the Bradley Commission, including consideration of the elements which should appropriately be included in a comprehensive disability and death compensation system.

I. PROBLEM

Under existing law, except for the statutory awards for anatomical loss or loss of use, the "Schedule for Rating Disabilities" is the basis of awards of disability compensation payable for service connected disability, and must be based upon the average impairment of earning capacity. The basic purpose of disability compensation is economic maintenance and the loss of earning capacity as the result of disability should continue to be the primary factor of rating criteria. However, it is recognized that human life has values beyond the economic sphere and other noneconomic factors, such as shortened longevity should be considered in arriving at an equitable basis for disability compensation.

Since the cornerstone of the disability rating schedule is the measurement of loss of earning capacity resulting from disability, any supplemental schedule of awards based upon a non-economic factor, such as shortened longevity, must necessarily be dependent thereon. It is, therefore, essential that we assume that the rating schedule is scientifically and factually valid for the purpose of arriving at any supplemental schedule of awards based on shortened longevity in an effort to provide an equitable basis for disability compensation.

With this brief background, it is clear that our problem concerns a noneconomic factor and involves the promulgation of a formula of compensation for loss of physical vitality (as meaning power of continuance) reflecting an impairment of health and constituting an infringement or impairment of the veteran's Godgiven right to a normal or average life expectancy, caused by disability due to service. We would emphasize that a shortened life expectancy is a personal loss to the individual existing in the present as well as a loss to the survivors.

The assignment of monetary value to years of life involves computation of numerical estimates in dollars and cents to intangible and nebulous factors of personal, ethical, and philosophic value. We are faced with a paradoxical situation wherein the only tangible guideline in estimating the dollars and cents value of this non-economic factor of shortened longevity revolves around the economic factor of earning capacity which is capable of numerical estimates. However, this does not imply that the only basis of compensation for shortened longevity should be the loss of future earnings.1

It is conceivable that there are many service-connected disabilities in varying degrees of severity, which contribute more or less toward shortened longevity. However, this study has been limited to those disabilities which can be determined to have a material effect on longevity. The problem involves a selection of these pertinent disabilities and the minimum prerequisite degree of severity to warrant entitlement to compensation.

The losses in longevity or the lethal power of a disease will vary according to the age at inception of the disability and the progressive attainment of increasing age. The problem arises as to the feasibility of applying varying losses with different ages as against average losses with average ages in establishing measurement criteria.

II. PROCEDURE

Ideally, original source data would be the best foundation upon which life tables should be established. In this respect, VA records covering a closed group contain probably the best source data from an actuarial point of view. This is confirmed by Edward A. Lew, chief actuary of the Metropolitan Life Insurance Co., who is recognized as one of the foremost knowledgable people in the business. Unfortunately, the VA case folders, as they exist, cannot be used to construct tables retroactively in the absence of rosters to establish cohorts. They can be used to provide current source data. A plan to this effect is contained in paragraph 5d of the memo from the CBD to the Deputy Administrator dated June 2, 1960, Subject: Study of Schedule of Supplemental Awards. Usable source data under this plan will be available from 2 to 4 years hence.

Since our own source data was not immediately available, contact with other agencies and research in medical monographs and literature comprised the next approach to the problem. A visit to the National Office of Vital Statistics disclosed

1 In their book, "The Money Value of a Man," which concerns the economic loss due to death, L. I. Dublin and A. J. Lotka state on page 3: "Esthetic value, sentimental value, of which much might be said with reference to man, with these we shall nct, here, be concerned. This is not because the authors are insensible to the deep significance of valuation, of a kind, given to intangible things. Quite on the contrary, realizing the supreme significance of these intangibles in human affairs, we shall expressly refrain from dealing with such spiritual values by methods wholly unadapted for their measurement."

the availability of statistics and tables showing various increases in average life expectancy resulting from exclusion of certain diseases. Although these were interesting statistics, they merely establish the impact of the lethal effect of certain diseases in shortening the average life expectancy of the entire population but do not show the individual lethal power of the disease or its potentiality for causing early death.

In a further effort to unearth all available source data from outside the VA, contact was made with the following: The offices of Directors of the National Heart Institute, National Cancer Institute and National Institute of Neurological Diseases and Blindness; the Director, Follow-up Agency, Division of Medical Sciences of the National Research Council; Statistical Directors of Group Health and Group Hospitalization; and the labor representative of the British Embassy. These contacts confirmed the fact that source data in government agencies outside the VA other than in the field of cancer is very fragmentary and can form no permanently valid basis to compute shortened longevity tables associated with specific disease.

