Printer Friendly

Use and cost of skilled nursing facility services under Medicare, 1987.

Use and cost of skilled nursing facility services under Medicare, 1987


Skilled nursing facilities (SNF's) certified under Medicare provide subacute skilled nursing or rehabilitative services for beneficiaries who are not sick enough to need the acute care services provided by hospitals, but are too sick to be cared for at home. The SNF patient may receive a variety of services that include skilled nursing care; physical, speech, or occupational therapy; drugs; blood; medical supplies such as splints or casts; and the use of durable medical equipment.

The SNF benefit was intended as a less costly alternative to continued hospital stay for post-acute medical or rehabilitation services: "Medicare's SNF benefit was specifically designed to provide only for relatively short-term subacute care needs" (Gornick and Hall, 1988). Thus, eligibility for 100 covered days of care under the SNF benefit was tied directly to a hospital stay of at least 3 days preceding admission to the SNF, and Medicare covered the SNF admission only if the conditions requiring SNF care were the same conditions treated during the prior hospital stay. The program required the attending physician to certify that the enrollee needed skilled nursing care, physical therapy, or speech therapy for these conditions. Transfer to an SNF had to occur within 30 days of the hospital date of discharge (the original 14-day requirement was changed to 30 days).

The SNF benefit of 100 days of care was linked to the benefit period, or spell of illness, which began the first day an enrollee used hospital services and ended when the enrollee had not been an inpatient of a hospital or SNF for 60 consecutive days. At the beginning of each new benefit period. Medicare hospital insurance (HI) coverage was completely renewed, signaling the availability of an additional 100 covered days of SNF care for the enrollee.

From 1967 through 1988, Medicare paid 100 percent of the reasonable costs for the first 20 days of covered SNF care. For the 21st-100th day, Medicare paid all reasonable costs except for the beneficiary coinsurance. If the beneficiary needed care that extended beyond 100 days during the benefit period, the beneficiary was responsible for all of the SNF charges.

An important law with potential to affect the use of the SNF benefit was the Social Security Amendments of 1983 (Public Law 98-21), which established, effective October 1, 1983, the prospective payment system (PPS). Although PPS specifically concerns short-stay hospitals (SSH's), the ramifications of the law were expected to be left by all providers furnishing post-hospital care. PPS gave SSH's an incentive to discharge patients as soon as medically feasible in their recovery period because Medicare SSH payments were predertermined prospectively for the entire stay, rather than based retrospectively on reasonable cost. A likely result of PPS would be a decrease in hospital length of stay and a corresponding increase in the transfer of hospital inpatients to SNF's.

As expected, the average covered days of care (CDOC) per discharge for short-day hospital Medicare inpatients decreased following the institution of the prospective payment system. The average CDOC fell from 10.1 days in 1983 to 8.4 days in 1985. However, the average CDOC held steady at 8.4 days in 1986 and then rose slightly, to 8.6 days, in 1987. The average CDOC for SSH's, therefore, decreased at an average annual rate of 5.2 percent from 1981 through 1987.

In comparison, the average CDOC for SNF patients declined from 29.2 days in 1981 to 21.5 days in 1987, an average annual decrease of 9.7 percent. Gornick and Hall (1988) reported that "the decline in mean number of covered SNF days per user reflects both an increase in short covered SNF stays and a decline in relatively long covered SNF stays . . . from 1983 to 1985, SNF stays with 7 or fewer covered days increased more than 56 percent and SNF stays with 31 or more covered days decreased 18 percent."

One factor that may have contributed to thee reduction in long covered SNF stays is the increase in the SNF coinsurance amount. The SNF coinsurance, which takes effect on the 21st day of a covered stay, is based on one-eight of the inpatient hospital deductible. In 1987, this deductible was $520. Thus, the SNF coinsurance was $65 per day in some cases, exceeded the SNF's full charge. There is some perception that this coinsurance liability may have induced patients and their families to seek alternative arrangements for continuing care. Also, because of PPS, there was increased physician attention to alternative arrangements that may have directed some of the patients into home health care.

