Back to Table of Contents

Rural Hospital Flexibility Program Tracking Project

Chapter 2
A Profile of Critical Access Hospitals

Stephanie Poley, B.A., Kathleen Dalton, Ph.D., Melissa Fruhbeis, M.S.P.H.,
Hilda Howard, B.S., and Thomas C. Ricketts, Ph.D.
North Carolina Rural Health Research and Policy Analysis Center
Cecil G. Sheps Center for Health Services Research, UNC-Chapel Hill

Introduction

As part of the effort to assess the progress of the implementation of the Rural Hospital Flexibility Program (RHFP, referred to throughout this report as the Flex Program), the North Carolina Rural Health Research and Policy Analysis Center at the University of North Carolina has developed a system to monitor and document the program activities in each of the participating states, especially the certification of Critical Access Hospitals (CAHs). This effort has enabled a steady stream of information about the program to be reported to the Federal Office of Rural Health Policy (FORHP), other government officials, and program participants at both the state and local levels. We aim to provide an overview of the overall progress of the program and assist the participants in understanding what opportunities and problems can be expected as the program evolves in their states. 

We began by identifying contacts who were knowledgeable about the process of program implementation in their states. Most often these contacts are Flex Program coordinators and/or Flex grantees. Through regular telephone and email contact we established working relationships with those contacts to continuously collect data. 

Contacts at the Centers for Medicaid and Medicare Services (CMS) (formerly known as the Health Care Financing Administration [HCFA]), were also identified in 1998 and 1999, and continue to provide us with information about newly designated CAHs as they enter the CMS reporting systems. Communication with our state Flex contacts occurs on a monthly basis, at which time updates to the CMS reporting systems are shared.

The data that were gathered form the basis of the data set that has been developed and described in Figure 1. The set of hospitals with their identification numbers and activity status, and contact names and numbers for states, serve as the core data set for linking other data. Additional data were extracted from various sources to provide a more detailed description of the states participating in the Flex Program, the communities in which CAHs operate, and the CAHs themselves. Such data include state and county characteristics, local health care resource availability, and hospital characteristics. The data that are collected from the contacts in the states and through federal agencies are then merged with other data sets to create a relational database. Chapter 2's Appendix B details the contents of this Flex Program database.

Information Gathered 

Hospitals Eligible for and Certified as Critical Access Hospitals

As of May 1, 2001, 401 hospitals in 38 states had been designated as Critical Access Hospitals.¹ State Flex Program coordinators estimate that more than 1,300 hospitals might be eligible for CAH status based on the states' designation criteria, but that a realistic projection of the number of hospitals expected to become CAHs is around 900. We know through our surveying efforts that many of the eligible hospitals have decided not to pursue CAH status for a variety of reasons discussed in Chapter 6 of this report. Our experience in monitoring the pursuit and designation of CAH status and our familiarity with small rural hospitals leads us to believe that the final number of CAHs may be somewhat lower than state estimates. Chapter 2's Appendix A describes hospital activities in individual states.

Nine of the 47 states participating in the Flex Program had not yet processed a CAH conversion as part of their Flex Programs as of May 2001; however, five of those states were approved for participation in the program only within the last year.² 

Growth in the number of certified CAHs has been steady since the program's inception and has accelerated substantially in the past year. Figure 2 illustrates the expansion in the number of certified CAHs since the program was authorized in 1997. In the initial months of the Flex Program, the hospitals being certified as Critical Access Hospitals were those that had been designated Rural Primary Care Hospitals (RPCHs) or Medical Assistance Facilities (MAFs) as part of the pilot projects to test the limited service hospital model. In fact, all but one of the first 39 facilities designated as CAHs prior to 1999 were located in states that had participated in the RPCH and MAF programs and had prior experience in implementing a Flex-like program.³

In 1999, the program began to gain momentum as states' rural health plans were approved by CMS and the CAH application and designation processes were finalized. With the exception of a few states, there was a period of at least five months, and in many states more than a year, between the approval of a state's rural health plan and its first CAH designation. In 1998 and the early part of 1999, CMS approved rural health plans for more than half of the states eligible for participation in the Flex Program, paving way for the first generation of CAHs that had not been RPCHs or MAFs. 

