It is a primal concept that addresses basic issues of the way we think, organize our thoughts, and record them for future use. Clinicians and administrators alike are probably used to thinking in a more linear fashion (today "A" happened, that led to "B" and I'll have to do "C" about the situation to lead to an outcome of "D." If a similar situation reoccurs next year, I hope my memory serves me well and I can make correlations necessary so as not to repeat mistakes.

The database concept says that we can break down a clinical situation, a laboratory experiment, a book of names and addresses, or a profile of patients into discrete "data elements." These elements can be categorized specifically as numbers, words (text), yes's and no's, etc. to make the recording of events and outcomes--and access to this information--repeatable and accurate. Having described all of the variables as "elements," we can begin to log "records"--the next name and address, the next patient profile--in the same consistent format. Having paid attention to definitions and scrupulous "data entry," we can now cull through hundreds or thousands of records, finding common events, finding records with specific characteristics.

This may sound like a tedious process, but it is the way all complex processes are kept straight. Imagine going to an "ATM" at the bank and having the machine tell you that it couldn't find your name because "name" wasn't one of the "data elements." Databases are all around us. They track our disciplinary histories (National Practitioner Data Bank), our credentialing procedures at hospitals, and the billing software in our front offices. In the near term, we will be expected to extend this capability into areas of medical practice improvement.

For example, it may no longer be adequate to inform a regulatory body that it is your common practice to immunize the children in your population. You will have to produce a database that lists all your children by age, by socioeconomic level, etc., with the immunizations given, the manufacturer of the serum, date given, and on and on. Thinking and recording information in this way reinforces a message given often by Larry Weed, MD, the father of the "problem-oriented medical record." He states that no one physician can ever know an adequate amount about either his patients (longitudinally) or the current literature supporting the care rendered patient. We will become increasingly dependent on databases of information about patients, their problems, and the literature to render high-quality care in the next century. It therefore behooves us to begin to think about the data accumulated around us in usable database form; if it's not in that form now (and most of it isn't), we need to get it there as soon as possible.

This is what the "computer-based patient record" (CPR) initiative is all about. As I am sure you are becoming aware, this is a project developing on many fronts--small entrepreneurial companies working with groups to write software; big traditional mainframe companies retooling and adding new, more user-friendly "front-ends" to their old systems; as well as "think tanks" of physicians and engineers designing new-age systems to organize our thoughts and practices. Though the CPR projects represent, perhaps, the central core of the informatics revolution, they are but one of a number of initiatives.

The next level of sophistication of organized data is the use of data to form "rules." For example, we collect potassium levels; we use the drug, digitalis. There is an important link here. If the potassium level drops, there is an effect on the dosage of digitalis. Why not use the information system to track these two variables, apply a set of rules we define, and give us "intelligent" messages such as, "Suspend dose for two more days and repeat level." Imagine this use of rules that connect our data like a huge spider web.

The vision of applying practice standards and protocols to our clinical work will involve the use of "rules-based logic" as described above. These expert systems are an example (and tiny subset) of the field known as "decision support" technology. Again, there's nothing new here. Industrial applications have existed for years where decision support is used to determine the best pathway through a process on an assembly line. The "assembly line" in our case is the clinical pathway, with diagnoses, drugs, and demographic data as the relevant data.

Another crucial concept in this overview of medical informatics is how the information flows around our organizations and to and from other organizations. Before, we exhorted physicians to write neatly so we could understand their charts, because the chart was the repository of the data. No more. Now one could realistically ask, "Exactly where does the information on my patient reside?" In fact, in the world of the paperless computerized record, the data reside in tables scattered all over the database, and in multiple databases. It doesn't really come together until you ask a question (a query) that requires the computer to assemble the "name" from one table with the "lab data" from another and the drug list from yet another. The "view" you have as the observer of the computer screen is deceiving. The "chart" is now "virtual" reality; only the discrete data elements are real. The next step is to see all of this information flowing around our organizations on a sea of wires, carried by microwave signals or along fiberoptic networks that President Clinton has promised to build in our lifetimes.

Now one can imagine the power of the concept--to be able to send documents as well as graphic images of xrays, CT scans, etc. around the city or the world, following the patient as he or she relocates or being made available to a consultant for review. It is definitely worth our while to learn more about the role of telecommunication and networking in our exploration of medical informatics.

A final point in this regard. This journal is directed toward the intelligent user. It is not limited to the designer, engineer, or "gearhead." Let me explain by analogy. To drive a high performance car, to appreciate the subtle and not so subtle aspects of that performance, and to have a general working knowledge of its specifications and functions, it is not necessary to fathom all that lies "under the hood." Another analogy-- "I can't define pornography, but I know it when I see it!"

Those of us who have spent considerable time working closely with computer engineers and technicians develop an enormous respect for their talents. Through this knowledge, we know what to ask for, what language to use in describing our needs, etc., but we will never be all that knowledgeable "under the hood." The end user must appreciate that, beneath the "user layer" we see on our desktop computers or the terminals connected to our mainframes, lie hidden layers of application software, operating system software, network software, and, finally, "machine language" written in codes that we will never have to understand.

This issue is raised to quell fears I have seen in some new users, who, as physicians, expect to be able to understand the minutiae underlying all technical things. Here, they meet their technical Waterloo! No matter-- like the connoisseur, appreciate the drive and learn to handle difficult turns without dwelling on technical issues. The technologists look to us to spend our time trying to figure out what is worth knowing--the executive view.

This is a hallmark of the "new age" medical informatics expert. In the past, we were asked to help interpret reams of computer reports produced by the finance department. Now we demand our own workstations where we write our own queries of data to meet needs as we find them in our work. More and more of us daily join the ranks of the "computer literate," although, as in most things, there is a difference between users and designers of systems. In any case, we will be prepared to make our mark on the management of an increasingly complex clinical and administrative world.

In the January 1994 issue of Physician Executive, Dr. Lowenstein and a distinguished collection of physician executives and other informatits experts will further explore this growing field and its implications for the medical management profession and the delivery of health care services.

Edward Lowesntein, MD, FACPE, is Medical Director, Orlando (Fla.) Health Care Group, He is Chair of the College's Forum on Medical Informatics.
COPYRIGHT 1993 American College of Physician Executives
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lowenstein, Edward
Publication:Physician Executive
Date:Nov 1, 1993
Previous Article:Quality management in ambulatory care: the future is now.
Next Article:Toward the new pattern of medical practice.

Related Articles
Depression after divorce: male call.
The Making of Modern Marriage: Matrimonial Control and the Rise of Sentiment in Neuchatel, 1550-1800.
Genuine letters help democratize our debate: letters give a window into how regular folks see the events of the day.
Assault ambiguity: marriage bans and domestic abuse.
Notes from the field.
Hospital Battlefield.
What will this election be about?

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters