Swing-bed services under the Medicare program, 1984-87.

Under Medicare, swing beds are beds that can be used by small rural hospitals to furnish both acute and post-acute care. The swing-bed program was instituted under the provisions of the Omnibus Reconciliation Act of 1980 (Public Law 96-499). Under Medicare, post-acute care in the hospital would be covered as services equivalent to skilled nursing facility level of care. Data show that the program has had a rapid rate of growth. By 1987, swing beds accounted for 9.7 percent of the admissions to skilled nursing facility services, 6.0 percent of the covered days of care, and 6.2 percent of the reimbursements. Over one-half of the swing-bed services are furnished in the North Central States.

Introduction

This article traces the growth in the use of swing-bed services by Medicare beneficiaries from 1984 through 1987. In the context of the Medicare program, swing beds are beds that can be used by small rural hospitals to furnish both acute and post-acute care. To be covered under Medicare, the post-acute services must meet the same level of care requirements applied to the reimbursement of services by skilled nursing facilities (SNFs). States have the option of also covering swing-bed services at the intermediate care level under their Medicaid programs.

The swing-bed concept was incorporated into the Medicare program by the provisions of the Omnibus Reconciliation Act of 1980 (Public Law 96-499). The law authorized the Medicare and Medicaid programs to cover swing-bed services furnished by rural hospitals with fewer than 50 beds. The provisions of the law were based on the experiences gained in demonstration projects that began in rural hospitals in Utah during the early 1970s and later expanded to Iowa, South Dakota, and Texas. The approach proved popular and received public and private sector support. The program takes advantage of the declining acute care occupancy rates and the surplus bed capacity that became increasingly common among rural hospitals during the 1970s. It provided these hospitals a means of obtaining additional revenues without incurring significant additional costs. At the same time, it provided greater access to post-acute nursing care services in rural areas where such services tend to be thinly dispersed.

The regulations governing Medicare coverage of post-acute services furnished in swing-bed hospitals were issued by the Health Care Financing Administration in July 1982. The method of paying for skilled nursing care services furnished by a swing-bed hospital was based on the assumption that these hospitals incur a relatively low incremental cost to provide post-acute care. They use the personnel, equipment, and facilities already in place to serve acute care patients. Additional service requirements to meet the special needs of nursing care patients (e.g., patient activities, discharge planning) would not require a major expansion of staff. Accordingly, the per diem reimbursement rate for the routine care component of post-acute services covered under Medicare in a swing bed was set at a rate equal to the average paid by the Medicaid program to SNFs for skilled nursing care during the prior calendar year in the State where the hospital is located. Ancillary services were to be reimbursed at cost.

The period following the issuance of the swing-bed regulations was marked by intense Federal efforts to contain the rise of hospital costs to the Medicare program. Several measures affecting payments to hospitals were passed during this period. The Tax Equity and Fiscal Responsibility Act (TEFRA) was passed in September 1982; the Social Security Amendments of 1983 instituted the prospective payment system (PPS) for hospital reimbursement; and the Deficit Reduction Act (DEFRA) of 1984 reinstated a new version of the Medicare separate reimbursement limits for hospital-based and freestanding SNF care that had been eliminated under TEFRA.

This rapid pace of change in the bases by which Medicare reimbursed hospitals for acute and post-acute care induced uncertainty among rural hospitals as to whether it was worthwhile electing the swing-bed option. This was reflected in the initial slow rate of applications by eligible hospitals for certification as a swing-bed facility. However, as the incentives provided by PPS at the acute and post-acute interface became clearer, the rate of election increased. This is reflected in Table 1 that shows the rate at which hospitals became certified to furnish swing-bed services.

Table 1

Number of certified swing-bed hospitals: Selected dates, 1983-87

Selected date	Number of hospitals
December 31, 1983	149
December 31, 1984	471
December 31, 1985	771
December 31, 1986	956
July 31, 1987	1,056

SOURCE: University of Colorado, Center for Health Services Research: Data from Health Care Financing Administration Contract, “Evaluation of National Rural Swing-Bed Program.”

