RESEARCH WITH A DEFINED POPULATION

133

Feldman, repeated cost analyses have shown that our direct operating costs can be maintained at about $12 for the laboratory phase of the multiphasic examination.

The long-term impact on outcome of patients is being determined by our multiphasic evaluation study conducted by Dr. Savitri Ramcharan. After a five-year follow-up, we are beginning to observe demonstrable effects, such as a statistically significant decrease in disability and time lost from work, for middle-aged males. These preliminary results are being prepared for publication. The study will be completed in 1974.

HEALTH Surveillance System

Our experience with health surveillance of large numbers of patients, through the utilization of our automated multiphasic screening program, led us naturally to the broader concept of health surveillance of a defined population for all important health hazards.

Since 1967, we have conducted a pilot study of an adverse drug reaction monitoring system in our San Francisco medical center. For an enrolled membership of 120,000 people in San Francisco, all relevant outpatient medical experience is entered and stored in our computer. Accordingly, physicians' diagnoses for 2,000 medical office visits, and drug data for 1,200 prescriptions each day, are added to our computerized medical data base.

By comparing the frequency of diagnoses reported in patients receiving specific drugs to the frequency of the same diagnoses in patients not receiving these drugs, our epidemiologist, Dr. Gary Friedman, can identify those diagnoses which constitute possible adverse drug reactions. He can also establish the incidence rates of adverse reactions, using the defined population as the denominator.

Our next interest was to extend our health surveillance system to include the monitoring of environmental health hazards. We have been conducting preliminary pilot studies to test serum on all multiphasic patients for important trace metal contaminants. We hope soon to be able to monitor, in both our Oakland and San Francisco populations, the blood levels for mercury, cadmium, copper, zinc, manganese, nickel, and lead. Thus, for example, the introduction of legislation to control the amount of lead in gasoline is expected to decrease the amount of this contaminant in the air. The effectiveness of such control measures can be monitored by periodic determination of the average level of blood lead in a defined exposed population.

HEALTH SERVICes Research Centers

In 1968, HEW's National Center for Health Services Research and Development established seven Health Services Research Centers in the United States. One of these centers was established in our Northern California Region and one in the Oregon Region.

The Health Services Research Center in Portland, under Dr. Merwyn Greenlick, coordinates the activities of the medical care research unit of the Kaiser hospital in Oregon, with major research and educational resources of four major universities. This joint effort involves the activities of senior investigators from Portland State University's Center for Sociological Research, Reed College, the University of Victoria, and Oregon State University's School of Pharmacy.

134

PROGRAMS IN EDUCATION, RESEARCH, AND DEVELOPMENT

The Portland center investigates significant individual, family, and organizational determinants of the medical care utilization of a known and identifiable population, and provides demonstrations in medical care which will implement these research findings.

Our Southern California Permanente medical group collaborates with Dr. Lester Breslow's Health Services Research Center at the University of California in Los Angeles. It has been especially active in conducting research in methods for best utilizing allied health personnel.

The Northern California Center. The objective of our Health Services Research Center in the Northern California Region is to develop a computerized medical data system and to establish a medical data base for both medical care and for health services research. As a result, there was established the capability for a centrally stored computer medical record for every one of our almost one million health plan members. The organization of our computer data files is based upon an integrated, continuous, life-time medical record. All records are individually retrievable by each patient's medical record number on a direct access basis within one-half second. Each individual's record within the computer's direct access storage facility is kept and moved together as a continuous string of data; that is, it is never divided for overflow, or any other reason.

Using our San Francisco medical center as the pilot hospital, we are implementing in a modular manner, a pilot medical data system, under the supervision of Dr. E. E. Van Brunt (Figure XII-3). The projects we have already described provided the first components. Our multiphasic testing laboratory was the first operational unit of the system. Diagnoses and drugs from our adverse drug reaction monitoring system were included next. The outpatient diagnosis recording system was initiated in November, 1967. San Francisco physicians note on special forms their diagnoses and any important procedures provided for each office visit. The data are subsequently entered into the appropriate patient computer record by clerks using on-line typewriters.

Within the San Francisco outpatient pharmacies, on-line typewriter terminals were installed in July, 1969. Pharmacists are entering complete drug data, for each prescription dispensed, directly into the centrally stored patient computer record. Subsequently, upon entering the patient's refill number, a new label is printed and the cumulative number of dispensed prescriptions is automatically recorded, thus facilitating the refilling of prescriptions.

A computerized laboratory data collection and reporting system with an associated patient bed census system is being implemented at the present time. Larger automated laboratory components, such as the AutoChemist and the Coulter Counter, automatically generate machine-readable punched card input. Other laboratory test data are entered by technicians using on-line keyboard terminals.

