CHAPTER VIII
Medical Records and Statistics After World War I
Maj. (later Brig. Gen.) Love1 was succeeded in the Medical Statistics Section in 1932 by Lt. Col. (later Maj. Gen.) George F. Lull, MC, who is at the time of this writing (1957) secretary and general manager of the American Medical Association. In 1936, Maj. (later Col.) Charles G. Souder, who served the section for 3 years, succeeded Colonel Lull. Colonel Souder was followed by Col. John W. Meehan, who continued in this position until his retirement in 1943. For a short period in 1943, Col. Charles G. Hutter served as director of the division and was succeeded by Capt. (later Lt. Col.) Harold F. Dorn, a medical statistician formerly with the U.S. Public Health Service, who had been commissioned specifically for this assignment. Captain Dorn served as director of the division until his relief from active duty in 1946, when he returned to the Public Health Service. At that time, the civilian position which had been established as director of the division was assumed by Eugene L. Hamilton, who, as an officer, had been on duty in the division in 1942-43 and in 1945-46, serving as deputy director in the later tour of duty. This appointment implemented a considered policy decision that the Army medical statistics program should be directed by a permanently assigned civilian medical statistician.
1Love's duty was not continuous from 1917 to 1932. The ate Col. William B. Borden was in charge of the Vital Statistics Division (formerly the Medical Records Section) from June 1922 to October 1923 and from January 1927 to January 1931. During the last 6 months of 1928 and during the years 1929 and 1930, Love devoted his entire time to the war casualty study and to the biometric studies on U.S. Army officers.
112
This period, 1932-57, has been one of much activity and frequent change in the history of medical statistics in the Army. In the course of this time, the Army medical statistics system further evolved, developed, and accommodated itself to contemporary needs and resources. The variability of these needs and resources can be visualized only when it is recalled that this period began in the depths of a worldwide economic depression (when a major Army responsibility was monitoring the activities of a large portion of the Nation's youth in the CCC (Civilian Conservation Corps)), continued through beginning conscription and partial mobilization in the face of war's outbreak in Europe, went on into the United States' entry into World War II, was followed by a brief respite which ended with the beginning of the Korean War, and terminated in the historical present, a 1957 with an absence of a major armed conflict but with a persistent cold war demanding constant readiness.
Although many changes have come about in the detailed procedures followed, in the forms used, and in the techniques applied in order to obtain statistics adapted to the specific needs of the time, it is a substantial tribute to the original architects of the system and to the solidity with which they built that the basic philosophy of reporting and the essential nature of that reporting remains unchanged. The prime parts of the system are, to this day, a brief current summary from each facility of the salient facts concerning the health and hospitalization of troops for immediate operating use and individual case records on all patients as a basis for the detailed medical and statistical analysis on which fundamental policies and longer range plans are based. The similarities between the earlier days experiences of the Army and its medical service and those observed today are more striking than the differences, despite the great strides in statistical theory and methodology and the major technological advances, including electronic developments in tabulating, computing, and communications. In fact, the mission of the Medical Statistics Division or the objective of the Army medical statistics program, when stated in broad and general terms, is found equally applicable in 1932, in 1957, or any of the years between. It is only in viewing the interpretations of this general mission, relative to specific situations and sets of circumstances, and in assaying the techniques of implementing
113
the program that differences worthy of note become apparent.
Since the establishment of what is now the Medical Statistics Division of the Surgeon General's Office, its purpose has been to provide the required statistical support for the Army's medical operations; to plan, devise, and operate a system of medical records and reports which will produce, at the right time and in the right amount, the quantitative facts needed to facilitate effective, continuous, current supervision of the worldwide medical service activities; and to provide a basis for realistic medium- and long-range plans for medical selection and separation of personnel, control of disease hazards and other environmental factors, provision of high-quality medical care to the sick and wounded, and efficient management of the hospitals and other medical resources of the Army.
In this section, the trends, developments, and changes are traced in the separate major parts of the medical statistics field as carried out by the Medical Statistics Division. This includes summary reporting on health and hospitalization; the individual medical records system; the Selective Service registrant examination records system; tabulating techniques, equipment, and procedures; the statistical reports, analyses, and publications; the standardization efforts relating to medical forms, reports, and statistics; and other procedures.
Summary Reporting on Army Health and Hospitalization
Army Medical Service personnel carry important responsibilities for the maintenance of the health of troops and for provision of the best possible care and treatment for the sick and wounded. This applies not only to The Surgeon General and his immediate office staff but also to the surgeon and staff of headquarters at every intermediate and subordinate echelon and to the commander and personnel of every medical treatment facility. Every day, important policy and operating decisions are made which must be based upon factual and up-to-date information. The mechanism by which much of this essential information is gathered and distributed is the summary reports system.
114
The past 25 years have seen many changes in the report forms and in the details of the procedures employed, but the general nature and purpose of this reporting has changed remarkably little. Current numerical summary data have always and still do provide such facts as the incidence and prevalence of, and the noneffectiveness and mortality from, disease, injuries, and wounds; the occurrence and spread of the acute communicable diseases, as the major part of the raw material of epidemiology; and the basic facts on which depend decisions to expand, reduce, or adjust the bed capacities within the hospital system. These reports are submitted by each functioning treatment facility and show the average strength or size of the population to which care is provided. They include patient counts, that is, the number of patients admitted, disposed of, and remaining on hand, by type of personnel, major medical type of case, and place of treatment (hospital or quarters); and they set forth pertinent items regarding bed capacity and beds occupied.
In a general sense, the use of these data continues to be the same throughout the years. The policymakers and operations people depend upon them as a means of having a realistic orientation to, and perspective on, the existing and developing situation and of being aware of the trends that are becoming manifest and the problems that come into the foreground in medical-service affairs. The preventive medicine officers and epidemiologists are, and have always been, in point of volume of instances of use and in point of continuity and consistency of use, the biggest customers for these summary statistics. In the Office of the Surgeon General and in many intermediate headquarters, these data are regularly summarized and distributed to all of those operating officials who require them in planning or carrying out their assigned duties.
Since their original use to the present time, the summary report data bear the same general relationship to data derived from the individual medical records, in that both stem from the same contacts between patients and medical service personnel. There are, however, important differences between them. The summary data submitted regularly each week or each month are more timely but less exact. They tend to overstate the incidence of such conditions as the common respiratory diseases but to understate certain conditions in which
115
establishment of the diagnosis is likely to be delayed. Summary data reflect disease deaths with a fair degree of accuracy but nearly always miss a proportion of the sudden deaths due to injuries resulting from aircraft and traffic accidents. These characteristic limitations are recognized as inherent in this type of system and have been sufficiently evaluated for appropriate allowance to be made for them.
As has been stated, the basic idea and the substance of summary reporting have changed relatively little in the past quarter of a century. However, important changes have come about in the mechanics of the system, the design of forms, and the techniques by which the system is caused to serve its purpose efficiently. Perhaps a fair analogy is that, although the essential nature of the physician-patient relationship has changed little since Hippocrates, extremely important advances have been made in the practice of medicine.
The statistical report, which is based on summary data, was first used in Vera Cruz, Mexico, about 1914 or 1915. Before 1932, many changes were made in its form and contents. Originally consisting of a single sheet, it was changed to a three-section report in the 1920's, each section of which was on a separate sheet of paper and was, in all but title, a separate report. The first section, Form 86a (figs. 21, 22, and 23), contained the strength, admission and disposition, bed capacity, and status information, together with a number of minor items. The second section, Form 86b (fig. 24), provided data on the communicable diseases. The third section, Form 86c (fig. 25), presented the statistics of Medical Department personnel and transportation equipment. In December 1933, a single change was effected by which Form 86a and Form 86b were physically combined on one sheet to make Form 86ab (figs. 26, 27, 28, and 29). No further change occurred until the revision of August 1940, although a specific Civilian Conservation Corps statistical report, which was essentially the same in content as the Form B6ab, had been put into use for this specific category of personnel. This specific CCC report was the only source of CCC medical statistics, since the corresponding individual medical records were reviewed and filed but not coded or otherwise processed.
Within the Office of the Surgeon General, the data received were edited and consolidated on an area basis, to provide totals
116
FIGURE 21. -Statistical Report, Form 86a, Medical Department, U.S. Army. (Authorized Feb. 14, 1921.)
117
118
119
FIGURE 22.-Statistical Report, Form 86a, Medical Department, U.S. Army. (Revised Jan. 8, 1926.)
120
121
122
FIGURE 23.-Statistical Report, Form 86a, Medical Department, U.S. Army. (Revised May 23, 1927.)