By the end of 1983, about 18 months following the issuance of the regulations, only 149 of an estimated 2,236 hospitals eligible to elect the swing-bed option had done so. By mid-1987, the proportion was approaching the halfway point.

The increasing participation of hospitals in the provision of post-acute skilled nursing care services resulted in swing beds gaining an increasing share of the Medicare SNF market. As summarized in Table 2 and detailed in Table 3, admissions to swing-bed hospitals for SNF services increased from 3.0 percent of all Medicare SNF admissions in 1984 to 9.7 percent in 1987. The swing-bed share of Medicare-covered SNF days increased from 1.5 to 6.0 percent during the same period. Reimbursements for swing-bed care increased from 2.0 percent of SNF reimbursements in 1984 to 6.2 percent in 1987.

Table 2

Percent share of skilled nursing facility admissions, covered days of care, covered charges, and reimbursement accounted for by swing-bed hospitals under Medicare: Calendar years 1984-87

Year	Swing-bed hospital
	Admissions	Covered days	Covered charges	Reimbursements
	Percent share
1984	3.0	1.5	1.8	2.0
1985	7.1	3.7	4.6	4.7
1986	8.5	4.7	5.7	5.6
1987	9.7	6.0	6.6	6.2

SOURCE: Health Care Financing Administration, Bureau of Data Management and Strategy: Data from the Medicare Statistical System; data development by the Office of Research and Demonstrations.

Table 3

Distribution of skilled nursing facility (SNF) admissions, days of care, charges, and reimbursements to nursing home and swing-bed hospitals under Medicare by area of residence: Calendar years 1984-87

Year	Admissions		Covered days of care				Covered charges						Reimbursements
	All SNFs	Swing bed	All SNFs		Swing bed		All SNFs			Swing bed			All SNFs			Swing bed
	Total	Total	Total in thousands	Per Admission	Total in thousands	Per admission	Total in millions	Per admission	Per day	Total in millions	Per admission	Per day	Total in millions	Per admission	Per day	Total in millions	Per admission	Per day
1984	332,746	10,084	8,864.4	26.6	133.1	13.2	$ 975.4	$2,931	$110	$17.8	$1,765	$134	$464.8	$1,397	$52	$9.1	$898	$68
1985	360,501	25,493	8,544.4	23.7	312.0	12.2	1,062.7	2,948	124	48.9	1,917	157	494.6	1,372	57	23.4	918	74
1986	347,418	29,426	7,769.8	22.4	365.5	12.4	1,122.7	3,231	145	63.9	2,172	175	501.4	1,443	65	27.9	948	76
1987	327,012	31,732	7,041.1	21.5	425.3	13.4	1,187.8	3,632	169	78.5	2,474	185	544.3	1,664	77	33.8	1,064	79
	Percent
AARG	−0.6	46.5	−6.0	−7.4	47.3	0.5	6.8	7.4	15.4	64.0	11.9	11.4	5.4	6.0	14.0	54.9	5.8	5.1

NOTE: AARG is average annual rate of growth.

SOURCE: Health Care Financing Administration, Bureau of Data Management and Strategy: Data from the Medicare Statistical System; data development by the Office of Research and Demonstrations.

Shaughnessy, Schlenker, and Silverman (1988) reported findings that help to interpret the data in Table 3. They found that swing-bed patients have substantially shorter stays and greater rehabilitation potential than do nursing home patients. Swing-bed patients, in greater proportion than nursing home patients, were found to need intense medical and skilled care for such problems as recovery from surgery, hip fractures within the past 6 weeks, shortness of breath, and the need for intravenous catheters. Nursing homes tend to treat patients with problems more typically seen in institutional long-term care settings; such as, incontinence, impaired cognitive functioning, and dependence in carrying out activities of daily living (e.g., feeding self, dressing). Each type of facility seems particularly suited to care for patients who can be, respectively, characterized as needing intense subacute care or as the traditional long-term care patient. The evaluation concluded, “At the subacute phase, the quality of services furnished by hospitals was found to be better overall than those services furnished by nursing homes. On the other hand, nursing homes provide higher-quality, traditional, long-term care services.”

