Nursing Home Care: Enhanced HCFA Oversight of State Programs Would Better Ensure Quality

Published by the Government Accountability Office on 1999-11-04.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                United States General Accounting Office

GAO             Report to the Special Committee on
                Aging, U.S. Senate

November 1999
                NURSING HOME
                Enhanced HCFA
                Oversight of State
                Programs Would Better
                Ensure Quality

          United States
GAO       General Accounting Office
          Washington, D.C. 20548

          Health, Education, and
          Human Services Division


          November 4, 1999

          The Honorable Charles E. Grassley
          The Honorable John B. Breaux
          Ranking Minority Member
          Special Committee on Aging
          United States Senate

          The federal government and the states are jointly responsible for ensuring
          that the nation’s more than 17,000 nursing homes provide adequate care to
          their highly vulnerable 1.6 million elderly and disabled residents. The
          Health Care Financing Administration (HCFA), within the Department of
          Health and Human Services, is responsible for ensuring that each state
          establishes and maintains a survey capability that effectively identifies and
          resolves problems in nursing homes that receive Medicare or Medicaid
          payments. Under contract with HCFA, state agencies conduct surveys at
          nursing homes to ensure that the homes provide quality care to residents.
          On the basis of their surveys, these agencies certify to the federal
          government that each home is in compliance with federal nursing home
          standards, which enables the home to receive federal payments. Federal
          payments to these nursing homes under the Medicare and Medicaid
          programs are expected to total $39 billion in 1999.

          In previous reports to you, we found that residents received an
          unacceptably poor quality of care in some nursing homes and that the
          federal and state programs designed to identify and correct these
          problems had significant weaknesses. For example, we reported that

      •   nearly a third of the 1,370 homes in California had been cited for care
          violations classified as serious under federal or state deficiency
      •   one-fourth of the nation’s nursing homes had serious deficiencies that
          caused actual harm to residents or that placed them at risk of death or
          serious injury and that 40 percent of these homes had repeated serious
      •   serious complaints alleging that nursing home residents are being harmed
          can remain uninvestigated for weeks or months, prolonging situations in

           California Nursing Homes: Care Problems Persist Despite Federal and State Oversight
          (GAO/HEHS-98-202, July 27, 1998).
           Nursing Homes: Additional Steps Needed to Strengthen Enforcement of Federal Quality Standards
          (GAO/HEHS-99-46, Mar. 18, 1999).

          Page 1                                        GAO/HEHS-00-6 Federal Nursing Home Oversight

    which residents may be subject to abuse, neglect resulting in serious care
    problems like malnutrition and dehydration, preventable accidents, and
    medication errors;3 and
•   when serious deficiencies are identified, federal and state enforcement
    policies have not been effective in ensuring that the deficiencies are
    corrected and remain corrected.4

    In response to these problems and our recommendations, HCFA has
    developed about 30 initiatives to strengthen federal standards, oversight,
    and enforcement for nursing homes.5 One of these initiatives is to enhance
    federal oversight of the state survey agencies to help ensure that the states
    are adequately protecting the health and safety of nursing home residents.
    When it is determined that a state agency is not adequately performing its
    survey responsibilities, HCFA has indicated it would develop appropriate
    sanctions to penalize the state agency, including terminating its contract.
    HCFA’s initiatives are fundamental to its ability to hold states accountable
    for reliably and consistently performing their contractual responsibilities
    for certifying that nursing homes meet Medicare and Medicaid standards
    and provide quality care for nursing home residents.

    Because an effective oversight program is critical to HCFA’s ability to gauge
    the states’ success in implementing HCFA’s many initiatives, you asked us
    to evaluate HCFA’s oversight programs of state agencies’ nursing home
    survey process. Specifically, we assessed (1) the effectiveness of HCFA’s
    approaches to assessing state agency performance, (2) the extent to which
    HCFA’s regional offices vary in their application of these approaches, and
    (3) the corrective actions available to HCFA when it identifies poor state
    agency performance. To do this work, we contacted HCFA’s 10 regional
    offices to obtain data about each region’s oversight programs from 1996 to
    the present; interviewed officials at HCFA’s headquarters in Baltimore as
    well as federal surveyors and their managers in HCFA’s regional offices;
    interviewed HCFA officials from the Atlanta, Kansas City, and Seattle
    regions and met with state surveyors and their managers in four states
    from these three regions—Florida, Missouri, Tennessee, and Washington;
    and reviewed data provided by HCFA and its regional offices regarding the
    number and types of oversight reviews conducted during the past 3 years.
    We conducted our work between March and September 1999 in
    accordance with generally accepted government auditing standards.