The remaining source of research material lies in medical literature and monographs. The Metropolitan Life Insurance Co. furnished us with several summaries of pertinent medical studies. One general conclusion that can be drawn from this literature is that the progress in medical science has focused primarily on diagnosis and treatment and has not been particularly concerned with the actuarial element associated with disability. A bibliography of pertinent medical literature is attached.

From a general consensus and on the basis of medical judgment, the following diseases can be classified as the leading causes of shortened longevity: Rheumatic heart disease; arteriosclerotic heart disease and myocardial infarction; hypertensive heart disease; heart disease, other; hypertensive vascular disease; kidney disease; diabetes mellitus; malignant neoplasm.

For the present time and until such time as VA source data becomes available, we believe that consideration of compensation for shortened longevity should be limited to those disabilities with a minimum schedular evaluation of 60 percent. Excluded from the list are diseases or disabilities that are already the basis of statutory awards, and diseases on which materially shortened longevity information is so spotty, fragmentary or conflicting at the present time that definite and reliable trends may not be established.

III. MEASUREMENT OF SHORTENED LONGEVITY

(A) Society of Actuaries 1951 impairment study.

This study is based on data furnished by 27 insurance companies representing 70 percent of the ordinary insurance in force in the United States and Canada on December 31, 1950. The total number of insurance policies in the impairment classifications covered by this investigation exceeded 725,000. The investigation covered the experience under ordinary insurance issued from 1935 through 1949, issued at ages 15 through 64. A summary of the findings of this study translated into reduction in life expectancy as compared to total veteran population is shown in exhibit A.

In attempting to draw conclusions from this impairment study, the following limitations of the data must be kept in mind:

(1) The mortality findings reflect the experience in a highly selected sample of persons insured under ordinary policies. This selected sample was drawn largely from among white adult males in the middle income classes and is, therefore, not representative of the general population.

(2) The mortality ratios were computed on the basis of the number of policies terminated by death during the period of exposure to risk. Since frequently several policies are in force on one life, the mortality ratios on the basis of policies terminated by death may be quite different from corresponding ratios computed on the basis of the number of lives involved.

(3) Many of the impairment classifications, especially those relating to conditions found on examination, do not represent specific clinical diagnoses, but rather reports of physical or laboratory findings. Such data and particularly the laboratory findings may not in some cases reflect any disease or its causal pathology yet may in other cases be symptomatic of any one of a number of diseases.

(4) The mortality findings relate only to the physical impairment status of individuals at the time of issue of the insurance. With the passage of time, some individuals become poorer risks; on the other hand, some become better risks and may voluntarily withdraw from the experience, especially if they have

substandard policies. Such changes in the composition of an impairment classification have a bearing on the mortality findings quite apart from the natural history of the impairment.

The mortality ratios determined by the Society of Actuaries, even with the above limitations kept in mind, are of value in confirming longevity data obtained from clinical studies of individual and group investigators.

(B) VA records 1946-54

Attached as exhibit B is a table of reductions in life expectancy prepared by the VA Insurance Department for the Teague committee in January 1960. This table is based upon a study made in 1956 in connection with the Bradley Commission Survey. In the 1956 study, mortality rates on an overall basis without regard to percentage rating among veterans receiving disability compensation were based upon veteran deaths as follows: For World War I veterans, covering the calendar years 1950-54 inclusive, wherein a total of 25,854 deaths were reported; for World War II veterans the period covered 1946-54, inclusive, and included 51,621 reported deaths. To obtain data for analyzing mortality experience by percentage of disability, using a sampling approach, information was obtained from the following folders: 4,671 World War I deaths during calendar years 1950 and 1954 and 7,634 World War II deaths during calendar years 1946, 1950, and 1954. As this table indicates, the percentage evaluations are not tied in with any specific diagnoses and for this reason, there are included disabilities which have no pertinent relationship to shortened longevity. In addition, the percentage evaluations include combined evaluations of more than one disability. From our experience with compensation ratings, it can safely be concluded that the 100-percent ratings represent proportionately a greater number of single disabilities than those ratings of less than 100 percent. This probably accounts for the wide difference in the longevity losses between the 60 -and 100-percent evaluations. For all of these reasons, it is reasonable to conclude that the reductions in longevity shown in this table are relatively conservative and represent minimum losses.