In addition to providing a partial explanation for the differences in length of stay, case-mix explains some of the differences in covered charges. The evaluation report estimates (based on 1985 data) that the more intense but shorter term care required by swing-bed patients results in costs about 20-percent higher per day than the average nursing home patient. This is reflected in the differences in the covered charges submitted. In 1987, swing-bed covered charges averaged $185 per day compared with $169 for all SNF days. Reimbursement of routine swing-bed services based on the State Medicaid program's average per diem reimbursement to skilled nursing facilities for routine care services during the previous year kept the difference in reimbursement per day to only $2 in 1987 ($79 to $77).

A second report evaluated the impact of Medicare's prospective payment system (PPS) on the swing-bed program (Shaughnessy et al., 1988). This evaluation found that, despite higher per diem costs for post-acute swing-bed services the overall costs for an episode of illness tended to be lower for patients discharged from a swing-bed hospital “… patients discharged from acute care in hospitals with swing-bed programs were more likely to receive swing-bed care than patients discharged from comparison hospitals. Such patients also received less Medicare nursing home (SNF) and home health care. Subsequent acute care use and cost also tended to be lower for patients discharged from acute care in swing-bed hospitals. The overall result was a slightly lower total cost of care (both excluding and including the cost of the initial acute care episode) for patients discharged from acute care in swing-bed hospitals.”

One factor that may explain the narrowing gap from 1984 to 1987 in the Medicare reimbursement per day is the decreasing average length of covered stay in all SNFs, including skilled nursing services furnished by swing-bed hospitals (Table 3). As shown in Table 3, this average decreased from 26.6 days in 1984 to 21.5 days in 1987. This would reflect the decrease in SNFs, since during the period 1984-87, the average length of nursing care stay increased in swing-bed hospitals. The shorter length of stay decreases the proportion of payment to SNFs made by beneficiaries because of the coinsurance kicking in on the 21st day. Thus, Medicare payments averaged over fewer coinsurance days increases the average Medicare payment per covered day.

Another factor narrowing the difference in the average reimbursement per day may be the method of reimbursing for post-acute routine care services by swing-bed hospitals. Ancillary services which include: supplies, operating room use, drugs, laboratory and radiology services, and anesthesia, are reimbursed at cost. The per diem amount that swing-bed hospitals receive for routine care services is based on the State Medicaid program's average per diem reimbursement to skilled nursing facilities for routine care services during the previous year. For the purposes of the ensuing discussion, accommodation charges will be referred to as charges for routine care services. Routine care charges are usually characterized as room and board charges, but embedded in the cost base on which the charges are established are allocations for such overhead costs as general and nursing administrative services, maintenance and repairs, operation of the physical plant, laundry and linen, housekeeping, dietary services, central services and supply, medical records, and social services. The per diem average amounts charged to Medicare from 1985 through 1987 by swing-bed facilities and SNFs for accommodations and ancillary services to skilled nursing care patients are shown in Table 4.

Table 4

Distribution of charges for skilled nursing facility (SNF) accommodation and ancillary services: Calendar years 1985-87

Year	Accommodation charges				Ancillary charges
	Swing beds		SNFs		Swing beds		SNFs
	Total in thousands	Per day	Total in thousands	Per day	Total in thousands	Per day	Total in thousands	Per day
1985	$22,426.3	$72	$710,933.8	$86	$27,288.5	$87	$344,898.3	$44
1986	28,510.6	78	706,319.0	95	36,630.9	100	397,806.7	56
1987	34,046.5	80	696,337.7	104	45,904.7	108	461,650.3	72
	Percent
AARG	23.2	5.4	1.0	10.0	29.7	11.4	15.7	27.9

NOTE: AARG is average annual rate of growth.

SOURCE: Health Care Financing Administration, Bureau of Data Management and Strategy: Data from the Medicare Statistical System; data development by the Office of Research and Demonstrations.

The average per diem routine care charges by swing-bed hospitals increased by about one-half the rate of increase of the SNFs (Table 4). Average per diem charges for ancillary services furnished by SNFs increased at more than double the rate of swing-bed hospitals although the latter was still 50-percent higher in 1987. The latter relationship is not unexpected, given the characteristics of post-acute swing-bed patients described earlier and the greater access to ancillary services generally available in hospitals. In interpreting these figures, the reader should bear in mind that from 1985 through 1987 total covered days of care furnished by SNFs decreased.

