Table 1 HOSPITALIZATION UTILIZATION RATES FOR PERSONS AGED 65 AND OVER, a/ Obtained by weighting the rates by age and sex by the estimated OASDI "eligible" population as of the beginning of 1960. Note: The figures shown above for "corrected rates" are the same (except The assumptions in the low-cost estimate produce costs only slightly above the Beneficiary Survey experience. This seems plausible for the near-future. For the long-range future, this low-cost assumption may be said to give recognition to the possibility of success of current efforts for progressive patient care, for reductions in hospitalization costs resulting from development of outpatient-hospital-diagnostic facilities, and for progressive cost-reducing trends in medical practice. Short Hospital utilization data from the National Health Survey, for July 1958 to June 1960 ("Hospital Discharges and Length of Stay: Stay Hospitals, United States, 1958-1960", Health Statistics from the U. S. National Health Survey, Series B No. 32, April 1962, Public Health Service, U. S. Department of Health, Education, and Welfare), have been used to develop utilization rates comparable with those obtained from the Beneficiary Survey data. These data for hospital utilization rates (average days per person per year) are shown in the following table: a/ Based on total hospital utilization with no maximum limitation b/ Obtained by weighting the rates by age (and, where applicable, In the aggregate, the hospital utilization rates derived from the NHS data are very close to those developed from the Beneficiary Survey (used for the purposes of this Actuarial Study). They are somewhat lower than the rates derived on the "low-cost" basis. Furthermore, it should be noted that the NHS data have some upward bias since they include utilization of Federal hospitals, which would not be covered under the Bill (about 10% of all hospital days--for persons of all ages--were in Federal hospitals). The hospital utilization rates derived from the Beneficiary Survey, modified as described above to allow for the effect of benefits being available as a right, must be corrected to allow for hospitalization used by persons dying during the survey year, who were not included in the Survey. For both cost estimates, this correction was obtained for each age-sex group by applying to the estimated proportion dying in a year an assumed average number of days of hospitalization for decedents (8 days for the low-cost estimate and 10 days for the high-cost estimate). As indicated by Table 1, the relative size of this correction naturally varies considerably by age and sex. For both cost estimates, the correction amounts to about 24% of the rate derived from the Beneficiary Survey for all ages combined, but it is as little as about 15% for women aged 65-69 and as much as 35% for men aged 75 and over. The absolute amount of the correction for decedents averages .53 days for a cost estimate intermediate between the lowcost and high-cost ones. After the basic work on these cost estimates had been completed, a more extensive study on the general subject of correcting hospital utilization rates derived from surveys so as to allow for decedents became available ("Hospital Utilization in the Last Year of Life, "Health Statistics from the U.S. National Health Survey, Series D No. 3, January 1961, Public Health Service, U.S. Department of Health, Education, and Welfare). This report presented a preliminary study using data for the Middle Atlantic states (New Jersey, New York, and Pennsylvania) for 1957. On the whole, after modifications to obtain comparability, the results of this survey agreed reasonably well with the adjustments made in the cost estimates for the effect of the exclusion of decedents from the Beneficiary Survey. fore, no changes were made. There The NHS report showed that for persons aged 65 and over, the unadjusted utilization rate was 1.67 days per person per year, while the rate adjusted for decedents was 2.33 days. This is a difference of .66 days, or a relative increase of 39%. The absolute correction for decedents of .66 days in the NHS report is somewhat higher than used in these cost estimates (.53 days on the basis of the current age-sex distribution of the eligibles). The correction based on NHS data, however, did not include the effect of a 60day maximum, which of course would have the effect of reducing the absolute correction (in days) and also the unadjusted utilization rate. Furthermore, it was derived from a population that is somewhat older on the average than the present OASDI "entitled" population (which includes those who are not current beneficiaries because of the retirement test), since the latter includes a higher proportion of the total aged population at the ages just beyond 65 than it does at the oldest ages. The percentage increase due to this correction factor was higher in the NHS report than in these cost estimates (39% vs. 24%), both because of the foregoing two elements and because the absolute increase of the "decedent" adjustment (in terms of days) was measured against a lower unadjusted rate, computed solely on the basis of reported experience of persons alive at date of interview (namely, 1.67 days in the NHS report as compared with the 2.21 days in the Beneficiary Survey). Current NHS statistics on hospital utilization by the population alive at date of interview are higher than formerly reported--as a