     Nursing Homes: Complaint Investigation Processes Often Inadequate to Protect Residents
    (GAO/HEHS-99-80, Mar. 22, 1999).
     GAO/HEHS-99-46, Mar. 18, 1999.
     A list of related GAO products is included at the end of this report.

    Page 2                                            GAO/HEHS-00-6 Federal Nursing Home Oversight

                   Since last year, HCFA has undertaken a series of initiatives intended to
Results in Brief   address quality problems facing the nation’s nursing home residents,
                   including redesigning its program for overseeing state agencies that survey
                   nursing homes to ensure quality care. The objective of HCFA’s oversight
                   program is to evaluate the adequacy of each state agency’s performance in
                   ensuring quality care in nursing homes, but the mechanisms it has created
                   to do so are limited in their scope and effectiveness. In addition, HCFA’s
                   oversight mechanisms are not applied consistently across each of its 10
                   regional offices. As a result, HCFA does not have sufficient, consistent, and
                   reliable data to evaluate the effectiveness of state agency performance or
                   the success of its recent initiatives to improve nursing home care. Given
                   the wide range in the frequencies with which states identify serious
                   deficiencies, HCFA cannot be certain whether some states are failing to
                   identify serious deficiencies that harm nursing home residents.
                   Furthermore, HCFA does not have an adequate array of effective sanctions
                   to encourage a state agency to correct serious or widespread problems
                   with its survey process.

                   HCFA’s  primary mechanism to monitor state survey performance stems
                   from its statutory requirement to survey annually at least 5 percent of the
                   nation’s 17,000 nursing homes that states have certified as eligible for
                   Medicare or Medicaid funds.6 But HCFA’s approach to these federal
                   monitoring surveys does not produce sufficient information to assess the
                   adequacy of state agency performance. To fulfill its 5 percent monitoring
                   mandate, HCFA makes negligible use of its most effective technique—an
                   independent survey done by HCFA surveyors following completion of a
                   state’s survey—for assessing state agencies’ abilities to identify serious
                   deficiencies in nursing homes. For the vast majority of states, HCFA
                   requires only one or two of these comparative surveys per state, per year.
                   Yet, in the 64 comparative surveys conducted from October 1998 to
                   August 1999, HCFA found deficiencies that were more serious than those
                   the state found in about two-thirds of the surveys, which suggests that
                   some state surveyors miss some serious deficiencies. But because of
                   elapsed time between the federal and state surveys, HCFA cannot tell
                   whether the differences between its survey results and those of the state
                   are attributable to poor state performance, such as underreporting by state
                   surveyors, or to conditions in the nursing home that changed since the
                   state survey. Rather than making extensive use of comparative surveys,
                   HCFA focuses 90 percent of its own survey efforts on “observational

                    The Omnibus Budget Reconciliation Act of 1987 refers to the surveys HCFA must conduct as
                   “validation surveys.” HCFA’s broad term for validation surveys is “federal monitoring surveys,” which
                   consist of comparative and observational surveys. Under HCFA’s current federal monitoring surveys,
                   observational surveys are technically referred to as Federal Oversight and Support Surveys.

                   Page 3                                          GAO/HEHS-00-6 Federal Nursing Home Oversight

surveys,” in which it relies on its regional surveyors to observe state
surveyors as they conduct at least a portion of their surveys. While this
approach is useful in many respects, including identifying training needs
for state surveyors, it also has a serious limitation as a way to evaluate
state performance. HCFA’s presence may make state surveyors more
attentive to their survey tasks than when they are not being observed (the
Hawthorne effect); therefore, this approach does not necessarily provide a
valid assessment of typical performance.