Based on the data available for this analysis, it is not possible to apportion reimbursements to routine care or ancillary services. Assuming there is a concomitancy between costs and charges, it is clear that reimbursements per day to SNFs have been rising in closer consonance to the rise in covered charges than has been the case for swing-bed hospitals (Table 3). This suggests that the current method of paying for routine swing-bed services may not be keeping up with the rate of increase in the hospital's costs of providing routine swing-bed services. However, in light of increasing participation in the swing-bed program, it may be supposed that swing-bed hospitals were still recovering the marginal cost of furnishing post-acute routine swing-bed services in 1987. Based on 1984 data, the evaluation report estimated that, on average, swing-bed hospitals incurred an incremental cost per day for routine post-acute care of about $33 to $34. The average routine care revenues received exceeded the costs by $8 to $10 per day. The 1987 data suggest that the difference between marginal routine care costs and revenues may be narrowing. However, given full cost reimbursement for ancillary services, the marginal revenue for otherwise empty beds seems to be attractive for eligible hospitals.

The geographic distribution of the use of and Medicare payments for swing-bed services in 1987 in relation to all SNF services is shown in Table 5. As expected, the number of swing-bed hospitals and the use of swing-bed services were concentrated in the North Central and South census regions which contain large expanses of rural areas. Of the 1,058 hospitals that submitted a bill for swing-bed services, almost one-half (504) were located in the North Central States. Another one-third (359) were located in the South. Only 16 hospitals in the Northeast Region were certified to furnish swing-bed services: 9 in New Hampshire, 4 in Vermont, and 3 in Pennsylvania. Of the 179 hospitals certified in the West to furnish swing-bed services, 131 (73 percent) were in the Mountain States.

Table 5

Distribution of skilled nursing facility (SNF) admissions, days of care, charges, and reimbursements to nursing homes and swing-bed hospitals under Medicare, by area of residence: 1987