For a variety of reasons, including increased emphasis on ambulatory surgery and perhaps the focused efforts of peer review organizations, the rates of SSH discharges per 1,000 HI enrollees decreased 4.5 percent from 1981 through 1987. During the same years, the number of SSH facilities declined from 6,067 to 5,850 facilities, decreasing at an average annual rate of 1.2 percent.

On the other hand, the SNF admission rate in 1987 was the same as in 1981, 10 per 1,000 HI enrollees. During this period, the number of SNF providers grew from 5,295 facilities to 7,379 facilities, increasing at an average annual rate of 11.7 percent. Included in the 1987 count of SNF providers were 1,058 swing-bed hospitals. The swing-bed concept was incorporated into the Medicare program by the provisions of the Omnibus Reconciliation Act of 1980 (Public Law 96-499). Under this law, rural hospitals with fewer than 50 beds could see the beds to furnish both acute and post-acute care (that is, SNF level of care). The Omnibus Budget Reconciliation Act of 1987 (Public Law 100-203) extended the swing-bed option to rural hospitals with fewer than 100 beds.

As a percent of SSH discharges, the SNF admissions increased from 2.6 percent in 1981 to 3.5 percent in 1986 before dropping to 3.2 percent in 1987. One reason is higher percentage of hospital discharges were followed by an SNF admission could be the incentives embedded in PPS for hospitals to discharge patients as soon as medically feasible, a practice characterized as "quicker and sicker." However, Gornick and Hall (1988) cite the increase in the number of procedures (previously done solely on an inpatient basis) that are being performed in an outpatient setting (e.G., cataracts). Patients having these procedures would not have been likely to need or use post-acute services. Thus, of the patients hospitalized today, a larger proportion are likely to need post-hospital care in an SNF.

In summary, the implementation of PPS was followed by a general decline in the rate of SSH discharges per 1,000 enrollees. Although the rates for SNF admissions per 1,000 HI enrollees remained fairly constant from 1981 through 1987, SNF admissions as a proportion of SSH discharges rose about 35 percent. The average covered days of care per SNF admission decreased.

Selected data highlights

For Medicare beneficiaries using HI benefits during the period 1981-87, trend data are displayed to show the patterns of both SSH and SNF use before and after the implementation of PPS. Table 1 includes the number of SSH and SNF providers, the number of SSH discharges and SNF admissions, rates of SSH and SNF user per 1,000 HI enrollees, the proportion of SNF admissions to SSH discharges, the average covered days of care, and the average annual rate of change (AARC).

During the period 1981-87, the count of SSH providers dropped 3 percent, declining from 6,067 hospitals in 1981 to 5,850 in 1987. In contrast, the number of SNF providers increased 39 percent, rising from 5,295 facilities in 1981 to 7,379 i 1987. The number of SSH discharges in 1981 (10.7 million) decreased 5 percent by 1987 (10.1 million), although SNF admissions rose from 273,000 in 1981 to 327,000 in 1987, an increase of 20 percent.

The SNF admission rate per 1,000 HI enrollees provides a measure to assess the impact of PPS on SNF use. SNF admissions increased from 10 per 1,000 enrollees in 1981 to 12 in 1985. However, the SNF admission rate declined to 11 per 1,000 in 1986 and returned to 10 per 1,000 in 1987. Therefore, according to this measure, there was little change in the use of SNF care following the implementation of PPS. Because of the changes in the patterns of hospital use resulting from the impact of PPS, SNF admissions as a percent of SSH discharges climbed 35 percent, rising from 2.6 in 1981 to 3.5 percent in 1986.

In noting the AARC from 1981 through 1987:

* The number of SSH providers decreased at an average annual rate of 1.2 percent. In contrast, the number of facilities providing SNF services grew at a rate of 11.7 percent; this included swing-bed hospitals.

* SSH discharges declined at a rate of 1.8 percent, and SNF admissions increased at the rate of 6.2 percent.

* Per 1,000 HI enrollees, SSH discharges decreased at a rate of 4.5 percent and the rate of SNF admissions remained steady.

* The average CDOC for each SSH discharge decreased at a rate of 5.2 percent, while the average CDOC per SNF admission decreased 9.7 percent.