123
124
125
for each corps area for the United States, for overseas, and for the total Army. Data for the overseas commands were submitted only in the form of a consolidated report for the command. Summary posting cards were maintained for each of the important health indexes and for the major diseases. In addition to the weekly data, regularly summarized, annual totals were also derived and distributed, as had been done for several years. In fact, the first annual summaries were pre?pared and published in the late 1920's,
In 1940, the Selective Service Act became effective, and in August of that year Form 86ab (fig. 27) was revised and expanded, in anticipation of the need for more detailed data to support the Army's broadened operations. For the reporting of both strength and beds occupied, a breakdown by components, that is, Regular Army, National Guard, and Reserve, was introduced. Whereas, on the 1933 form (fig. 26) the patient data by disease, injury, and battle casualty had been reported for hospital and quarters cases combined, the 1940 revised Form 86ab provided for the separate reporting of the hospital cases and of the quarters cases, as well as for the total of the two groups. In the reporting of bed status, a new distinction was introduced between beds in permanent buildings and those in temporary buildings,
The 1940 revised Form 86ab continued in use for about 3 years, during which time the situation had changed completely. The problems changed from those of a greatly expanded training mission to those of fighting and supporting a full-scale, worldwide war. Needless to say, this involved important changes in the needs for data. The new version of Form 86ab (fig. 28), titled ''Statistical Health Report,'' was published in October 1943. The changes which it incorporated provided a reflection of the type of information which nearly 2 years of war experience had shown to be necessary. The strength section was changed to provide information on Women's Army Corps personnel and a breakdown of strength by race. The officer-enlisted distinction in strength was dropped, as was the 'by component' information. In the section of the report on patients occupying beds at the end of the period, separate information was called for on patients in convalescent facilities as distinct from hospitals; separate data were required on Army Air Force personnel, Women's Army Corps personnel, Allied
126
FIGURE 24.-Statistical Report, Form 86b, Medical Department, U.S. Army. (Revised Jan. 8, 1926.)
127
128
129
FIGURE 25.-Statistical Report, Form 86c, Medical Department, U.S. Army. (Revised Feb. 4, 1922.)
130
131
132
FIGURE 26.-Statistical Report, Form 86ab, Medical Department, U.S. Army. (Authorized Dec. 22, 1933.)
133
134
135
and neutral forces, and prisoners of war; and all of the bed occupancy data were cross classified by race. A section was introduced to give information on days lost by Army patients, cross classified by the type of case (disease, injury, or battle casualty) and place of treatment (hospital, quarters, or convalescent facilities). A days-lost table for venereal disease by race and type of personnel (Army, excluding Women's Army Corps) was also included. New information was required on bed capacity and bed status, differentiating between fixed and nonfixed hospitals; within fixed hospitals, between buildings and tentage; numbers of beds in locked wards; and status of bed credits in other than Army hospitals.
Soon after the 1943 Form 86ab was published, it became apparent that information was also needed on the personnel losses that were occurring by reason of separations for disability. The Statistical Health Report was therefore again revised in January 1944 (fig. 29) to show the number of patients disposed of by CDD (Certificate of Disability for Discharge). Other minor additions were also effected, such as reporting on the number of Army patients on sick leave and furlough and the number of deaths among Army patients in non-Army hospitals.
In July 1945, the Statistical Health Report was again revised (fig. 30), bringing it within the new pattern of identification of forms, established by The Adjutant General, whereby all medical forms were to be numbered in the 8-series. The substantive changes, apart from adding a brief section on the hospitalization status of patients in the reconditioning program, were relatively minor and few in number and were, generally, in the direction of further simplifying the form, For example, the officer-enlisted men distinction and the race breakdown were dropped from the table on patients occupying beds. The lines in the admission and the dispositions sections of the patients' table labeled 'hospital to or from quarters' were deleted, since, although these lines had been intended as an accounting convenience, they had proved to be a source of confusion. In the 'reportable conditions' section, a specific line was provided for trenchfoot, which had previously been a write-in entry. In addition, a separate Statistical Health Report was required to be prepared and submitted on the enemy prisoner-of-war population, which had grown quite large. This version of the
136
137
FIGURE 27.-Continued (Second Section)
138
FIGURE 28.-Front of Statistical Health Report, W.D., M.D. Form No. 86ab, 18 October 1943
139
FIGURE 28.-Continued. Back of Statistical Health Report
140
FIGURE 29.-Front of Statistical Health Report. W.D., M.D. Form No. 86ab, 24 January 1944
141
FIGURE 29.-Continued. Back of Statistical Health Report
142
FIGURE 30.-Front of Statistical Health Report. W.D. A.G.O. Form 8-122, 1 July 1945
143
FIGURE 30.-Continued. Back of Statistical Health Report
144
FIGURE 31.-Front of Morbidity Report (DD Form 442)
Statistical Health Report continued in use until the middle of 1951.
It was at that time, in 1951, that the new family of summary reports which had for more than a year been under development in the Defense Department Committee on Standardization of Medical Forms, Recording and Reporting Procedures, was placed in use in Army, Navy, and Air Force facilities throughout the world. This represented the most basic and far-reaching change ever effected in summary statistical reporting in the Army. The new summary reports system included three report forms (figs. 31, 32, and 33). The Morbidity Report and
145
FIGURE 31.-Continued. Back of Morbidity Report
the Beds and Patients Report together might be thought of as replacing the Statistical Health Report, while the Outpatient Report was to serve the purpose which had been little more than touched on by the very limited outpatient information provided on the monthly Report Sheet of Sick and Wounded.