Area of residence	Number of hospitals providing swing-bed services	Covered SNF admissions			Covered days of care					Covered charges					Reimbursements
		Total	Swing bed		Total		Swing bed			Total		Swing bed			Total			Swing bed
			Number	Percent of total	Number	Per admission	Number	Percent of total	Per admission	Amount in thousands	Per day	Amount in thousands	Percent of total	Per day	Amount in thousands	Percent of charges	Per day	Amount in thousands	Percent of total	Percent of charges	Per day
All areas	1,058	327,012	31,731	9.7	7,041,052	21.5	425,251	6.0	13.4	$1,187,820.9	$169	78,500.7	6.6	$185	$544,276.5	45.8	$77	$33,751.8	6.2	43.0	$79
United States	1,058	326,257	31,730	9.7	7,027,623	21.5	425,240	6.1	13.4	1,185,971.4	169	78,500.0	6.6	185	543,353.8	45.8	77	33,751.6	6.2	43.0	79
Northeast	16	53,385	455	0.9	1,578,320	29.6	7,080	0.4	15.6	223,888.4	142	1,244.1	0.6	176	95,591.6	42.7	61	424.1	0.4	34.1	60
North Central	504	101,895	18,358	18.0	1,971,967	19.4	230,530	11.7	12.6	336,796.3	171	42,476.2	12.6	184	161,788.3	48.0	82	21,203.2	13.1	49.9	92
South	359	82,098	9,772	11.9	1,787,008	21.8	150,041	8.4	15.4	294,292.5	165	27,661.6	9.4	184	123,758.3	42.1	69	9,088.8	7.3	32.9	61
West	179	88,879	3,146	3.5	1,690,328	19.0	37,597	2.2	12.0	330,994.2	196	7,118.1	2.2	189	162,215.6	49.0	96	3,035.5	1.9	42.6	81
New England	13	9,410	299	3.2	251,022	26.7	4,270	1.7	14.3	36,849.2	147	747.7	2.0	175	16,540.0	44.9	66	283.9	1.7	38.0	66
Maine	0	778	3	0.4	18,375	23.6	37	0.2	12.3	4,200.7	229	7.6	0.2	205	2,267.2	54.0	123	2.2	0.1	28.9	59
New Hampshire	9	683	197	28.8	14,076	20.6	2,402	17.1	12.2	2,484.8	177	428.6	17.2	178	972.6	39.1	69	145.4	14.9	33.9	61
Vermont	4	351	90	25.6	9,484	27.0	1,725	18.2	19.2	1,201.3	127	288.0	24.0	167	498.4	41.5	53	130.4	26.2	45.3	76
Massachusetts	0	2,689	3	0.1	74,886	27.8	46	0.1	15.3	13,968.6	187	5.1	*	111	6,500.2	46.5	87	2.1	*	41.2	46
Rhode Island	0	1,144	1	0.1	34,895	30.5	7	*	7.0	3,621.0	104	1.1	*	157	1,502.0	41.5	43	0.4	*	36.4	57
Connecticut	0	3,765	5	0.1	99,306	26.4	53	0.1	10.6	11,372.8	115	17.4	0.2	328	4,800.0	42.2	48	3.3	0.1	19.0	62
Middle Atlantic	3	43,975	156	0.4	1,327,298	30.2	2,810	0.2	18.0	187,039.2	141	496.4	0.3	177	79,051.6	42.3	60	140.2	0.2	28.2	50
New York	0	21,161	10	*	699,281	33.0	123	*	12.3	97,116.4	139	22.7	*	185	38,840.4	40.0	56	7.6	*	33.5	62
New Jersey	0	3,726	9	0.2	109,883	29.5	67	0.1	7.4	15,745.1	143	13.1	0.1	196	7,102.2	45.1	65	4.4	0.1	33.6	66
Pennsylvania	3	19,088	137	0.7	518,134	27.1	2,620	0.5	19.1	74,177.7	143	460.7	0.6	176	33,109.0	44.6	64	128.2	0.4	27.8	49
East North Central	97	57,374	2,922	5.1	1,248,104	21.8	37,552	3.0	12.9	191,120.7	153	6,962.6	3.6	185	88,071.7	46.1	71	3,638.8	4.1	52.3	97
Ohio	7	11,974	173	1.4	230,475	19.2	1,823	0.8	10.5	34,857.8	151	552.9	1.6	303	15,655.2	44.9	68	116.3	0.7	21.0	64