Thus, there was a more notable decrease in the average CDOC per SNF admission. The average CDOC declined from 29.2 days in 1981 to 21.5 days in 1987, representing a drop of 24 percent. This pattern was similar to that reported in the Health Care Financing Review. The authors reported that the decrease in the average CDOC per SNF admission during the period 1981-85 was entirely in PPS States (Guterman et al., 1988). Gornick and Hall noted that this declension "reflects both an increase in short covered SNF stays and a decline in relatively long covered SNF stays."

For the period 1971-87, trend data are presented in Table 2 to identify patterns in the use and cost of SNF services. The data are arrayed by calendar year and include the number of CDOC, the amounts of covered charges, the total amounts of program payments under Medicare Part A and Part B, and the amounts of program payments to SNF's

Covered charges under the SNF benefit rose 417 percent, increasing from $230 million in 1971 to $1.2 billion in 1987, an AARC of 10.8 percent. Program payments to SNF's increased 217 percent, rising from $179 million in 1971 to $544 million in 1987, an AARC of 7.2 percent. The widening divergence in covered charges and program payments from 1971 through 1987 (Figure 1) probably occurred, in part, because beneficiary coinsurance payments progressively represented a larger proportion of total SNF expenditures (program payments plus beneficiary coinsurance).

* From 1971 through 1983, the number of CDOC increased 21 percent, rising from 7.5 million days to 9.0 million days, respectively.

* However, from 1984 through 1987, the number of SNF CDOC decreased 21 percent, declining from 8.9 million days to 7.0 million days.

* Program payments to SNF's as a percent of total Medicare Part A and Part B program payments, fell from 2.4 percent in 1971 to 0.7 percent in 1987.

* Thus, SNF payments were increasing at a much slower AARC (7.2 percent) than total Medicare Part A and Part B payments (15.6) percent) during the period.

* The average SNF Medicare payment per day increased from $24 a day in 1971 to $77 in 1987 (Figure 2), an AARC of 7.5 percent.

* SNF program payment per enrollee increased from $8.62 in 1971 to $17.09 in 1987, an AARC of 4.4 percent.

The use of SNF services during 1987 are examined by area of residence in Table 3. Covered admissions and CDOC are shown along with the corresponding amounts of covered charges and program payments.

Of the U.S. census regions, the rate of covered SNF admissions ranged from 8 per 1,000 enrollees in the Northeast and South Regions to 16 per 1,000 in the West Region. Per 1,000 enrollees, the highest number of CDOC was reported in the West Region (306 days); the lowest, in the South Region (170 days). The average number of SNF CDOC per admission ranged from 19.0 days in the West Region to 29.6 days in the Northeast Region.

Average covered charges per admission were lowest in the North Central Region ($3,305) and highest in the Northeast Region ($4,194). Average program payments were lowest in the South ($1,507) and highest in the West Region ($1,825$. The largest amounts of average covered charges per day ($196) and Medicare program payments per day ($96) were recorded for the West Region.

In the United States, 10 out of every 1,000 Medicare HI enrollees were admitted to SNF's in 1987. Overall, the States in the South Atlantic Division showed the lowest admission rates per 1,000 enrollees. Delaware, the District of Columbia, Georgia, and Maryland admitted Medicare patients to SNF's at the rate of 4 admissions per 1,000 enrollees. Virginia and North Carolina's rate was 5; Florida's rate was 6; South Carolina and West Virginia registered 7 SNF admissions for every 1,000 enrollees.

On the other hand, the States with the highest admission rates per 1,000 enrollees were located in the West North Central Division. These States were well above the national admission rate of 10. Iowa and North Dakota, respectively, reported 23 and 24 admissions per 1,000 enrollees. Kansas (20) and Nebraska (19) were followed in rank by Missouri (16) and Minnesota (13), while South Dakota (11) tied the national rate.

Among the States, Arkansas and South Dakota had the lowest average CDOC (13.3 days) per admission, and Hawaii had the highest (35.4 days). The smallest amount of SNF program payments per day was shown for Delaware ($41) and the largest for Louisiana ($159).

Data on the use of SNF services by Medicare hospital insurance beneficiaries are presented in Table 4. The number of covered admissions, CDOC, amounts of covered charges, and program payments are displayed by the age, sex, and race of the Medicare beneficiary.