The most significant growth in the program has occurred in the most recent 17 months (January 2000 through May 2001). Since the beginning of 2000, some 280 hospitals have been certified as CAHs. Changes in the federal criteria for CAH eligibility, authorized by the Balanced Budget Refinement Act (BBRA) of 1999, expanded the number of facilities that could potentially participate in the program and may explain part of the rise in the number of certifications in 2000 and 2001.⁴ In the year since our last report to the Office of Rural Health Policy, 231 hospitals have been certified as CAHs, more than doubling the total number designated as of May 1, 2000. Twelve states have experienced their first CAH conversions and CMS has approved all but one state's rural health plans.⁵ Of the 231 new CAHs, 42 (18%) are located in twelve of the states that had no CAHs one year ago. Additionally, 56 (nearly 25%) of the 231 CAHs certified in the past 12 months are located in 3 states: Nebraska (20), Texas (19), and Iowa (17). In the first 5 months of the year 2001, 79 facilities joined the CAH program and we expect that the total number of new CAHs added during 2001 will be similar to the number of conversions in 2000. 


Geographic Distribution of Critical Access Hospitals

As Map 1 demonstrates, a pattern is evident in the geographic location of the 401 certified CAHs. The majority of the hospitals that have been certified are clustered in the middle of the country. More than one-third of the total number of certified CAHs are located in five states: Nebraska (48), Kansas (33), Texas (25), Iowa (20), and South Dakota (20). A substantial proportion of CAHs (more than 25% of the current CAH population) are located in the states that participated in the EACH and RPCH demonstration program or the MAF Program. The growth in CAH conversions in these states is likely strengthened by their prior experience with limited service hospital models. 

Critical Access Hospitals and their Communities: Analysis of Secondary Data

Hospitals Included in the Analysis

Critical Access Hospitals differ from other hospitals in many respects, and in order to demonstrate these differences, we have defined multiple groups of hospitals for our analyses. 

Our first study group, Group 1, includes 401 hospitals certified as CAHs as of May 1, 2001. During the time period for which these cost reports were filed, many of the hospitals now certified as CAHs were still filing for reimbursement under the Prospective Payment System (PPS). In order to examine the differences between the facilities that filed for cost-based reimbursement (CBR) and those that filed under PPS, we have divided the hospitals in Group 1 into two subgroups for analysis. 

The resulting subgroups include only those CAHs for which cost report data were available (N=327). The groups were assigned based on whether the hospital filed its 1998 CMS cost report using a CAH provider number. Group 1a includes CAHs that filed with a CAH provider number (i.e. filed for cost-based reimbursement). Group 1b includes CAHs that filed under a PPS provider number (i.e. did not file for cost-based reimbursement). 

For comparison, we also defined two groups of rural hospitals to analyze. We began by identifying all short-term acute care hospitals in rural areas using 1999 Metropolitan Statistical Area (MSA) designations. A total of 2,192 facilities fit this definition and are presented as Group 3. Note that Group 3 includes 380 CAHs (21 CAHs are in metropolitan areas and therefore excluded from this group). 

Because Group 3 includes many large suburban hospitals, we felt it was necessary to further dissect the group and create a subset of only the smallest rural hospitals, a group that would closely resemble, but not include Critical Access Hospitals. For the purpose of this analysis, we defined small hospitals as those with an Average Daily Census (ADC) (excluding swing patients) of 15 or less. Because this definition is founded on hospital operating statistics taken from CMS cost reports, only those facilities with data available (1,971 of the 2,192 hospitals in Group 3) were eligible for inclusion in the subset. After applying these criteria to facilities to Group 3 and removing all CAHs, 787 hospitals were retained for inclusion in Group 2.

Sources of Data for Analysis

In addition to the information collected through telephone interviews by the North Carolina Rural Research and Policy Analysis Program team, various secondary data sets have contributed to the description of hospitals and the communities in which they are located. State, hospital, and community-level data reported in our analyses come from the following sources:

• CMS Online Survey and Certification Reporting System (OSCAR).

• Hospital Cost Report Information System (HCRIS) for federal fiscal year 1998 (includes reports filed for hospitals with fiscal years ending between September 1998 and August 1999)⁶.

• Bureau of Primary Health Care listings of Federally Qualified Health Centers, Medically Underserved Areas and Health Professional Shortage Areas.

• Bureau of Health Professions, DHHS, HRSA Area Resource File.

What Are the Operating Characteristics of Critical Access Hospitals?

Capacity, Occupancy and Services: By program design, facilities certified as Critical Access Hospitals are small and are less complex in services available. Table 1a describes the basic operating characteristics of hospitals certified as CAHs and illustrates the differences between CAHs and other rural hospitals. 