Indiana	8	9,296	441	4.7	179,755	19.3	5,892	3.3	13.4	23,870.0	133	1,071.4	4.5	182	13,102.4	54.9	73	588.0	4.5	54.9	100
Illinois	22	15,988	879	5.5	325,838	20.4	10,393	3.2	11.8	73,115.2	224	2,162.9	3.0	208	33,571.1	45.9	103	886.6	2.6	41.0	85
Michigan	0	13,841	25	0.2	383,822	27.7	384	0.1	15.4	41,606.6	108	73.4	0.2	191	16,810.1	40.4	44	40.3	0.2	54.9	105
Wisconsin	60	6,275	1,404	22.4	128,214	20.4	19,060	14.9	13.6	17,671.1	138	3,101.9	17.6	163	8,932.9	50.6	70	2,007.5	22.5	64.7	105
West North Central	407	44,521	15,436	34.7	723,863	16.3	192,978	26.7	12.5	145,675.6	201	35,513.6	24.4	184	73,716.6	50.6	102	17,564.4	23.8	49.5	91
Minnesota	54	7,223	1,360	18.8	153,564	21.3	12,904	8.4	9.5	19,599.0	128	2,245.4	11.5	174	8,706.7	44.4	57	1,112.1	12.8	49.5	86
Iowa	91	10,106	4,349	43.0	138,179	13.7	50,800	36.8	11.7	31,556.0	228	9,942.4	31.5	196	18,971.2	60.1	137	6,199.3	32.7	62.4	122
Missouri	47	12,318	2,058	16.7	189,805	15.4	24,547	12.9	11.9	52,955.4	279	5,265.7	9.9	215	27,492.2	51.9	145	3,087.3	11.2	58.6	126
North Dakota	33	2,238	1,166	52.1	45,312	20.2	21,144	46.7	18.1	5,049.8	111	2,721.3	53.9	129	2,226.8	44.1	49	1,161.4	52.2	42.7	55
South Dakota	34	1,137	874	76.9	15,160	13.3	10,867	71.7	12.4	2,400.8	158	1,841.7	76.7	169	875.5	36.5	58	589.6	67.3	32.0	54
Nebraska	63	4,419	1,852	41.9	80,541	18.2	26,880	33.4	14.5	18,306.5	227	4,788.0	26.2	178	7,455.2	40.7	93	2,326.6	31.2	48.6	87
Kansas	85	7,080	3,777	53.3	101,302	14.3	45,836	45.2	12.1	20,308.4	200	8,709.0	42.9	190	7,989.0	39.3	79	3,088.0	38.7	35.5	67
South Atlantic	82	30,093	1,773	5.9	745,447	24.8	34,613	4.6	19.5	102,730.0	138	5,319.2	5.2	154	44,889.9	43.7	60	1,419.0	3.2	26.7	41
Delaware	0	295	0	0.0	8,099	27.5	0	0.0	0.0	787.9	97	0.0	0.0	0	336.0	42.6	41	0.0	0.0	NA	0
Maryland	0	1,753	6	0.3	41,984	23.9	100	0.2	16.7	5,061.0	121	14.0	0.3	140	2,317.3	45.8	55	3.1	0.1	22.1	31
District of Columbia	0	265	0	0.0	6,444	24.3	0	0.0	0.0	891.0	138	0.0	0.0	0	464.1	52.1	72	0.0	0.0	NA	0
Virginia	5	3,469	101	2.9	102,488	29.5	1,482	1.4	14.7	14,228.8	139	317.0	2.2	214	6,177.2	43.4	60	121.5	2.0	38.3	82
West Virginia	10	2,090	375	17.9	51,003	24.4	6,791	13.3	18.1	6,426.2	126	1,263.9	19.7	186	2,486.1	38.7	49	382.6	15.4	30.3	56
North Carolina	21	3,863	647	16.7	102,780	26.6	13,529	13.2	20.9	11,863.4	115	1,792.5	15.1	132	4,588.9	38.7	45	424.6	9.3	23.7	31
South Carolina	13	2,628	325	12.4	71,121	27.1	7,766	10.9	23.9	10,499.5	148	1,102.5	10.5	142	4,695.2	44.7	66	204.9	4.4	18.6	26
Georgia	24	2,607	190	7.3	50,360	19.3	3,173	6.3	16.7	7,133.8	142	417.8	5.9	132	2,994.4	42.0	59	128.6	4.3	30.8	41
Florida	9	13,123	128	1.0	311,168	23.7	1,764	0.6	13.8	45,838.4	61	404.2	0.9	229	20,830.6	45.4	67	152.8	0.7	37.8	87
East South Central	105	22,833	4,471	19.6	552,723	24.2	73,859	13.4	16.5	71,911.8	130	12,959.8	18.0	175	28,398.6	39.5	51	4,504.3	15.9	34.8	61
Kentucky	13	5,126	349	6.8	139,356	27.2	7,077	5.1	20.3	17,231.0	124	1,071.6	6.2	151	6,856.2	39.8	49	451.3	6.6	42.1	64
