TYPE OF PRACTICE by physicians has changed greatly in past three decades. There are now many more specialists and more doctors in hospital service and working for the Government. Information is from American Medical Association and U.S. Public Health Service.

In theory and in hope the diminishing numbers of general practitioners, who exemplified the virtues of the personal doctor, could provide the continuity and discriminating guidance to the confusing multiplicity of specialist services that lie beyond the scope and judgment of the patient. The medical schools are often blamed for the disappearance of the general practitioner, and unfairly. The expansion of medical knowledge has forced the schools to precede the profession into specialization. Against this pressure many schools have experimented with comprehensive patient-care clinics to help the student to integrate the training he receives piecemeal from specialists representing fields as diverse as biochemistry and psychiatry.

Medical students, in spite of the attractions of other careers in the sciences, are still a highly selected group. In facing the challenge of medical education, however, they run the gamut from brilliant to plodding. The able and promising are welcomed into the ladderlike training programs of the university hospitals that lead to certification in one or another specialty. As a study by Fremont J. Lyden of the Harvard Medical School has shown, students of higher rank obtain teaching-hospital internships more frequently than poorer students and go on to further postgraduate training with substantially greater probability than those who have interned in nonteaching hospitals. The least able must be content with more modest institutions that provide shorter training with more meager educational resources. Although theory would dictate that the less able need the longer and better training, the conflicting interests of education result in the recruiting by the teaching hospitals of the best for themselves. This is a practice that is not exclusive to medicine.

Several years ago a general study of general practitioners was undertaken at the University of North Carolina by Leon P. Andrews, Bernard G. Greenberg, Robert S. Spain, and myself. We expected to find variations in skill, but we were entirely unprepared for the extremes we found. The best of the practitioners performed at a level that would have been acceptable in the outpatient clinic of a university hospital. Many were not conspicuously skillful but were adequate. At the other end of the scale there were some whose practices would have been unsatisfactory in a senior medical student.

Since the general practitioner daily sees a variety of diseases, trivial and serious, his major problem is to make a diagnosis or at least to recognize a dangerous illness when it is present. In the study we observed, randomly selected doctors at their work over a period of 32 to 4 days to see how often and how well a clinic history was taken, a physical examination was given and a few common laboratory procedures were used when patients presented such potentially serious complaints as chest pain, bleeding, or weight loss. Description and classification of these procedures, made in many cases independently by more than one of us, proved to be highly reproducible. Furthermore, the doctors' performances turned out to be markedly consistent from day to day and from technique to technique. If we observed a doctor taking the clinical history of a patient in a thorough fashion, we could predict that he would make a good physical examination and order a sound laboratory procedure. This made it possible for us to classify the doctors by their skill and to study the elements in their background significant to their performance.

We found that a number of variables were related to excellence. The doctor's record as a medical student proved to have some prognostic value, although there were enough exceptions to make it perfectly clear that a poor student could become a good doctor and vice versa. The extent of hospital training in internal medicine proved to be one of the most important determinants of skill; the average performance was consistently better and the variation less in those whose training was the most prolonged. Some observations could not be correlated with performance on a quantitative scale. Among these were the depth of a doctor's interest in clinical medicine; this seemed to vary, in individual physicians, from intensity to boredom. One of our more significant quantitative findings, on the other hand, was that a doctor's success, as measured by the number of patients he saw during the course of the week, bore no relation to his knowledge and skill. The choice of a physician by the patient would seem to depend on factors other than these, which are obviously difficult for the layman to judge.

Kenneth F. Clute, of the school of hygiene at the University of Toronto, has applied the techniques developed in our study to an evaluation of the work of general practitioners in several areas in Canada. He found the same marked variation in performance, with good performance plainly correlated to internship and residency in teaching hospitals and poor performance to postgraduate experience in nonteaching hospitals. What is more, the duration of training in the teaching hospital showed a clear positive correlation to the doctor's skill. No such connection could be detected where training had been secured in nonteaching hospitals-an indication of the relative ineffectiveness of these institutions.