The differences between the morbidity information provided by the Statistical Health Report and the Morbidity Report are substantial and significant. Instead of providing, as did the Statistical Health Report, for some 40 diagnostic categories all of which, except for neuropsychiatric conditions and trenchfoot, were acute communicable diseases, the Morbidity Report
146
FIGURE 32-Front of Beds and Patients Report (DD Form 443)
included nearly 100 diagnostic categories covering, singly or in groups, the entire range of diseases and injuries. Previously, no data except those pertaining to beds occupied had been regularly available for Navy and Air Force personnel or for the nonmilitary personnel treated in Army facilities. With the advent of the Morbidity Report, however, a separate sheet was prepared for each major category, giving full admission, disposition, and diagnosis information for each. Whereas the Statistical Health Report had provided for the inclusion of CRO (carded for record only) cases in the basic patient counts with a "remarks" section note on the total of such cases, the Morbidity Report provided for a completely separate accounting of these cases, which permitted their inclusion or exclusion, depending upon the purpose for which the data were required.
By establishing, in the Beds and Patients Report, a separate form to be submitted by hospitals only and pertaining only to hospitalized patients, it was feasible to make provision for more
147
FIGURE 32.-Continued. Back of Beds and Patients Report
key factors directly relevant to operation of a hospital system. The Beds and Patients Report consists of six sections: In the first section, admissions are shown by source and dispositions by type, together with patient census and bed-occupancy data, all cross classified by Army, Navy, Marine, Air Force, and other. In the second section, abbreviated admission and disposition and census and occupancy data are given for the major component of the 'other' category of part I, such as dependents, retired personnel, short-tour active-duty military personnel Veterans' Administration beneficiaries, Allied and neutral military personnel, prisoners of war, and so forth. The third section sets forth data on beds occupied and available on each of the clinical services operated by the reporting hospital, classified by selected specific categories of personnel. The fourth section gives information on hospital cases in various stages of processing by physical evaluation boards for retirement or dis?ability separation. The fifth section indicates the source of
148
FIGURE 33.-Outpatient Report (DD Form 444)
admission of the case, that is, whether by transfer from hospital or by evacuation from overseas for active-duty military personnel, separately for each military department. The sixth section presents much more detailed information concerning bed capacity, bed status, and bed utilization than had ever previously been available.
The Outpatient Report was designed to meet a specific need. Outpatient treatment had always been a hotly debated question. Many doubted the wisdom of treating military patients in quarters, since there could not be the same control over these patients as could be achieved in hospitals. However, in the continuous effort to make the most efficient and economical use of the available medical resources, emphasis was placed on reducing the duration of hospital stay and treating without hospitalization, that is, as quarters patients or as outpatients, all those cases in which the procedure was medically sound. In the course of implementing this policy, outpatient care assumed a growing importance. To plan and operate this important segment of the medical operation effectively, more and better
149
basic information was required. This requirement was met by the Outpatient Report.
On the Outpatient Report, data were recorded regarding outpatients' visits, treatments physical examinations, and im?munizations, for 14 specific categories of personnel, with 'by Department' detail given where indicated, to facilitate tri-service interchange of information. Previously, the only breakdown available had been Army or non-Army. The new report subdivided outpatient treatments into 13 categories, on a clinical service-type axis, providing data on such groupings as dermatology, ophthalmology, obstetrics, pediatrics, physiotherapy, and surgical-dressing room. Immunizations were itemized by type, together with a count of the number of reactions to each type, classified as mild, moderate, or severe. Outpatient visits were accounted for 'by type of personnel,' with separate identification of those which were visits of outpatients to the facility and those which represented visits of treatment-facility personnel to the patient's home or quarters.
After about 3 years of experience in the use of the Outpatient Report, it was decided that there were a number of ways in which it could be improved to serve its purpose more adequately. This task was undertaken by the Defense Department Committee on Standardization of Medical Forms, and in December 1955 a revised Outpatient Report (DD Form 444) (fig. 33) was published. This revision included several major changes, including the abandonment of the concept of an outpatient 'treatment' as the basic unit of measure and replacing it with the term outpatient 'visit.' Although conceptually and theoretically sound, outpatient treatments had proved to be difficult to define without ambiguity, and many problems had developed in securing standard and uniform reporting on this basis. The new term, although slightly grosser in concept, had the advantage of being susceptible of clear, if arbitrary, definition,
In addition, another change was the reduction and attendant simplification in the number of personnel categories separately reported. The clinical service-type breakdown was replaced by a simple list of major diagnostic categories. In order that approximations of incidence might be developed, a column was provided on the form for showing the number of initial visits
150
for each of these diagnostic groups, which was included in the total. Data on immunizations by specific type were eliminated, as was the detailed information on reactions, since it was determined that total immunizations by major types of personnel would be adequate for minimum operational needs. A brief supplementary table was included wherein outpatient visits of military dependents were cross classified by department and by age (under 14 years, 14 years and older). A new table was also added to permit better workload evaluation by showing the number of adjunct services performed in X-ray, laboratory, pharmacy, and physiotherapy. For Army use, a supplement to the basic form was required, on which those facilities having separate specialized clinics organized would show the number of clinic visits for each, with an 'outpatient,'' 'quarters patient,' and 'inpatient' breakdown of this workload.