Still under development is a system of 24 visual display terminals for the nurses' stations of the San Francisco hospital. The system is currently being tested in the Oakland computer center and is scheduled for installation in the San Francisco hospital at the end of 1971. By means of these terminals, physicians will be able to enter medical orders and patients' diagnoses directly into the computer by using a light pen to select appropriate items from the visual displays. Using the same method, nurses

RESEARCH WITH A DEFINED POPULATION

135

will be able to chart administered drugs directly into the computer, which in turn will generate listings of such patient data by electric typewriters at the nursing stations either periodically or on demand.

Following completion of the fully operational pilot medical data system, an evaluation will be conducted by Dr. R. Richart as to its efficiency, cost, and impact. Since our own baseline studies show that the proportion of time for all health personnel, spent in patient care-communication was 25 to 30 percent, it is essential that improved medical information systems be developed and evaluated.

A NEW ENTRY SYSTEM TO MEDICAL CARE

A new systems approach to medical care delivery appears now to be of high priority. If the components of medical care which have been developed during the past ten years are to be utilized effectively, an improved systems engineering for interfacing operational modules would be advantageous. In July of 1970, Medical Methods Research Department obtained a contract from the National Center for Health Services Research and Development to design, test, and evaluate a new concept described by Dr. Sidney R. Garfield in the April, 1970 issue of Scientific American. The goal, in essence, is to attempt to screen patients as to their health needs and to direct them into the proper components of the delivery system which will most effectively utilize health personnel time and ability.

In our current system, as elsewhere in the nation, the demand by the well, the worried well, the early sick, and the very sick for physician appointment time has never been regulated. Consequently, response to such demands for sick care services is unrelated to priority of need. Dr. Garfield suggests that the result is an overloaded system wherein the excess use of physician time by the well and the worried well becomes an actual barrier to the needed ready-access to physician appointments by the sick-as exemplified by long waiting times for physician appointments. He proposes

FIGURE XII-3 Kaiser Foundation Health Plan, Department of Medical Methods Research, Pilot Medical Center Data System

136

PROGRAMS IN EDUCATION, RESEARCH, AND DEVELOPMENT

to correct this problem by employing allied health personnel under physician supervision, supported by health testing, health education, and computer data processing (Figure XII-4).

This new entry system project will be operational in the Oakland facility by October of 1971. Triage procedures are being developed for suitably directing patients to the multiphasic health testing services. To meet these requirements, the multiphasic health checkup is being appropriately modified to provide a new entry system to medical care that can separate the variable mix of patients into three basic components: the well (about one-third of all current multiphasic patients), the asymptomatic and early sick, and the overtly sick (Figure XII-5). A multiphasic health testing unit is also being installed for children, under the direction of Dr. R. Bachman.

Additional new services are being developed under the supervision of Dr. K. Soghikian, which include health care services for the well, and preventive maintenance services for the asymptomatic and early sick. This leaves the conventional sick care services for the overtly sick. The health center is being augmented in Oakland so as to add these necessary components of health education, prevention, and monitoring care (to take care of the well, the worried well, and the early sick). Together, this will provide a demonstration model of this new medical care delivery system within the Oakland medical center.

FIGURE XII-4 Kaiser Foundation Health Plan, Department of Medical Methods Research, Garfield Model for Health Care Delivery

RESEARCH WITH A DEFINED POPULATION

137

FIGURE XII-5 Kaiser Foundation Health Plan, Department of Medical Methods Research, New Medical Care Delivery System (MCDS) Study, Patient Flow Pattern

The extent to which this new delivery system achieves its objectives of improving services to patients and conserving physician time will be evaluated by Dr. R. Richart. A cost-effectiveness study of the new medical care delivery system as compared to our current system for providing medical care will be conducted. The defined Oakland population of 120,000 members has been equally and randomly divided by systematic assignment of their medical record numbers. One-half will be directed through the new entry system into the new health care and preventive maintenance services, and the other half will continue to be provided only our conventional sick care services. Baseline measurements have already been completed. It is expected that this large controlled study will be conducted over the entire year of 1972.

TRAINING

In addition to the activities described, for the past few years our Medical Methods Research Department has conducted training programs in health services research and computing science in collaboration with the University of California in Berkeley. M.D. postdoctoral candidates receive their academic work at the Berkeley campus and all of their field training within Medical Methods Research. We are presently exploring with Stanford University a similar collaborative arrangement.

The research center in Portland has also provided a significant emphasis on health services research training. They have instituted a joint Ph.D. program in urban sociology and health services research with Portland State University.

Due to the increasing importance of health services research and the scarcity of qualified investigators, training in health services research will acquire more importance with time.