Table 1a.  Capacity, Occupancy and Services

	Group 1 All CAHs	Group 1a CAHs that filed as CAHs in FY1998	Group 1b CAHs that were not yet filing as CAHs in FY1998	Group 2 Small Rural Hospitals (excluding CAHs)	Group 3 All Rural Hospitals (including CAHs)
	(N=327)	(N=73)	(N=254)	(N=787)	(N=1971)
Beds 1     Median     Mean	26 27	18 21	28 29	35 36	46 64
Percent of Facilities with Swing Beds1	84%	92%	82%	78%	58%
Average Daily Census (ADC Acute + Swing Beds1     Median     Mean	5.6 6.6	5.5 4.0	6.2 7.0	9.1 9.8	15.2 25.8
ADC (Acute Only)1     Median     Mean	3.2 3.9	1.6 2.4	3.7 4.4	7.2 7.3	12.6 24.0
Occupancy (Acute and Swing Beds)1     Median     Mean	22% 25%	21% 26%	22% 24%	26% 28%	34% 35%
Occupancy (Acute Patients Only)1     Median     Mean	12% 14%	10% 12%	13% 15%	19% 21%	28% 30%
Percent of Facilities with Some Obstetric Services1 (Defined as Facility Reporting at Least 3 Nursery Days)	32%	16%	39%	58%	61%
Percent of Facilities with an ICU1 (Defined as Facility Reporting at Least 1 ICU Day)	13%	3%	15%	30%	52%
1 Source:  Hospital Cost Report Information System Minimum Data Set, PPS 15, Health Care Financing Administration, 1997-1998.

With regard to bed size, CAHs are the smallest of the small rural hospitals. Under federal legislation, CAHs are limited to 25 beds, with no more than 15 to be used at any given time for acute care patients. Data obtained from CMS describing CAHs verify that the maximum size of CAHs is 25 beds, but as can be seen from the entries for Group 1b, over half of the recent converters had to reduce their reported bed capacity at the time of conversion. Not only do CAHs have fewer beds than most rural hospitals, they are much more likely to use some of those beds as swing beds for long-term care. While 84 percent of CAHs have licenses to operate swing beds, only 58 percent of all rural hospitals are authorized to use theirs as swing beds (swing beds are only authorized for hospitals with fewer than 100 routine beds). 

Consistent with their bed size, Critical Access Hospitals have a low average daily census (ADC); the ADC for CAHs is 3.9, compared to 7.3 for other small rural hospitals and 24.0 for all rural hospitals as a group. When swing patients are included in the calculation, the CAH daily census increases by 70 percent to 6.6 patients per day, implying that these facilities are making use of the swing beds afforded to them. With or without the swing patients, the average census in CAHs is lower than the average in the other small rural comparison group, and well below the statutory limit of 15 acute or 25 acute plus swing patients.

Based on the number of open, acute-care beds reported in the Medicare cost report, occupancy rates averaged only 34 percent for all rural hospitals. The average Critical Access Hospital used only 25 percent of its available bed days in fiscal year 1998. If swing days were excluded from the calculation, the acute-care occupancy rates for CAHs would be only 14 percent, compared to 21 percent for other small rural hospitals and 30 percent for all rural hospitals. 

We identified hospitals that provided obstetric and critical care services from cost report data, based on whether they reported at least three routine newborn days of care or any days of care in a segregated intensive care unit (ICU). Not surprisingly, Critical Access Hospitals were less likely to offer these services than other rural hospitals. While 61 percent and 52 percent of all rural hospitals offered obstetrics and had ICUs, respectively, only 32 percent of CAHs provided obstetrical services and 13 percent had an ICU. These numbers were much lower for Group 1a, which includes the former RPCHs and MAFs. We will not know until next year whether the more recently converted facilities in Group 1b will report similar participation rates after their conversion date.

Average Length of Stay: Critical Access Hospitals do not differ dramatically from other small rural hospitals in the average length of stay (ALOS), although the ALOS for all rural hospitals is somewhat higher (3.7 days, compared to 3.3 days for CAHs and 3.4 for other small rural facilities). Medicare patients also average shorter stays in CAHs than they do in rural hospitals in general (3.7 days, compared with 4.0 days in small rural comparison hospitals and an average of 4.4 days for all rural hospitals). The differences are somewhat more pronounced for Medicaid patients (2.7 days in CAHs, compared with 3.5 and 3.6 days in other small rural and all rural hospitals, respectively).