The public meanwhile has been working out its own adjustment to the disappearance and decline in the status of the general practitioner. H. Jack Geiger, formerly in the Department of Preventive Medicine of the Harvard Medical School, has conducted a study of the doctor-patient relationship in an economically mixed suburb of the type in which Americans are living in increasing numbers. He found that the family doctor who delivered the baby, set the bone, removed the tonsils, and gave comfort where he could do no more, had ceased to exist in the experience of most of the population. Working-class families still have a "family doctor" in the sense that they secure most of their medical care from general practitioners. The remainder of the population seek in their own way to obtain the best of modern medicine by employing specialists in internal medicine, pediatricians, obstetricians, orthopedists, and others--each to treat a pain or problem the patient believes to be "in his field."

PLACE PATIENT IS SEEN has also changed. Chart shows sites of visits by or to the doctor outside hospitals. Decline of home visits reflects greater efficiency in medical practice.

PROFESSIONAL QUALIFICATIONS of physicians who performed surgery were studied in 1957-58 survey of nearly 3,000 families representing a sample of U.S. population. Study was made by Health Information Foundation. Forty-nine percent of the surgical procedures performed on these people were done by physicians having the special credentials indicated. (The sum is less than the parts because some doctors are certified by two boards.) Of physicians claiming some surgical specialization, 21 percent had no special credentials.

ALIVE PHYSICIANS PER 100,000 CIVILIAN POPULATION (1859) PHYSICIAN-POPULATION RATIO for each State in 1959 shows that only 11 States are above the national average (vertical broken line at 119). The States with the more favorable ratios also tend to have a larger proportion of specialists and of physicians who are employed in hospitals.

Good medical care requires ample physician time. Under the combined pressures of eagerness on the part of the patients and the economic arrangement of payment by the visit, however, general practitioners see between 25 and 35 patients per day on the average in different settings. Clearly the number must reach a point where the carefully detailed history and examination necessary to make a diagnosis becomes impossible. That the number is excessive is suggested by comparison with other countries. The U.S. citizen visits a doctor 5.3 times per year on the average, in contrast with the British and Swedish figures, which are 4.7 and 2.5 respectively. The lower death rates and the greater longevity of the British and the Swedes suggest that health need not be a hazard.

Studies conducted by Vergil Slee, of the Commission on Professional and Hospital Activities, and Paul A. Lembcke, of the University of California Medical Center in Los Angeles, provide an objective index of the variation in the quality of care obtainable from the assemblage of general practitioners and specialists available in the community. Slee examined the records of 15 hospitals to determine how many operations for appendicitis had been justified by the finding of diseased tissue in the pathology laboratory. The percentage of diagnoses thus confirmed varied from about 70 percent in one hospital to less than 20 percent at the other extreme. Needless to say, 20-percent correct diagnosis is not good. Lembcke, taking the same disease as a criterion, found that the number of people operated on for appendicitis in more than 20 different hospitals varied from 2.9 to 7.1 per 1,000 population per year. It is, of course, unlikely that the incidence of the disease is so variable. He has also shown that a study of surgery in hospitals sharply reduced the frequency of operations that can be criticized as too extensive, too restricted, or unnecessary.

Other studies show that the quality of care depends not only on the native ability and training of the doctors but also on the organization of medical services, the availability of facilities and the intellectual stimulation of colleagues. One of the most pertinent of these studies was conducted by Ray E. Trussell, now Commissioner of Hospitals of the City of New York, with his associates at the Columbia University School of Public Health and Administrative Medicine. Their findings were the subject of local controversy that drew attack on their methods; they are supported, however, by other data. Trussell and his associates concluded that about 80 percent of patients who are cared for in hospitals affiliated with medical schools receive good or excellent treatment. At the other end of the scale, in hospitals that have no internship or residency programs and are not accredited to give such training, only a third of the patients receive such care. Significantly these investigators also found that specialists certified by appropriate boards and societies and working in good hospitals give good care, whereas similarly accredited specialists working in poor hospitals give care no better than that rendered by doctors with no evidence of such training.

These conclusions are supported by the work of Odin W. Anderson of the Health Information Foundation and Jacob J. Feldman of the National Opinion Research Center. About half of the surgery in the United States, they estimate, is done by doctors who are neither certified as specialists nor members of the societies that require evidence of competence or training. Doctors who have qualifications show up more frequently in accredited or large hospitals. Anderson and Feldman found a similar distribution of obstetrical talent and training. In the Nation as a whole 38 percent of the deliveries are performed by obstetricians who are certified or who give sufficient time to their practice to be classified as specialists. "The larger the hospital," these authors observed, "the more likely that a specialist attends the obstetrical case."