With advancing age, the number of SNF covered admissions per 1,000 enrollees increased dramatically. For enrollees aggregated into the various age cohorts under 75 years of age, admissions ranged from 2 to 8 per 1,000 enrollees. Medicare patients 75-79 years of age were admitted at a rate of 12 for each 1,000 enrollees. In the group of enrollees 80-84 years of age, the admission rate per 1,000 climbed to 21; and for the elderly people 85 years of age or over, the rate was 34 per 1,000 enrollees.

* Excluding beneficiaries under 65 years of age, the CDOC rate per 1,000 enrollees rose with advancing age. For beneficiaries 80-84 years of age, the CDOC rate (451) was twice as high as that for all beneficiaries (221).

* For beneficiaries 85 years of age or over, the CDOC rate (753) per 1,000 enrollees was more than 15 times higher than that for beneficiaries 65-66 years of age (50).

* Females averaged 22.2 CDOC per SNF admission, or 10 percent more than males, who averaged 20.1 days.

* Enrollees under 65 years of age had the highest program payments per admission ($1,808) of any age group, about 9 percent above the national average ($1,664).

* Eleven per 1,000 white enrollees were admitted to SNF's in 1987; the rate for enrollees of other races was 8 per 1,000.

* Beneficiaries who were white had shorter average stays per admission (21.3 days) than beneficiaries of other races (24.9 days).

Data in Table 5 reflect the use of SNF services in 1987 by the 12 leading principal admitting diagnoses, that is, those conditions most frequently reported by the attending physician as responsible for the patient's admission to an SNF. The medical coding for the principal diagnosis was taken from the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) (Public Health Service and Health Care Financing Administration, 1980).

* The 12 leading diagnoses for SNF Medicare patients accounted for 141,809 admissions, or nearly 43 percent of all SNF admissions (327,012).

* Two of the 12 leading diagnoses, hip fracture (ICD-9-CM code 820) and acute cerebrovascular disease (ICD-9-CM code 436) accounted for nearly 25 percent of all SNF admissions and charges.

* Of the 12 leading diagnoses, the average CDOC per admission ranged from 15.6 days for malignant neoplasm of the trachea, bronchus, and lung (ICD-9-CM code 162) to 33.9 days for chronic ulcer of skin (ICD-9-CM code 707).

* Covered charges per admission ranged from $2,497 for heart failure (ICD-9-CM code 428) to $5,959 for chronic ulcer of skin.

Days of care cohorts are the focus of Table 6, and the data are arrayed by Medicare status. During the first 20 days of care in each benefit period, Medicare reimbursed the SNF's for all covered services. For stays of 21-100 days, the beneficiary coinsurance amounted to $65 per day in 1987. For SNF stays of 41-100 days, beneficiary coinsurance payments of $65 per day were progressively higher than were the Medicare program payments. That is, beneficiaries paid more than the average per diem payment made by Medicare. For example, Medicare SNF program payments amounted to $46 per day for admissions within the cohort of 81-100 days.

* About two-thirds (215,623) of all SNF admissions had 20 or less CDOC, accounting for 30 percent (2.1 million) of all SNF covered days on care (M90 million) and 42 percent ($227.8 million) of all SNF program payments.

* Approximately 30 percent (69,984) of all admissions entailed 1-8 SNF CDOC.

* The average program payment per SNF admission increased substantially with each successive CDOC interval, going from $528 for 1-8 CDOC to $4,362 for 81-100 CDOC.

* Conversely, as the number of SNF CDOC increased, the average SNF program payment per day decreased, for the most part, declining from $106 to $46.

* For SNF stays over CDOC, the average SNF program payment per day was progressively less than the beneficiary coinsurance payment per day of $65 (one-eighth of the HI inpatient hospital deductible).

* Disabled beneficiaries accounted for only about 4 percent (11,471) of all SNF covered admissions during 1987. The use and cost of services per admission for disabled beneficiaries was somewhat higher than that for aged beneficiaries.

Definition of terms

Admission--The formal admission of a patient into an SNF participating in the Medicare program. Admissions include those who died during an SNF stay or were transferred to another SNF. The admissions shown in this article reflect beneficiaries who received at least 1 covered day of SNF care under the Medicare HI program.