Tennessee	27	8,883	1,248	14.0	259,526	29.2	16,278	6.3	13.0	34,562.2	133	4,091.4	11.8	251	13,581.9	39.3	52	1,298.6	9.6	31.7	80
Alabama	14	5,925	285	4.8	101,999	17.2	3,413	3.3	12.0	11,700.4	115	534.7	4.6	157	4,794.1	41.0	47	153.3	3.2	28.7	45
Mississippi	51	2,899	2,589	89.3	51,842	17.9	47,091	90.8	18.2	8,418.0	162	7,262.0	86.3	154	3,166.3	37.6	61	2,601.1	82.1	35.8	55
West South Central	172	29,172	3,528	12.1	488,838	16.8	41,569	8.5	11.8	119,650.7	245	9,382.6	7.8	226	50,469.8	42.2	103	3,165.5	6.3	33.7	76
Arkansas	31	2,985	636	21.3	39,715	13.3	7,885	19.9	12.4	9,331.4	235	1,472.1	15.8	187	5,294.7	56.7	133	567.9	10.7	38.6	72
Louisiana	34	7,210	842	11.7	109,272	15.2	10,624	9.7	12.6	42,263.1	387	3,660.8	8.7	345	17,418.8	41.2	159	777.6	4.5	21.2	73
Oklahoma	25	4,322	471	10.9	61,354	14.2	5,881	9.6	12.5	18,715.8	305	1,014.1	5.4	172	8,827.8	47.2	144	354.2	4.0	34.9	60
Texas	82	14,655	1,579	10.8	278,497	19.0	17,179	6.2	10.9	49,340.5	177	3,235.6	6.6	188	18,928.5	38.4	68	1,465.8	7.7	45.3	85
Mountain	131	19,998	2,361	11.8	342,881	17.1	28,651	8.4	12.1	59,105.4	172	5,211.3	8.8	182	29,809.6	50.4	87	2,246.2	7.5	43.1	78
Montana	29	2,581	567	22.0	53,192	20.6	7,331	13.8	12.9	5,829.7	110	1,090.9	18.7	149	2,557.7	43.9	48	364.9	14.3	33.4	50
Idaho	17	1,525	322	21.1	22,748	14.9	3,195	14.0	9.9	2,921.5	128	648.0	22.2	203	1,539.6	52.7	68	333.1	21.6	51.4	104
Wyoming	13	506	297	58.7	8,177	16.2	4,488	54.9	15.1	1,236.2	151	698.3	56.5	156	500.0	40.4	61	257.5	51.5	36.9	57
Colorado	31	5,584	438	7.8	87,672	15.7	4,952	5.6	11.3	18,380.3	210	1,036.3	5.6	209	9,326.5	50.7	106	465.0	5.0	44.9	94
New Mexico	14	1,157	274	23.7	21,554	18.6	2,911	13.5	10.6	4,834.6	224	761.7	15.8	262	2,285.7	47.3	106	396.3	17.3	52.0	136
Arizona	10	4,105	196	4.8	71,501	17.4	2,599	3.6	13.3	12,555.5	176	448.6	3.6	173	6,792.1	54.1	95	234.6	3.5	52.3	90
Utah	14	3,542	230	6.5	57,981	16.4	2,829	4.9	12.3	10,127.5	175	458.0	4.5	162	5,218.9	51.5	90	170.9	3.3	37.3	60
Nevada	3	998	37	3.7	20,056	20.1	346	1.7	9.4	3,220.1	161	69.5	2.2	201	1,589.2	49.4	79	24.0	1.5	34.5	69
Pacific	48	68,881	785	1.1	1,347,447	19.6	8,946	0.7	11.4	271,888.8	202	1,906.8	0.7	213	132,406.0	48.7	98	789.3	0.6	41.4	88
Washington	19	4,735	279	5.9	82,499	17.4	3,033	3.7	10.9	12,157.7	147	488.3	4.0	161	6,449.2	53.0	78	238.8	3.7	48.9	79
Oregon	2	4,203	49	1.2	96,752	23.0	468	0.5	9.6	15,223.8	157	101.7	0.7	217	6,721.8	44.2	69	66.6	1.0	65.5	142
California	19	58,983	376	0.6	1,136,025	19.3	4,543	0.4	12.1	238,777.6	210	1,092.0	0.5	240	116,810.4	48.9	103	354.5	0.3	32.5	78
Alaska	5	122	26	21.3	2,537	20.8	239	9.4	9.2	630.1	248	100.8	16.0	422	290.0	46.0	114	51.1	17.6	50.7	214
Hawaii	3	838	55	6.6	29,634	35.4	663	2.2	12.1	5,099.8	172	124.0	2.4	187	2,134.4	41.9	72	78.4	3.7	63.2	118
Outlying areas1	0	755	1	0.1	13,429	17.8	3	0.0	3.0	1,849.3	138	0.7	0.0	233	922.8	49.9	69	0.0	0.0	0.0	0