consequence of the improved data-collection procedures now followed. Accordingly, when measured against this higher base, the days used by decedents would raise the estimated days used by all the aged (derived from the experience of survivors) by a significantly lower amount than 39%, especially after further adjustment for a 60-day limit and for age distribution. Therefore, the use of a 24% correction factor for the data used in this Study appears reasonable. As a further point of comparison between the NHS data and the assumptions in these cost estimates, the average number of days of hospitalization per decedent was 9.57 for the former, as against the assumption here of 8 days for the low-cost estimate and 10 days for the high-cost estimate. A growing body of additional data on hospitalization experience of persons aged 65 and over, subdivided by health-insurance ownership and other relevant characteristics, is available from the National Health Survey. In some respects these findings are at variance with those from the Beneficiary Survey, partly because of the later time period and differing population groups represented, and partly because of differences in survey techniques. On balance, the present cost estimates would be little changed if NHS data were substituted for corresponding Beneficiary Survey data. The foregoing discussion has related to the derivation of hospital utilization rates on the basis of a 60-day maximum provision. It is assumed that such rates apply with equal accuracy whether the maximum relates to a calendar year, a benefit year, or a benefit period as defined in the proposal. Proceeding from those basic cost factors, modifications have been made for proposals considered from time to time in the past that have had different maximum-duration periods or that introduced deductible periods (whether expressed in terms of the first "n" days of hospitalization, a flat dollar deductible regardless of length of hospitalization, or a uniform dollar deductible per day for the first "n" days of hospitalization). The relative effect on the cost factors of increasing the maximum duration of benefits from 60 days to various other durations is as follows: days - 9%; 120 days - 10; 180 days 12%; and 360 days 15%. Conversely, if the maximum duration is reduced from 60 days to 21 days, the cost is lowered by 15%. These factors have been derived from consideration of data from the National Health Survey and from private insurance experiences. In considering the effect of a deductible provision on hospitalization-cost factors, it is necessary to have what is termed a hospitalization continuance table applicable to the particular beneficiary group involved. Such a table was derived from data in the National Health Survey (Health Statistics, Series B No. 7) and is shown in Table 2. Average Daily Cost of Hospitalization The second element in hospitalization-benefit cost factors is the average daily cost (including both room and board and other hospital costs). The 1959 Hospitalization Report derived a figure of $21 a day for persons aged 65 and over in 1956 (see pp. 79-80). This figure was used as the basis for the long-range actuarial cost estimates made for that Report, since all the actuarial cost estimates for the OASDI system made at that time used the 1956 general earnings level. The figure, however, was adjusted upward by 14% (to $24) to take into account the fact that, before 1956, hospital charges had been increasing more rapidly than the general wage level and would probably do so for at least a few more years. The basis of the 14% increase was the assumption that over the next 4 or 5 years after 1956, hospital charges might increase at an average rate of about 6% (perhaps 7-8% in the beginning and lessening amounts thereafter) before an assumed leveling-off so as to have the same rate of increase as the general wage level. Thus, during this period, the cost estimates made in 1959 assumed that the "real increase" of hospital costs in relation to the general wage level might begin at 3-4% a year and then decline, so that a cumulative relative increase of 14% would precede the leveling-off at the end of the 4-5 year period. An analytical study was made in 1959 as to the reasonableness of assuming that after this 14% relative increase, there would be a levelingoff as between hospitalization costs and the general wage level. The data seemed to indicate that in the years since World War II, hospital daily costs have been increasing in a linear manner (at a rate of about $1.60 per year), and that wage rates have been increasing geometrically. Accordingly, although in the recent past the difference between these two trends series has been about 3-4% per year, this difference seems to be declining somewhat. In early 1962, the long-range cost estimates for the hospitalization benefits were again re-examined, this time on the basis of the 1961 earnings levels and considering the relative recent trends of hospital costs, taxable wages, and total wages. In brief, the results of this reconsideration were that both hospital-benefit costs and the "savings" to the OASDI system from raising the earnings base were increased--the former rising somewhat more than the latter. The long-range cost estimates of this Study are based on levelarnings assumptions, at the 1961 level. Another--and equally acceptable-way of describing the earnings-assumption basis of the long-range cost |