Covered charges--The SNF's covered charge for room, board, and ancillary services as recorded on the billing form (HCFA-1450 or HCFA-1453).

Covered day of care--A day of SNF care during which the services (determined to be medically necessary) covered by Medicare were furnished to a person eligible for HI benefits. The day of discharge is not counted as a day of care.

Principal diagnosis--The principal diagnosis is the condition established after study to be chiefly responsible for the admission of the patient to a SNF. All diagnostic information in this article are classified according to the ICD-9-CM. Three, four, or five-digit codes are assigned for each principal diagnosis.

Program payments--Payments under the HI program are based on interim reimbursement rates reported on processed bills. The interim rates are established to reflect current costs as closely as possible. These are usually established a per diem amount or as a percentage of the total charges. Figures shown exclude amounts for which the patient is responsible such as deductibles, coinsurance, and charges for noncovered services. The final amount of program payments due under Medicare to each provider of medical services is determined after the end of the fiscal year on the basis of the provider's audited reasonable cost of operations.

Skilled nursing facility--An institution providing inpatient skilled nursing and restorative care services and meeting specific regulatory certification requirements. The SNF must be certified under Medicare in order to be reimbursed.

State--Refers to the State where the beneficiary is living, not the State where he or she receives services.

Sources and limitations of data

Data are derived from a 100-percent count of billing forms submitted by participating SNF's for reimbursable inpatient SNF services. Data are based on records processed and recorded as of December 1988.

It is estimated that the totals for the covered days of care and reimbursements are approximately 95 percent of the eventual totals. Thus, the rates may be less than would prevail if completed data were available. Comparisons of rates should be taken as indicative of differences by relevant characteristics rather than as final measures of the rates.

The data for SNF covered days of care should be used cautiously. The decline in the average covered days of care does not necessarily indicate a decline in the patient's actual length of stay.


The authors greatly appreciated the technical expertise and advice of their colleagues David Gibson and James Hatten of the Bureau of Data Management and Strategy and Marian Gornick, Margaret Hall, Charles Helbing, Elizabeth Cornelius, and Herbert Silverman of the Office of Research and Demonstrations. The authors would also like to thank Will Kirby, Thaddeus Holmes, and Beverly Ramsey for providing the data files, graph services, and secretarial services, respectively.


Guterman, S., Eggers, P., Riley, G., Greene, T., and Terrell, S.: The First 3 Years of Medicare Prospective Payment: An Overview. Health Care Financing Review. Vol. 9, No. 3. HCFA Pub. No. 03263. Office of Research and Demonstrations, Health Care Financing Administration. Washington. U.S. Government Printing Office, Spring 1988.

Gornick, M., and Hall, M.: Trends in Medicare use of post-hospital care. Health Care Financing Review. 1988 Annual Supplement. HCFA Pub. No. 03275. Office of Research and Demonstrations, Health Care Financing Administration. Washington. U.S. Government Printing Office, Dec. 1988.

Public Health Service and Health Care Financing Administration: International Classification of Diseases, 9th Revision, Clinical Modification. 2nd ed. DHHS Pub. No. (PHS) 80-1260. Public Health Service. Washington. U.S. Government Printing Office, Sept. 1980.

Reprint requests: Viola B. Latta, 2502 /ak Meadows Building 6325 Security Boulevard, Baltimore, Maryland 21207.
COPYRIGHT 1989 U.S. Department of Health and Human Services
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Latta, Viola B.; Keene, Roger E.
Publication:Health Care Financing Review
Date:Sep 22, 1989
Previous Article:State policies and the financing of acquired immunodeficiency syndrome care.
Next Article:Physician charges and utilization trends.

Related Articles
Changes in Medicare skilled nursing facility benefit admissions.
Medicare-covered skilled nursing facility services, 1967-88.
Skilled nursing facilities.
Changes in Medicaid nursing home beds and residents.
Creating a MEDPAR analog to the RUG-III classification system.
PPS isn't working.
Deadline pressure: Lawmakers zero in on Medicare bills.
Skilled Nursing Providers Call for Changes to Chairman's Mark.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters