A non-Federal private hospital, on the other hand, is established to provide in-patient medical care to a population who, in general, have other sources from which they can receive medical attention. When a civilian is ill enough to require hospital care, and can afford to pay all or part of the cost, he generally is admitted to a hospital, accompanied or preceded by his physician who has an exact and immediate knowledge of the history and current status of the patient. Treatment can begin promptly and a whole series of economic factors are immediately put into play, the action of which tends to force the patient out of the hospital at the earliest possible moment. In addition, the patient's physician is continuously available on call and can visit the patient at his home. Therefore, the convalescent period is usually not spent in hospital.

Veterans' Administration hospital services are somewhere between these two extremes. In most instances the admitted patient is a medical unknown. He has not been seen prior to admission, and he may have traveled a long distance to be admitted. The Veterans' Administration is also bound by legal requirements which make it very difficult to provide convalescent care outside the hospital. Moreover, the VA patient is discharged after receiving treatment for all disabilities which are amenable to therapy, not merely the one which acted as the precipitating cause for hospitalization.

Under these circumstances, it may be expected, in general, that a patient with a particular illness will be hospitalized longest in a hospital of the Armed Forces and for the shortest period of time as a paying patient in a private hospital. His stay as a Veterans' Administration beneficiary would be somewhere between these limits.

To go further, a Federal hospital is rarely directly comparable with a private hospital because each treats different types of patients. In the Veterans' Administration there are institutions primarily for the care of mental or tuberculosis patients, as well as G. M. and S. hospitals which, in many instances, have NP and TB services-facilities not usually found in the private general hospital.

APPROACHES TO THE PROBLEM OF LENGTH OF STAY

The Veterans' Administration has been actively aware of the problem of length of stay since about 1945. Then, in the face of a rapidly rising demand for admission, bed capacity was expanding too slowly. The Veterans' Administration was forced to think about the length of time that general medical and surgical patients were retained in hospital. Certain steps had to be taken to reduce length of stay. A policy was established which brought to the attention of the medical staff in hospitals the necessity of employing to the greatest extent possible the practice of placing patients on convalescent leave rather than of retaining them in the hospital until they were ready for discharge.

More recently, the Hoover Commission and the Bureau of the Budget prompted a general Federal interest in this matter, and during the past year the Veterans' Administration has worked with an interagency Federal committee seeking further to define both the problem and the methods for its resolution by means of surveys of selected Federal and private hospitals.

The most meaningful way of proceeding is probably to describe a few of the things which were found at one of the larger VA G. M. and S. hospitals located in a large metropolitan area and to present a comparison of its experience with that of two excellent non-Federal institutions located in the same city. Since the time of this study, the fact that certain generalizations may be drawn has been established beyond a doubt on the basis of further investigation.

In setting up the study it was necessary to face the problems which were mentioned earlier and answer some additional questions. First of all, what types of patients were going to be studied? It was necessary to define the study group in such a way that it would be reasonably certain that any differences which were observed might be ascribed to administrative or professional practices in the care of those patients and not to extraneous forces. Consequently, it was decided to study the experience of patients for whom the clinical record indicated medical care was given for only one condition and whose hospital course was not affected by any major complications.

Chart 3 summarizes the findings for patients admitted for herniotomy in the selected VA hospital. Take a specific example: Case 1-a 37-year-old patient in the VA hospital. On the day of admission a workup, consisting of blood tests and urinalysis, was performed, and a chest X-ray given. On the second day serology tests were made, and surgery was performed. On the 10th day he was discharged.

Case 4 was a 27-year-old male, admitted on a Thursday. For him, workup was not begun until the second day. X-ray was made on Saturday. Sunday passed, and the patient was not operated on until Monday. He was discharged on the 12th day after admission. In this instance, there is some evidence of one of the problems which we face-mischeduling of the admission of the patient with respect to the availability of operating-room facilities, or with respect to the fact that a weekend will intervene between admission and initial surgery. It may be noted in the next to the last column that this patient lived only 20 miles from the hospital, and so any argument that he lived so far away that he had to be admitted promptly is not particularly pertinent.

Let us look next at case 14, a 34-year-old male who was not operated until the 10th day after admission. In this case, however, secondary diagnoses of arthritis and backache account for the lapse of nine full days after admission, during which various diagnostic procedures were performed, with respect to the secondary diagnoses. In this instance, note that the hospital placed the patient on a 2-day pass midway through the diagnostic period, and in so doing made some attempt to reduce the cost of hospitalization. This pass was given on a weekend. A summary of this experience indicates that, out of 16 patients, 1 was operated on the second day after admission, 3 on the fourth day, 1 on the fifth, 3 on the sixth, 4 on the seventh, 2 on the eighth, 1 on the tenth, and 1 on the eleventh. With respect to length of stay, 1 of these 16 patients stayed 10 days, 2 stayed 11 days, 1 stayed 12 days, 4 stayed 14 days, 3 stayed 15 days, 2 stayed 16 days, and 1 each stayed 17, 18, and 19 days, respectively.

Chart 4 refers to a municipal hospital in the same city. It will be noticed by comparison of the location of the shaded boxes in this table, to the location of similar boxes in the table pertaining to the VA hospital, that surgery was generally performed sooner than in the VA hospital, although in only one instance did an operation occur on the day of admission. And in only two others did it occur on the second day.

The experience of a university hospital in the same city is described in chart 5. In all but four instances these patients were operated the second day after admission. In these four cases, operation occurred on the third post-admission day. The general conclusions to be drawn from the analysis of the herniotomy study have been demonstrated in almost all of VA's special analyses. Another illustration is a comparison of the same hospitals as above with respect to patients CHART 5

admitted for tonsillectomy (chart 6). The same situation is noticeable, but perhaps in a much more striking way. The private hospital provided information on 12 patients, all of whom were operated the day of admission, and discharged the next day. Six of these patients were ward cases; six were private cases. All of the ward cases returned to the outpatient department for checkup 9 days after discharge. All of the private cases were given post-hospital care by their own physician at home or in his office.

The story was much the same for the municipal hospital. Here, of seven patients, all were operated on the second day after admission, and with one exception were discharged the next day. All returned for checkup to the hospital outpatient department 1 week after discharge.

The question as to why operation in the municipal hospital was not on the day of admission, as was the case with the university hospital, is immediately raised. Here we begin to see the influence of one of the factors that lengthens stay in hospitals in the Veterans' Administration system. These patients were unknown to the municipal hospital prior to admission, and they too had to be worked up to a greater extent than was the case for patients in the university hospital.

In the top section of chart 6 is shown the experience of the Veterans' Administration hospitals for 20 tonsillectomy patients. The earliest day of operation for any of them was the third day. All of the cases except one were operated on either the third, fourth, fifth, or sixth day of hospitalization. Special diagnostic procedures and additional diagnoses account, in a number of the cases, for some of this delay, as indicated in the "remarks" section of the table.

Veterans' Administration has explored another aspect of length of stay-one which may well be a significant approach to the problem, and one which it is possible for use in the central office through the medium of records received from VA hospitals. The VA medical records system is roughly the same as that for the other Federal medical services. At the time that a patient is discharged from hospital a copy of the abstract of the clinical record is received. From this the length of stay can be related to the established diagnosis or to