Table 1b.  Average Length of Stay
	Group 1	Group 1a	Group 1b	Group 2	Group 3
	All CAHs	CAHs that filed as CAHs in FY 1998	CAHs that were not yet filing as CAHs in FY 1998	Small Rural Hospitals (excluding CAHs)	All Rural Hospitals (including CAHs)
	(N=327)	(N=73)	(N=254)	(N=787)	(N=1971)
ALOS (Excluding Swing) 1
Median	3.2	3.0	3.3	3.3	3.5
Mean	3.3	3.2	3.4	3.4	3.7
Medicare ALOS (Excluding Swing) 1
Median	3.7	3.1	3.7	3.9	4.3
Mean	3.7	3.3	3.8	4.0	4.4
Medicaid ALOS (Excluding Swing) 1
Median	2.6	2.5	2.6	2.6	2.7
Mean	2.7	2.6	2.7	3.5	3.6
1 Source: Hospital Cost Report Information System Minimum Data Set, PPS 15, Health Care Financing Administration, 1997-1998

Medicare and Medicaid Utilization: The majority of all rural hospital inpatients are Medicare beneficiaries, but Critical Access Hospitals tend to be the most Medicare-dependent facilities. Table 1c shows that 62 percent of discharges and 68 percent of acute care days in the CAHs are attributable to Medicare beneficiaries. These are somewhat higher utilization rates than those for other small rural hospitals (54% and 61%, respectively) and higher still than the averages for all rural hospitals (52% and 60%, respectively).

As a group, Critical Access Hospitals have slightly lower Medicaid utilization rates than rural hospitals in general. This pattern is to be expected due to the small number of CAHs that provide obstetrical services. Medicaid accounts for 8 percent of CAHs total inpatient cases, compared to 10 percent for the comparison group of other small facilities and 13 percent for all rural hospitals. 

 

Table 1c.  Medicare and Medicaid Utilization
	Group 1	Group 1a	Group 1b	Group 2	Group 3
	All CAHs	CAHs that filed as CAHs in FY 1998	CAHs that were not yet filing as CAHs in FY 1998	Small Rural Hospitals (excluding CAHs)	All Rural Hospitals (including CAHs)
	(N=327)	(N=73)	(N=254)	(N=787)	(N=1971)
Utilization (Acute Discharges, Excluding Swing) 1
Medicare
Median	64%	70%	62%	56%	52%
Mean	62	68	61	54	52
Medicaid
Median	6	5	7	10	11
Mean	8	7	9	12	13
Utilization (Acute Days, Excluding Swing) 1
Medicare
Median	71%	74%	69%	62%	60%
Mean	68	72	45	61	60
Medicaid
Median	5	3	6	8	9
Mean	7	6	7	10	10
1 Source: Hospital Cost Report Information System Minimum Data Set, PPS 15, Health Care Financing Administration, 1997-1998

Financial Performance Indicators:  We reviewed the CMS cost reports for federal fiscal year 1998 and extracted data for Medicare inpatient costs. We also extracted figures for net operating income and total income from summarized financial statements that are included in these reports. 

The average cost per Medicare discharge during federal fiscal year 1998 was higher for CAHs than for other small rural hospitals. The CAHs in Group 1a (many of which come from the EACH/RPCH-MAF states) showed comparable costs per case to those in Group 2. However, CAHs that were still paid under the PPS, those in Group 1b, had Medicare costs per case that were more than 12 percent greater than those in the small hospital comparison group. The difference was even more pronounced when costs were summarized per day rather than per discharge; the Medicare cost per day was 18.6 percent higher for Group 1b than for other small rural hospitals. The differences between CAHs and all rural hospitals (Group 3) were not as great, but the average acuity for cases in all rural hospitals is also expected to be higher. The case mix for Group 2 is more likely to be comparable to that of CAHs. To some extent, we must consider this as evidence that, within the group of eligible CAH facilities, conversion is occurring where it was intended - that is, among those whose high unit costs prevent them from succeeding under prospective payment.

The negative mean and median financial margins experienced by CAHs provide clear evidence of the severity of their financial condition. At least half the CAHs had operating margins (net operating income expressed as a percent of net patient service revenues) that were below -11.4 percent. Those that converted early enough to file 1998 cost report as CAHs (Group 1a) showed slightly worse margins than those that converted later. Operating margins for all rural hospitals were low (mean -5.9%, median -3.3%) but CAHs, as a group, appear to have been in a substantially worse position. Perhaps even more important, the data on total margins indicate that hospitals in both Group 2 and Group 3 appeared to have been more successful in obtaining non-operating income to offset their operating losses. Total margins differ from operating margins in that they include additional income or fund transfers from tax-supported, philanthropic, or parent-company sources, in the computed ratio. The mean total margin for CAHs was still -3.4 percent, compared to -0.4 percent (nearly break-even) for the comparison group of other small rural hospitals, and +1.8 percent across all rural hospitals combined.