A second HCFA oversight mechanism also has significant shortcomings.
About 3 years ago, HCFA implemented the State Agency Quality
Improvement Program (SAQIP), a program under which the state agency
does a self-assessment to inform HCFA, at least once a year, whether the
state is in compliance with seven standard requirements. For instance, all
states are expected to evaluate their surveyors’ ability to correctly
document deficiencies in nursing homes and to conduct complaint
investigations effectively. SAQIP is limited as an oversight program,
however, because HCFA (1) does not independently validate the
information that the states provide, so it is uncertain whether all serious
problems are identified or whether identified problems are being
corrected, and (2) has no policy regarding consequences for states that do
not comply. For example, in our prior work, we found that some states
were not promptly reviewing complaints filed against nursing homes and
that these states had not identified this problem in their SAQIP reports to
HCFA.7 SAQIP also includes four indicators of state performance that HCFA,
rather than the states, assesses. SAQIP specifies, for example, that HCFA will
determine whether states conduct nursing home surveys within specific
time frames and enter the survey results into HCFA’s database. However,
the four indicators do not address some important aspects of a state
agency’s performance, such as the predictability of the timing of state

In addition to these weaknesses in its oversight programs, HCFA regions are
uneven in the way they implement them, resulting in limited assurance
that states are being held equally accountable to federal standards,
including the recent initiatives. Although HCFA established the current
federal monitoring surveys to develop a uniform national approach for
regions to follow in conducting federal oversight surveys, the regions use
different methods for selecting oversight reviews and conducting them.
Some regions, for instance, comply with HCFA guidance to select homes
with no established pattern of deficiencies, while other regions focus on

 GAO/HEHS-99-80, Mar. 22, 1999.

Page 4                               GAO/HEHS-00-6 Federal Nursing Home Oversight

             homes that the state has already identified as having serious deficiencies.
             Regions that take the latter approach are unlikely to identify situations in
             which state surveyors underreport serious deficiencies. The regions also,
             on average, spend very different amounts of time on observational
             surveys. While the average time spent on these surveys is 52 hours, the
             regions range from about 27 hours to about 71 hours to complete these
             surveys, thus raising questions about the level of effort some regions
             devote to gauging state performance. In addition, HCFA regional officials
             have varying views about SAQIP’s effectiveness as an oversight program. As
             a result, some regions supplement SAQIP information by extensively
             analyzing available survey performance data, while other regions do not
             believe there is a need to use these supplemental data to assess state
             survey performance.

             Even if HCFA identifies inadequate state agency performance, it currently
             does not have a sufficient array of effective remedies or sanctions at its
             disposal to ensure adequate state performance. When HCFA identifies poor
             state agency survey performance, it can employ one or more of several
             remedies, such as requiring the state to submit a plan of correction or
             providing special training to the state surveyors. If these remedies do not
             bring the state agency into compliance with survey standards, HCFA has
             two sanctions available—reducing a state’s survey and certification
             funding or terminating the agency’s survey contract. Because of the
             extreme nature of these sanctions, HCFA has only once reduced state
             funding and has never terminated a state agency’s contract. Although HCFA
             is considering additional sanctions, on the basis of our review of them, we
             believe that their potential to compel a state to improve its performance is

             To assist HCFA in effectively overseeing state agencies and achieving the
             goals of its broader initiatives, we are recommending that HCFA improve
             the scope and rigor of its state oversight mechanisms, improve the
             consistency of its oversight across its regions, and further explore the
             feasibility of additional remedies and sanctions for states that prove
             unable or unwilling to meet HCFA’s performance standards.

             HCFA is required by statute to establish an oversight program for evaluating
Background   the adequacy and effectiveness of each state’s nursing home survey
             process.8 If HCFA determines that a state agency’s survey performance is
             inadequate, it is authorized to impose appropriate remedies or sanctions

              See sections 1819 (g)(3) and 1919 (g)(3) of the Social Security Act.

             Page 5                                           GAO/HEHS-00-6 Federal Nursing Home Oversight

                                      against the state agency. Among HCFA’s remedies and sanctions are
                                      requiring the state to submit a written plan of correction explaining how it
                                      plans to eliminate the identified deficiencies; reducing federal funds for
                                      state survey and certification activities; and, ultimately, terminating HCFA’s
                                      contract with the state. HCFA surveyors in its 10 regional offices carry out
                                      the oversight of state agencies.

                                      Every nursing home that receives Medicare or Medicaid funding must
                                      undergo a standard survey conducted by the state agency no less than
                                      every 15 months. This survey entails a team of state surveyors spending
                                      several days on-site conducting a broad review of whether the care and
                                      services delivered meet the assessed needs of the residents.9 The Omnibus
                                      Budget Reconciliation Act of 1987 (OBRA 87) requires HCFA surveyors to
                                      conduct federal oversight surveys in at least 5 percent of the nursing
                                      homes in each state each year within 2 months of the state’s completion of
                                      its survey.10 The following table shows the number of nursing homes per
                                      HCFA region, the number of federal monitoring surveys each region is
                                      required to conduct in fiscal year 1999, and the number of federal
                                      surveyors who conduct nursing home monitoring surveys as of
                                      August 1999.

