The Last Priesthood:
The Coming Revolution in Medical Care Delivery

Stephen S. Hyde

Stephen S. Hyde is the founder and president of IG Ventures, Ltd. He was the founder of Peak Health Care, Inc., a multistate HMO company.

During the past few years, very significant developments in artificial intelligence have yielded computerized expert systems that approach or equal medical doctors' levels of expertise in diagnosing and prescribing treatment for a wide range of medical problems. For example, a company named N-Squared Computing in Silverton, Oregon, sells a package of software called Symptom Analyst Plus. It permits health caregivers to perform the differential diagnosis of any combination of 537 signs, symptoms, and laboratory tests for 337 diseases and disorders. The possible diagnoses are displayed in rank order according to their levels of support. Each such diagnosis can then be compared with its disease reference in the data base, which displays all the characteristics of that disease. This software program is contained on a single 3-1/2-inch disk and costs $295.

A recent review of the literature turned up more than eighty such programs, covering a broad range of medical disciplines, including backache, infectious disease, psychiatry, heart disease, coronary artery disease, sports medicine, pediatrics, emergency medicine, dermatology, dentistry, diabetes, general medicine, geriatrics, hypertension, medical history, allergy, endocrinology, enterology, hematology, laryngology, nephrology, neurology, obstetrics and gynecology, ophthalmology, otology, proctology, pulmonology, rheumatology, urology, anesthesia, and blood chemistry. Currently, almost all such software is offered only to medical doctors.

The availability and comprehensiveness of such software provides strong evidence that many, if not most, nonsurgical problems lend themselves to diagnosis and treatment recommendation by computer. I call this the algorithmization of medical practice.

Indeed, it is not unreasonable to forecast a full range of nonsurgical primary physician expertise being available to much less intensively trained personnel in the form of inexpensive, widely available software running on home computers linked with on-line computer networks.

I would like to suggest just one of the many directions such technology could take that could result in major societal benefits. Such benefits include large health care cost savings, better integration of prevention regimens with treatment protocols, immeasurably improved medical research data, markedly increased availability of care in rural and third-world settings, and a heightened personal awareness of and responsibility for one's own health.



The vehicle for such benefits would be a new category of certified health care provider called the certified home practitioner (CHP). Certification would be available to essentially any person who successfully completed a formal course of study lasting perhaps two to six months and who then passed a certification examination. The formal courses would focus on four main areas: first, practical anatomy and applied physiology and pharmacology; second, physical assessment, in which the student learns to examine a patient (this typically requires only a single semester of nursing or medical school); third, basic treatment procedures such as giving injections and first aid; and finally, the use of certified, computer-based, medical expert systems.

Once certified, the certified home practitioner would--within carefully circumscribed limits--be able to practice medicine, but only upon herself (use of the female gender is arbitrary), her legal spouse, their children, and their parents. She would be able to examine her patients, order diagnostic tests, prescribe drugs, administer treatment, and refer her patients to medical doctors for diagnosis and treatment beyond her allowed scope of practice. She would not be allowed to charge a fee. All care would be under the specific direction of certified expert systems software, which would embody conservative, current, generally accepted medical practice and expertise.

The key element is this conservatism. Uncertain diagnoses where incorrect treatment could jeopardize the patient would trigger an expert system directive for the CHP to refer the patient to a medical doctor. Likewise, such referrals would be directed for any condition that requires a medical doctor's intervention, such as possible malignancies, conditions requiring hospitalization, surgical procedures, or heart problems.



I believe that there would be a number of significant benefits from such a program. First, creation of the CHP program would extend the trend of increasing personal responsibility for one's own health. Already, a very high percentage of health care is personally administered. Of the remainder, approximately 80 to 90 percent can be performed by a competent primary care physician (family practitioner, pediatrician, internist, and, sometimes, gynecologist). Between 25 and 44 percent of all physician office visits are estimated to be medically unnecessary. The CHP could eliminate virtually all of those. At least a third of all hospital admissions, hospital lengths of stay, emergency room encounters, and medical referrals to specialists are unnecessary and preventable by a properly trained primary care physician. The CHP, directed by her expert system, would replace not only much of the primary care physician's diagnostic and treatment role, but much of his role as utilization control gatekeeper in directing patients to the other components of the health care delivery system, such as specialty physicians, lab tests, x-rays, prescription drugs, and hospital emergency rooms. Thus, there would be major cost savings from the reduction in unnecessary utilization of medical care.

Also, since we are seeing an increasing number of prescription drugs gaining over-the-counter status, such as cortisone, antifungals (Lotrimin), certain antihistamines, decongestants, and pain medications, and prospectively Tagamet and its cousins and Feldene, the antiinflamatory, the CHP's authority to prescribe drugs would represent a prudent extension of that trend. Such increased authority would also serve to minimize the now significant tendency of the patient either not to get his prescriptions filled in the first place or not to continue the full course of drug treatment after he begins to feel better.

An additional benefit is that the expert systems could incorporate preventive and life-style instruction and record keeping. The patient-specific software would include an expanding, time-tagged data base on each patient over the course of his life. Data would be input annually, monthly, daily, hourly, or even continuously through remote monitors, as indicated and desired. Those data would be continually evaluated by the expert systems for indications favoring health-inducing behavior modification, identification of cause and effect (for example, relating one's stress EKG interpretation to his long-term dietary practice), or identifying recurring patterns of illness such as stress-induced illness. Dietary weight loss could be monitored daily to provide positive feedback and to monitor for adverse reactions. Life expectancy could be continually updated, as measured by behavioral factors such as drinking alcohol, smoking, diet, sleep, exercise, stress, and occupation.

Another benefit is that the CHP would reduce the delay between detection of symptoms and subsequent diagnosis and treatment because there would be no waiting for a doctor's appointment.

Better medical records could result. A recent study indicates that, currently, 11 percent of lab tests have to be reordered because the results are not in the patients' records. In addition, 10 percent of the time a patient's age is not included in the medical record, 27 percent of the time the doctor fails to record the patient's chief complaint, and 40 percent of the time a doctor does not record his diagnosis. The CHP's computerized record keeping system could significantly reduce, if not eliminate, those omissions.

As much as half of all medical care provided to a patient is estimated to be consumed in the last year of his life. The presence of a family CHP, providing home treatment in a family setting, could greatly reduce that expenditure, as decisions regarding living wills, preferred treatment settings, and the desire for heroic, life-saving measures could be made in advance within the family. One or both members of elderly couples would be ideal candidates for CHP training and certification.

Many of the medically indigent could gain improved access to health care services by virtue of inexpensive scholarship programs to provide CHP training and at least a minimal degree of coverage for basic diagnostic and therapeutic services.

As CHPs become more widespread, one can envision the creation of a whole new industry supplying high volumes of low-cost data collection, diagnostic, and treatment equipment to the CHPs. Such equipment could include $50 hand-held devices to record diet, symptoms, and events throughout the day for connection and entry into the computer at the end of the day, or hand-held $100 blood analyzers with $.50 cartridges requiring a couple of drops of blood, or rentable pocket-sized recording EKG machines, or a full range of $2 diagnostic test kits available from the local pharmacy. Currently, only pregnancy and diabetes test kits are available because the American medical profession argues strongly that individuals need counseling when they take tests for diseases. Perhaps we would see equipment cost curves following those of computers. Twelve years ago my first home computer cost me $3,500. Its equivalent today would be hand-held and cost perhaps $250.

Since the CHP's computer would also receive and hold data on patient progress and response to treatment as frequently as hourly, or even continuously, drug efficacy research on a national scale would become much more accurate and much less expensive. Outpatient triple-blind studies could be expeditiously and inexpensively conducted, with many more participants and data points than are currently practicable.


Questions, Risks, Problems, and Arguments

What are some of the problems and risks of the CHP proposal? I have thought of several. Consider, for example, computer malpractice. Who is liable when the software instructions are in error and the patient is harmed? That would be minimized by having each discrete algorithm reviewed and certified by a panel of medical experts, who themselves would be exempt by law from malpractice liability. Second, only very conservative diagnostic and treatment protocols would be allowed in the expert systems. Third, when errors do occur, I would suggest that actual damages (not pain and suffering) be determined by arbitration and paid from a fund created for that purpose and perhaps funded by a percentage of software sales. Payment of patient's legal fees could either be nonexistent or tightly capped.

Next consider CHP malpractice. What happens when the CHP commits malpractice? Remember, the CHP is allowed to treat only herself and immediate family members. Thus, CHPs could be exempt, by law, from malpractice liability, except where gross negligence or intentional harm was involved.