The Morbidity Report and the Outpatient Report were required to be submitted by each functioning medical treatment facility, while the Beds and Patients Report was required only of hospitals. The Beds and Patients Report was originally submitted on a weekly basis for the 7-day period ending Wednesday midnight, while the full Morbidity Report and the Outpatient Report were submitted for the 4- or 5-week report month ending on the last Wednesday of the calendar month. Certain key items of the Morbidity Report data were, however, required to be submitted weekly by electrical means as a weekly morbidity telegraphic report. In February 1954, a change to regulations was issued providing that the complete Beds and Patients Report be submitted on a monthly basis (4- or 5-week report month) but that a few selected key items on bed occu?pancy be submitted each week as a part of the morbidity telegraphic report.
In thus reviewing the experience of a quarter of a century with statistical reporting on health and hospitalization in the Army, it can be seen that, although the general purpose and nature of such reporting have remained essentially unchanged, the key to proper and realistic evaluation of the system lies in consideration of the detailed provisions in each period of time and in each set of circumstances. This reporting must be dynamic and adaptive. Its efficacy depends in large part upon its responsiveness to current problems and developing trends in military and medical affairs.
151
Individual Medical Record System
The individual medical record is the prime document in the medical statistics system of the Army. For most of the 25-year period under discussion, the individual medical record was a special Medical Report Card, completed at the time of disposition of the patient, which served not only as a source of essential medical statistics but also as an important legal document for the adjudication of claims and other official actions.
In December 1955, as a part of the continuing program to reduce the workload in the field installations and to effect maximum processing economy, a procedure was instituted whereby a carbon copy of the Clinical Record Cover Sheet (fig 34), prepared for each case and containing a full diagnostic and administrative r?sum? of the case, was transmitted to the Office of the Surgeon General to serve as the source document for statistical tabulations and analyses. The original copy of the Clinical Record Cover Sheet was used as the legal document in the case and was filed as a part of the clinical record in the treatment facility. Thus, the need for a separate medical report card was eliminated. In forward areas in active combat, however, the Emergency Medical Tag (fig. 35) and, where indicated, the Field Medical Card continued to be prescribed throughout this period. Where these latter forms are used, the original copy of the Emergency Medical Tag is submitted to the Office of the Surgeon General to serve both the statistical and the legal needs for the case.
In content, the Emergency Medical Tag and the Field Medical Card are essentially similar to the Medical Report Card and the Clinical Record Cover Sheet. The principal difference lies in physical format whereby the field forms are designed to fit their intended field use. The Emergency Medical Tag is a small form on linen or tough paper stock, with a wire or string inserted to be used for attaching it to a patient's clothing. It is made up in books, with carbon paper interleaves to provide for retention of a copy of the record by the medical officer or first medical aidman who attended the casualty and prepared it. The Field Medical Card is a multiple-section folding form which permits successive entries by treatment facilities through which a casualty is evacuated, with only a onetime
152
FIGURE 34.-Clinical Record Cover Sheet (DD Form 481)
153
FIGURE 35.-Front of Emergency Medical Tag, W.D., A.G.O. Form No. 8-26, October 1944
entry required of each of the identifying and administrative items.
The content of the individual medical record has remained substantially the same for the past 25 years, despite the variations in the size and shape of the sheet or card on which it is recorded. The items have always included the name, grade, and service number of the individual; his age, race, and length
154
FIGURE 35.-Continued. Back of Emergency Medical Tag
of service; the date of admission and place or source of admission; the cause of admission and additional diagnoses, including, in the case of injury, an indication of how, when, and where, and by what mechanism the injury was incurred; the treatment given, including surgical operations; and indication of whether or not the disease or injury was incurred in line of duty; the nature of the disposition of the case, whether by return
155
to duty, death, disability separation, or other action; the date of disposition and recapitulation of the days lost from duty by place of treatment; and appropriate identification of the reporting facility. These were the basic items required and are, therefore, available on this record throughout this period. Certain other items have been required during parts of this time, but not continuously. For example, for a time the State or country of nativity was required to be entered. Over the years, with the increasing mobility of the population, nativity came to be considered as of doubtful significance as a factor in most medical analysis and was therefore deleted. Organization and arm or service of the individual has usually been a required item, However, for a time, this was not called for on the Emergency Medical Tag, since it was believed that such an entry should not be made in forward combat areas because of its potential value to the enemy in 'order of battle' intelligence.
With the adoption of the Clinical Record Cover Sheet in 1955, the changes effected were slight. In view of a change in personnel procedures in the Army, the military occupational specialty for enlisted men was substituted for the entry of arm or service. A new entry was required regarding any change in the physical profile of the individual during the course of treatment or hospitalization. In the case of deaths, a full transcription of the medical certification of cause of death, as it appears on the official death certificate, is now required. The information regarding days lost during the episode of treatment is now required to be reported in somewhat greater detail, with days in a bed-occupancy status in hospital being specifically identified.
The individual medical record is completed at the time of disposition of the patient, by the reporting medical treatment facility. Once each report month (4- or 5-week period), a transmittal of these records is made to The Surgeon General, Up to the middle of 1951, this was known as the monthly Report of Sick and Wounded and consisted of the individual records and a Report Sheet of Sick and Wounded (MD Form 51 until 1944, thereafter WD AGO Form 8-23) (fig. 36) which identified the reporting facility, specified the number of records submitted, provided information on the composition and strengths of the troop units served, noted any important moves or changes in status during the period, and gave a limited number of items
156
FIGURE 36.-Front of Report Sheet of Sick and Wounded, W.D., AGO. Form No. 8-23, 1 August 1944
of data regarding outpatient service, physical examinations, and immunizations. When, in 1951, the new summary reports triad became effective, it was decided that the Morbidity Report strength data would be adequate for individual medical records purposes. Further, since the outpatient information
157
FIGURE 36-Continued. Back of Report Sheet of Sick and Wounded
was not only replaced but also greatly improved upon by the new Outpatient Report, it was decided that the Report Sheet of Sick and Wounded could be discontinued. A simple transmittal letter stating the number of records which were enclosed was required in its place.
158
Before 1943, the reporting facility included in its Sick and Wounded Report the records of all patients disposed of during the month, including those disposed of by transfer to other hospitals. This made it necessary for the Medical Statistics Division to maintain, alphabetize, and continuously process a tremendous file of incomplete records for the purpose of matching all of the cards for a given case and deriving a complete record for coding. In 1943, this practice was discontinued; a procedure was instituted whereby the individual record on a transferred case would be forwarded to the receiving hospital with the patient. At the time of final disposition of a patient (that is, disposition by return to duty separation from the Army, or death), the complete record of the case from admission to final disposition, including the records from all intermediate hospitals, was sent to The Surgeon General.
It should be noted that the Sick and Wounded Report Card (p. 58) formerly served as an information card as well as a medical report card. It was the only source of information available in the War Department from which congressional and other inquiries about the individual soldier could be answered. Love recalls a distinguished Federal judge who came to the office twice to pore over a bloody diagnosis tag, the only intimate record available of the death of his only son in battle in the AEF. When personnel reports were inaugurated by the Adjutant General's Office, they replaced the use of Sick and Wounded Report Cards for such information purposes.
Late in 1944, the entire basic regulation on the individual medical records system, with greatly amplified, explicit instructions for this important function, was revised and republished. Also, in 1944, a new four-digit classification of diseases and injuries was adopted for use in the coding of the individual medical records. This code was developed with the advice and assistance of many medical and surgical specialists who were on the staff or were serving as consultants to The Surgeon General.
Processing.-The tabulations based on individual medical records are ordinarily prepared on a once-a-calendar-year basis. Since the records are received after final disposition of the patient and since some patients are hospitalized for extended periods of time, it is apparent that many records would not be received, in the ordinary course of events, until long after the
159
end of the year in which the admission occurred. Since it is necessary for each year's tabulations to account for all of the admissions and all of the disease incidence which occurred in that year, some special mechanism was required to make possible the inclusion of these cases. The system used for this purpose required that a 'remaining record' be submitted with the February report for every patient who was admitted in the preceding calendar year, or earlier, who had not been disposed of by the end of February. This record is similar to the final record of a case, except that it cannot provide final-disposition and duration information. These remaining records, together with information from January and February final-disposition records on cases admitted in the preceding years, made possible the timely completion of tabulations, with adequate coverage of admission, incidence, and noneffectiveness.