Less than 0.05 percent.

Includes Puerto Rico and other outlying areas.

NOTE: NA is not applicable.

SOURCE: Health Care Financing Administration, Bureau of Data Management and Strategy: Data from the Medicare Statistical System; data development by the Office of Research and Demonstrations.

In the North Central States, 18 percent of all admissions for SNF services were to swing-bed hospitals. In the South, almost 12 percent of SNF admissions were to swing-bed hospitals. In the largely urbanized Northeast, less than 1 percent of the admissions for SNF services were made to swing-bed hospitals. However, New Hampshire and Vermont are notable exceptions to the patterns of the Northeast. In these two States, more than one-fourth of the admissions for SNF services were to swing-bed hospitals. Admissions to swing-bed hospitals are based on the residence of the patient. Where admissions to swing-bed hospitals are noted in States with no swing-bed facilities, admission to a facility in a neighboring State is the probable explanation.

The West census region presents a dichotomy between the Mountain States and Pacific Coast States. In the Mountain States, almost 12 percent of the admissions for SNF services were to swing-bed hospitals. In four of the Mountain States (Montana, Idaho, Wyoming, and New Mexico), more than 20 percent of the admissions for SNF services were to swing-bed hospitals with Wyoming having almost 60 percent going to swing-bed hospitals. The remaining Mountain States show less than 10 percent of the admissions for SNF services going to swing-bed hospitals. Only 1 percent of the admissions for SNF services in the Pacific Coast States went to swing-bed hospitals; Alaska, with 21 percent, was the only Pacific Coast State with more than 7 percent using swing-bed hospitals for SNF care. Alaska had a total of only 122 admissions for SNF services.

The States showing more than 50 percent of the admissions for SNF services going to swing-bed hospitals were: North Dakota, South Dakota, Kansas, Mississippi (the highest at 89 percent), and Wyoming. Delaware and the District of Columbia were the only jurisdictions with no admissions for swing-bed services. Figure 1 displays the geographic patterns of admissions to swing-bed hospitals as a percent of all SNF admissions.

Figure 1

Percent of admissions for skilled nursing facility services admitted to swing-bed hospitals: United States, 1987

For the individual States, the relationship among admissions, covered days of care, charges, and reimbursement is about that indicated for 1987 in Table 2. A notable exception is Mississippi. As previously mentioned, about 89 percent of the admissions for SNF services in Mississippi went to swing-bed hospitals. Swing-bed hospitals accounted for almost 91 percent of the covered SNF days of care and received 82 percent of SNF reimbursements. Mississippi was the only State in which the average length of SNF stay in a swing-bed hospital (18.2 days) exceeded the statewide average (17.9 days).

Summary

The data presented in this article and the findings of the evaluation indicate that the rural hospital swing-bed program has been working as might have been anticipated:

Swing-beds have assumed the provision of a significant portion of post-acute care services in many States with large rural areas.

The post-acute case mix in swing-bed hospitals represent more short term, intense level of care requirements than those in SNFs. Swing-bed hospitals seem better suited to meeting nursing care needs of these types of patients than do rural SNFs, which seem more suited to meeting the needs of the traditional long-term care nursing home patients.

Higher average total charges per day for swing-bed patients suggest that they tend to be more expensive to care for than are the patients in SNFs; especially in the use of ancillary services.

Per diem reimbursements for swing-bed services have been growing at an average annual rate of about one-third of that for SNFs.

The latter finding raises question as to whether the current basis for reimbursing for post-acute routine care services in swing-bed hospitals causes per diem revenues to rise at a slower rate than per diem costs. The current difference between marginal costs and revenues seem sufficient to attract increasing participation by rural hospitals with fewer than 50 beds. However, given the different behavior of the overhead as well as the direct cost components of the costs for routine care services in hospitals and SNFs, the current method of paying for routine swing-bed services may require re-examination some time in the future. This may become more apparent when the experiences of the larger rural swing-bed hospitals brought into the program by the Omnibus Budget Reconciliation Act of 1987 (Public Law 100-203) are analyzed. Under this legislation, the swing-bed option was extended to rural hospitals with fewer than 100 beds. Providing an incentive to small rural hospitals to continue rendering swing-bed services may require re-examination of the bases on which payment for these services are made.