Can someone really be taught how to examine patients and identify symptoms in only two to six months? One can if she does two things. First, she should limit the universe of patients to immediate family members. The CHP may not be able to divine the subtleties of competing, multisystemic diagnoses, but she can quickly learn what normal looks like for the patients she is treating. Second, she should limit the range of diagnoses and treatments to those that can be safely and effectively taught within the training period. According to my physician sources, much can be taught in two to six months.

Then there is the potential of physicians' opposition. With a properly structured CHP program, many physicians suggest that they would be less likely to oppose the CHP program. For example, the CHP would relieve the medical doctor of unnecessary, uninteresting patient complaints. Second, the physician could be a fully participating member of the CHP program. Third, physicians could handle more problems by phone, a service for which they could be paid.

Then there is the potential for FDA opposition--a major concern, as the FDA is not generally recognized as progressive. But consider. There are currently two types of legal drugs: over-the-counter and prescription. Any six-year old can walk into a pharmacy and purchase any over-the-counter medication. But to buy a prescription drug, one must have a prescription written by somebody who has had seven years of intensive medical training. There is nothing in between. Perhaps there should be.

Who would certify the medical expert systems and the CHPs? One way would be to follow existing models of state licensure, whereby state-appointed boards would set standards and set up regulatory mechanisms. That function could also be done privately, perhaps with some sort of state-granted franchise.

Who would create and distribute the CHPs' medical expert systems software? Private vendors would--with appropriate regulatory oversight.

Would there not be serious privacy issues, given that the medical expert systems would reside in computer networks. Yes, but the issues of network integrity and security are ones that are well advanced.

Who would become CHPs? Nurses and other health care professionals who are not medical doctors would perhaps be the first, and most enthusiastic, candidates. The depleted but still significant ranks of nonworking spouses and mothers would be another obvious inclusion, as would many fathers. Rural families would benefit from having a CHP family member. Retirees as CHPs could have a dramatic impact on the costs of providing health care to the elderly. Employees, encouraged by employer programs and financial incentives, would comprise perhaps the largest eventual component of the CHP complement. Training for welfare mothers and the medically indigent is certainly not out of the question. HMOs would have a strong incentive to encourage and assist covered members to become active CHPs.

Another issue that arises is the current financing system incompatibility. Who would pay for all this--especially the lab tests, x-rays, and prescription drugs ordered by the CHPs, not to mention the software? The training and software would be paid for by the CHP, with augmentations by scholarship, employer assistance, and government and third-party payer assistance. The tests and drugs would be covered by most health benefit plans, after effective and repeated demonstration of the cost-effectiveness of the program. Again, HMOs would probably be at the forefront of coverage and encouragement.

Who benefits from the creation of a CHP program? Beneficiaries would include the patient, the CHP, the employer, the HMO or insurer, federal, state, and local budgets, the purveyors of diagnostic services, the drug companies, and the economy in general because of a healthier, more productive work force. There may even be political advantages for astute backers of such a system with its strong profamily orientation.

Who loses from the creation of such a program? I believe the losers would be marginal at worst. Perhaps some doctors would find themselves underutilized. Hospitals would probably suffer lower occupancy rates, although attribution would be difficult, given the welter of bed-emptying measures extant today. Such losses would be short-term and ultimately beneficial.



The certified home practitioner concept, with its multitude of variations, is, I believe, a logical extension of two trends in health care: first, the algorithmization of diagnosis and treatment of disease, and second, the increasing recognition of the role of personal responsibility in maintaining and promoting one's own health.

Today's medical profession is perhaps the last priesthood that operates according to the principles of the medieval guild. That is, it enjoys a state-enforced monopoly, requires excessively long education and apprenticeship, employs arcane language, and exhibits a circle-the-wagons mentality (to mix my metaphor) when faced with any threat. It has been nearly five hundred years since Martin Luther told people it was permissible to talk to their deity directly without priestly intercession. It has been barely fifteen years since the computer priesthood was opened to mere mortals. Perhaps now is the time to use the twin wedges of education and technology to pry open the bolted door of medical practice by at least a notch.

Regulation is published four times a year by the Cato Institute. Editorial and business offices are located at 1000 Massachusetts Avenue, N.W., Washington, D.C., 20001. For subscription information, please write to Circulation Department, Cato Institute, same address, or call (202) 842-0200. Send email inquiries to, or subscribe online via the World Wide Web at:

| Regulation | Home | Order Regulation | Publications | Search