Inherent in the nature of the individual medical records system is the fact that complete and meaningful tabulations of the data from these records are possible only on a centralized basis. In any command below Department of the Army level, the records available for tabulation would be incomplete to whatever extent patients originating in that command had been transferred elsewhere, and would be further distorted by the lack of remaining-record type information. For this reason, the system of transmission of individual medical records was designed, with the idea in mind of facilitating and expediting the receipt of these records in the Office of the Surgeon General. Within the United States, for example, the report goes directly from the reporting treatment facility to The Surgeon General and not through any intermediate headquarters. In the oversea areas, the reports are channeled through the office of the surgeon of the command, primarily because of the assistance he may be able to render in maintaining effective control of coverage and promptness in submission.
Since the Office of the Surgeon General does not maintain any permanent individual medical records files, it was necessary to provide for prompt and properly controlled forwarding or disposition of these records after editing and coding. Until 1955, when use of the carbon copy of the Clinical Record Sheet, was begun, all individual medical records were transferred to The Adjutant General after processing, except for those pertaining
160
to foreign nationals and to personnel of the Department of the Navy or the Department of the Air Force. The records of foreign nationals were ordinarily transmitted to the embassy of the government concerned and those pertaining to Navy personnel or to Air Force personnel were transmitted to those Departments. Since 1955, the originals of DD Forms 481 are sent to The Adjutant General, as part of the clinical record files, and it has been feasible for the Medical Statistics Division to destroy the carbon copies for Army personnel after coding. The Emergency Medical Tags and Field Medical Cards are still sent to The Adjutant General after coding, since they are original documents, and the Form 481 copies for foreign nationals and for other departments are forwarded, as before, to those concerned, since they would not otherwise receive these documents.
The general pattern of processing and use of the individual medical records within the Medical Statistics Division is the same in 1957 as it was in 1932. A coverage control is maintained to insure that a report is received each month from every functioning medical treatment facility. A comparison is made between the number of records listed on the transmittal letter and the number received, The individual records are examined and edited to determine whether they are complete, internally consistent, and prepared in conformance with the pertinent regulations. The records are sorted into appropriate group for assignment to the coding clerks, who translate the relevant personal, demographic, and medical facts about the case from words into numerical form. The coded information is punched into tabulating cards to permit rapid and accurate compilation of the data they contain. The cards are then mechanically sorted and tabulated in the appropriate sequences and groupings to produce the required compilations. The tabulated data are posted to appropriate table formats, and rates, ratios, and other relevant statistical constants and parameters are computed. The data are analyzed and interpreted, and the study or report for which they are compiled is written and distributed.
The major difference which has come about in the last 25 years is in the area of the selection of records for coding. Up to 1943, all of the records which were received were coded according
161
to the predetermined pattern. As long as the size of the Army remained reasonably constant and the size of the Medical Statistics Division staff remained generally proportionate to it, this presented no problem. By 1943, however, the size of the Army had tremendously increased, and the flow of records had increased even more because of the higher illness rates among unseasoned troops and because of the additional records on battle casualties. The size of the coding and tabulating staff was no longer adequately proportionate to the strength, and the full coding of all records could no longer be sustained, These considerations led to the development and adoption of a system of scientific sampling of the records for coding, a system which, in varying forms, has been in effect ever since. Despite this demonstrated necessity for curtailment of the volume of coding, punching, and tabulating, it was recognized that, for certain types of cases, the requirement for complete, detailed, and precise information is so great that sampling is not feasible. For example, deaths, battle casualties, separations from the service, cases invalided home from oversea areas, repatriated American prisoners, and a few selected categories of non-Army personnel were processed on a nonsampled 100-percent basis. In addition, certain other types of cases, representing a combination of relatively low frequency of occurrence with a very high degree of interest and concern, were not able to be satisfactorily sampled.
Sampling.-Where sampling was found feasible, the records were subjected to a variety of methods of sampling to reduce the amount of coding and processing required. From 1943 through 1954, the basic tabulations of medical statistical data for the Army were derived from a randomly selected 20-percent sample of medical records. In preparing the records for coding, the following procedure was adopted: After the deaths, battle casualties, and other cases not subject to sampling had been removed from the file, the residue was sorted into two groups, on the basis of the last digit of the patient's service number-those ending in 3 or 5 constituting the 20-percent sample and those ending in all other digits making up the 80-percent sample. All of the items on the punchcard, in the current tabulating pattern, were coded for cases in the 20-percent sample for every year. In 1943-44 and 1947, 1948, and 1949, it was
162
found possible, in terms of staff and other workload, to code a much abbreviated card, with a sharply limited number of items, for the 80-percent sample cases.
In order to make available for final tabulations a complete 20-percent sample, the punched cards representing the 100-percent sample of special types of records (death, separations, and so forth) were sorted to select those having a service number ending in 3 or 5. The cards so selected were reproduced and combined with the 20-percent sample file of the nonspecial cases. In final processing, the number of cases in the overall 20-percent sample was multiplied by 5 to provide an estimate of the total cases for which records were received.