(C) Medical monographs and other actuarial studies on shortened longevity related to specific diseases

(1) Analysis of series of 540 cases of heart disease studied by Metropolitan Life Insurance Co.-Exhibits C, C-1, and D represent data taken directly from the study of 540 cases of heart disease in white males, cases reported, and/or admitted to disability, Ordinary Department, Metropolitan Life Insurance Co., between 1934 and 1936, and traced to the anniversary of such admission in 1952. All these cases were on waiver and annuity and the great bulk of them had to qualify under the total and permanent disability provision of the insurance contracts. Satisfactory evidence of continued disability was required at intervals and benefits were terminated if the insured had resumed some type of work or was found able to do so. Cases which terminated fatally soon after a heart attack would, obviously, not be represented in this experience. These cases were considered seriously impaired by reason of objective evidence based upon medical investigation independent of the insured individual's private physician. It has been argued that the 166 cases of coronary occlusion in this series are unfavorably weighted because the data were drawn entirely from survivors of coronary attacks who claimed total and permanent disability. Nevertheless, the gross survival experience of this group is not greatly different than that reported in other American and European studies of survival following established coronary artery disease. Moreover, the results of further insurance and clinical studies suggest radical revision of the commonly accepted belief in poor survival outlook in coronary artery disease.

The attached actuarial exhibit E entitled, "Reduction in Expectation of Life in Certain Categories of Heart Diseases" was prepared by Carl J. Singer of the Department of Insurance of the Veterans' Administration by programing and subjecting to computer analysis the Metropolitan Life Insurance Co. data contained in exhibit D. The reductions in life span shown at various ages for various types of heart disease indicate the type of information that may in the future be obtained from VA records. However, since the number of cases in the study is relatively small, the exact reductions in life span by disease and age are important only in establishing trends and guidelines for comparison with other diseases.

Exhibit C, which sets forth the percentage of survivors at yearly intervals for the main types of heart disease, is of more importance for purposes of the present problem. On the one hand, it shows that hypertensive and/or arteriosclerotic

heart disease complicated by cerebral and renal involvement has the poorest outlook for survival, followed respectively in terms of better prognosis by valvular heart disease, then coronary heart disease, and finally by uncomplicated arteriosclerotic-hypertensive heart disease which has the best long-term survival outlook. Secondly, the survivorship tables offer a means of comparing the relative "lethal potential" of heart disease with other diseases such as malignant neoplasms.

(2) "Cancer in Connecticut 1935-51".-At the time of its publication in 1955 this volume of cancer statistics on 75,000 cases published under the above title was one of the most comprehensive reports published on the subject up to that time. It includes survivorship rates (percent) at 1-, 3-, 5-, 8-, 10-, 13-, and 15year intervals. We have drawn extensively on the data in this volume and the relevant material is included in exhibit I attached.

(3) "Survival Experience of Patients with Malignant Neoplasms," published by Public Health Service, HEW.-This is a report, authored in part by the same investigators that worked on "Cancer in Connecticut." It was prepared by the end-results group for the National Cancer Conference, Minneapolis, September 13-15, 1960. It contains either the 1- or 5-year survival experience covering 212,000 cancer patients in the United States. The data on 1-year and 5-year survival in cancer appearing in exhibit I has been drawn from this source if available for the specific entity.

(4) "Analysis of a Series of 453 Cases of Hypertension—17-Year Survival Experience," by Robert Sterling Palmer, M.D.-This data is summarized in exhibit F attached and will be discussed in paragraph IV-B.

(5) Expectation of life and mortality rate at selected ages among diabetics. data is summarized in exhibit G and will be discussed in paragraph IV-B.

IV. MONETARY VALUES FOR SHORTENED LONGEVITY

(A) With no differentiation between diseases

This

As indicated above, we believe the right to compensation for shortened longevity is a noneconomic presently existing intangible right. The assignment of a dollars and cents evaluation to the loss of years of life can only be done arbitrarily. However, as one tangible guideline, average earning capacity offers a possibility in the measurement of the value of a year of life. On the basis of our disability rating schedule, which contains evaluations of disability based on average economic impairment due to disability, and the current rate of disability compensation, total disability representing total economic impairment or loss of total economic capacity is compensated at the rate of $2,700 per year. On the assumption of an equal lethal power among those disabilities selected for compensation for shortened longevity, the attached exhibit H, based upon exhibit B, referred to in paragraph III-B above, indicates a possible theoretical compensation table using $2,700 as the value of one year of life. It should be emphasized that this table does not represent future earnings and for this reason, the process of discounting to determine present worth is not in order and is not shown.