At the time of its institution, the 20-percent sampling system was adequate for most purposes. As the size of the Army decreased and as the magnitudes of the incidence rates reached lower and lower levels, there came to be an increasing number of instances where the sample of cases available was not entirely satisfactory, particularly for subcategories of diseases of low incidence. It was necessary to devise a method whereby, with the available personnel, it would be possible to obtain a more fully satisfactory final result. The solution adopted was a revised plan of sampling. The new pattern adopted in 1955 involves a 5-, a 50-, and a 100-percent sample, depending upon the category being sampled. It had been found, under the earliest 20-percent sampling arrangement, that, although for many uses the number of cases of medium and low incidence conditions was not sufficient, the number of cases of high frequency conditions, such as acute respiratory disease, was actually much greater than that needed for a sample of adequate stability. Therefore, in the 1955 plan, the group of cases of special interest (that is, the deaths, separations, battle casualties, and patients invalided home) were selected as a 100-percent sample, as before; a limited list of some 22 conditions of very high incidence or very low interest were sampled on a 5-percent basis; and all other cases were sampled on a 50-percent basis.
After the cards were punched, the 100-percent sample cards were reproduced for the establishment of auxiliary files of deaths, separations, and so forth. Then, all of the original cards, whether 100-, 5-, or 50-percent, were included in one morbidity file, The tabulating machines were wired in such a way that each 5-percent sample card would count as 20, each 50-percent
163
sample card would count as 2, and each 100-percent sample card would count as 1. By this means, the tabulations produced were already expanded to the estimated full-case count, and any need for further computation or clerical processing for this purpose was obviated. This sampling method, which is currently in use, had made it possible to achieve a more satisfactory end product, with a cost approximately the same as that for the 20-percent sampling technique.
During this 25-year period, essentially the same kinds of items of information have been coded from the individual medical records for the purpose of analysis. They have always included identification of the patients personal characteristics such as age and race; the diagnoses for which he was treated; the causative agent in cases of injury; surgical operations; the nature of the disposition of the patient; and the number of days lost. The general framework was solidly designed and has remained substantially unchanged. Many significant modifications have been made, however, in the degree of detail and specificity with which these various elements are portrayed and in the numbers and nature of the corollary items which it has been necessary to include in the coding pattern in various portions of this period. As with the summary reports system, it is necessary to tailor the individual medical records and tabulating system to fit the operational and the medical realities of the era and the circumstances in which it is to be used.
Selective Service Registrant Examination Records
During periods of conscription, equitable and efficient use of the Nation's manpower resources in building the required military forces depends to a major extent upon the determination and establishment of realistic and appropriate physical standards for acceptance. Rational evaluation of the physical standards which are established depends largely upon the availability of adequate statistical data reflecting the experience derived in the application of these standards. Proper management of the system of examining and induction stations also relies heavily on a sufficient and continuous flow of current operating statistics.
During the quarter century, 1932-57, there have been three periods in which conscription or selective service was in effect.
164
The first of these extended from 1940 until 1946; the second was for about 3 months, beginning the latter part of 1948 and ending early in 1949; and the third began in the middle of 1950 and has continued until the present time (1958). During these periods, a dual reporting system, paralleling that in use for reporting on Army health and hospitalization, has been placed in use. This system provided for an individual record of the physical examination of each registrant called up for preinduction examination, whether he was subsequently qualified or found acceptable, and of each registrant disqualified by his local board. It also provided for a monthly summary statistical report from each examining or induction station, giving overall information on the number examined, the number found acceptable, and the number disqualified, by major categories of cause for disqualification. Data from the individual records regarding the personal and socioeconomic characteristics of the registrant, his physical measurements, the results of the various tests administered, and any existing diseases or defects provided the basis for the analyses and studies relevant to evaluation of the physical standards and to appraisals and estimates concerning the manpower pool. The summary reports presented a running representation of the levels of rejection, both in total and for the major categories of causes. They offered a basis of comparisons of experience among the various stations and areas and made possible a quick determination of the broad effects of major changes in policy.
The broad pattern in the Medical Statistics Division of processing of the individual examination records has included coverage control, editing for completeness and internal consistency, translating of pertinent personal and medical items into numerical codes, punching these coded items into tabulating cards, preparing tabulations in appropriate sequences and cross classification, and, finally, analyzing, interpreting and publishing the tabulated data. These analyses often suggested changes in standards, to bring resources and requirements into better alinement. From them, it became possible to predict the effects of proposed changes in standards, made it feasible to anticipate health and treatment problems in the Army from data on nondisqualifying diseases and defects among inducted men, and provided a health inventory for important segments of the population.