From the standpoint of equity, rates based on age at inception and attained age would appear to be acceptable. However, exact determinations of age at inception would present difficult problems in the adjudication of claims, especially in cases of service-connection established by aggravation of preservice disability, and those cases involving slowly progressive insidious diseases. This adjudicative problem could be eliminated by the use of the crude arithmetical averages shown in exhibit H. This would conform very nicely with the following observations made by the Bradley Commission Report No. VIII, part A, on page 242:

"Despite the fact that no two disabilities are likely to be precisely alike, influencing the future of the disabled persons in exactly the same fashion, a mass compensation program cannot be administered by attempting independent judgment in each case on the particular and special facts which may be involved. social insurance program cannot be turned into a judicial system.

A

"To achieve administration efficacy, social insurance relies on the 'magic of averages' to arrive at overall equity and social justice. This means inescapably that one individual may get somewhat more and another somewhat less than precise individual justice would indicate. If this suggests arbitrariness, one has only to observe the methods used in nonscheduled permanent disability cases,

where administrators are given full discretion to determine each case on its merits."

As already stated, compensation based on averages without differentiation between the selected diseases although equitable only as to those in the middle, might be acceptable if the lethal power were the same for all the killing diseases. Unfortunately, as a study of the tables and statistics in Exhibits attached hereto will show, the lethal power of the selected diseases tends to vary considerably from disease to disease particularly in the field of cancer and even within the same disease depending on the vital organ or part affected. These factors are based upon what appears to be a definite trend and cannot be ignored.

(B) With differentiation between diseases

Exhibit I is a comparison of survivorship rates (percent) at 1-, 5-, 10- and 15year intervals between various categories of heart disease and malignant neoplasms. Heart diseases having different survival curves have been arranged into three categories empirically labeled "category I" "category II" and "category III" in order of increasingly favorable long-term survival rates. Similarly, various localized and systemic cancers have been arranged into groups having similar survival curves, labeled empirically “group I, II, III and IV" in order of increasingly favorable long-term survival rates.

Comparison of survival rates as between heart disease and cancer reveals that cancers listed in groups I and II have an unfavorable mortality experience that is considerably in excess of even the poorest outlook heart category, group I cancers having an average expected lifespan of only a few months to 2 or 3 years. On the other hand, group III cancers have survivorship rates that approximate the experience in heart categories I and II; and group IV cancers have a survival picture that approximates category III of heart disease.

With these relationships established, it is possible to arrange heart disease and cancers into four levels based on the “lethal potential” as follows:

Group I: Consisting of group I cancers.

Group II: Consisting of group II cancers.

Group III: Consisting of group III cancers and category I and II heart disease. Group IV: Consisting of group IV cancers and category III heart disease. With respect to diabetes, exhibit G shows reductions in lifespan due to this disease at attained age 20 of 15.8 years, at age 30 of 12.4 years, at 40 of 9.6 years, at 50 of 7.8 years, and at 60 of 5.9 years. Metropolitan Life Insurance Co. analysis of causes of death among diabetics in the Joslin Clinic, 1950-56, shows that in recent years diseases of the cardiovascular-renal system have accounted for about three-fourths of the total mortality. Heart disease alone is responsible for nearly half of the deaths; two out of every three heart disease deaths were due to coronary artery disease. The life table for diabetics shown in exhibit G is recognized as understating recent improvement in the diabetic's outlook. basis of evidence now available and the relationship to "cardiovascular-renal disease" there is a sound basis for considering diabetes as similar to heart category III survival experience and equivalent to group IV cancers.

On

Hypertensive heart disease survival rates have already been discussed. Hypertensive vascular disease is associated with increased mortality, as well, because of its close association with heart, cerebral and renal vascular complications. Exhibit F, attached, which has been extracted from the survival experience on 453 cases demonstrates survival experience in cases called merely hypertension. The grade IV group in exhibit F, i.e., hypertension without recognizable organic or functional change, has the highest survival rate. Grade III in exhibit F, which constitutes the largest numerical group in the study, consisting of hypertension with organic changes in head, heart, or kidney with no functional failure in any of these areas has a survival rate very similar to hypertensive heart disease in exhibit I and consequently, an equivalent of group IV cancers.

The shortened life expectancy from nephritis is likewise due largely to heart and renal complications, similar to cardiovascular-renal disease, and is therefore appropriately classified for shortened life expectancy as similar to hypertension and hypertensive heart disease, i.e., category III heart disease or group IV cancer. The 1951 Society of Actuaries impairment study confirms such a conclusion as well as clinical study of the natural history of nephritis.