Table 1: Nursing Homes, Required
Federal Surveys, and Federal                                                            Federal monitoring               Federal nursing
Surveyors Available, by HCFA Region                                                    surveys required to               home surveyors
                                                                            Nursing meet 5% requirement in                available as of
                                      Regional office                        homes         fiscal year 1999                 August 1999
                                      Boston                                   1,170                              63                        12
                                      New York                                 1,020                              56                         7
                                      Philadelphia                             1,526                              84                        12
                                      Atlanta                                  2,772                            139                         18
                                      Chicago                                  3,784                            189                         22
                                      Dallas                                   2,398                            122                         11
                                      Kansas City                              1,693                              84                        12
                                      Denver                                     666                              37                         8
                                      San Francisco                            1,681                              89                        11
                                      Seattle                                    497                              32                         9
                                      Total                                   17,207                            895                     122

                                       The standard survey is used to meet HCFA’s requirement to certify homes for Medicare and Medicaid
                                       A minimum of five reviews must be conducted in each state each year, even if this brings the total
                                      number of required reviews to more than 5 percent.

                                      Page 6                                           GAO/HEHS-00-6 Federal Nursing Home Oversight

HCFA’s recent initiatives relating to assessing state agency performance are
the latest in a series of approaches HCFA has used since OBRA 87 was
enacted. Until 1992, HCFA conducted only comparative surveys, in which
federal surveyors performed an independent survey of a home and
compared their results with the state’s. Since 1992, HCFA has used a mix of
comparative and different variations of observational surveys. For
instance, from 1992 until July 1995, the regions used surveys in which they
directly observed individual state surveyors as they performed a survey,
but did not communicate with them until the last day of the survey.
Starting in 1995, HCFA regional surveyors observed the state surveyors and
actively communicated with them during the survey. Under this approach,
federal surveyors provided on-the-spot training to the state surveyors.
Starting in July 1996, HCFA allowed the regions to develop variations of this
approach, and by 1998, multiple regional variations existed. Among these
were partial observational surveys that focused on only parts of the
survey, and participatory surveys in which federal surveyors became
members of the state agency teams.

As part of its broader nursing home initiatives, in October 1998 HCFA
introduced its current program of overseeing state survey agency
performance, referred to as the federal monitoring survey. This program
modified HCFA’s prior oversight programs and has two components. The
first component is a comparative survey, in which a team of federal
surveyors conducts a complete, independent survey of a nursing home
after the state has completed its survey, and then compares the results
with the state’s.11 The second component, which is HCFA’s primary
monitoring technique, is an observational survey, in which generally one
or two federal surveyors accompany state surveyors to a nursing home
either as part of the home’s annual standard survey or as part of a revisit
or a complaint investigation.12 During these observational surveys, federal
surveyors watch the state surveyors perform a variety of tasks, give the
surveyors verbal feedback, and later provide a written rating of the state
surveyors’ performance to state managers. Basically, the current
observational surveys represent an extension of the several types of
observational surveys that HCFA’s regions have used over the previous 6
years. However, unlike earlier observational surveys, the revised surveys
are intended to have a national standard protocol, a national focal point

 In conducting these surveys, state and federal surveyors must use the survey protocol (that is, the set
of survey procedures) as set out in HCFA’s regulations.
  Revisits are surveys that are required after a nursing home has been found to have certain
deficiencies, in order to determine whether the home has corrected the deficiencies. A complaint
investigation is made when a complaint has been filed against the home.

Page 7                                           GAO/HEHS-00-6 Federal Nursing Home Oversight

                         for collecting data about the surveys, and a single national database for
                         tracking survey results.

                         In addition to the formal review activities required under the federal
                         monitoring surveys, HCFA has other sources of information available for
                         evaluating state agency performance. One such source, SAQIP, initiated in
                         1996, requires states, in partnership and collaboration with HCFA, to
                         develop and implement quality improvement action plans to address
                         deficiencies in the state’s survey process that either the state or HCFA has
                         identified. In addition to SAQIP, a few regions also use information from
                         HCFA’s database on survey results to assess state performance in areas
                         such as timeliness of providing information to nursing homes regarding
                         identified deficiencies and the timeliness of enforcement actions.

                         HCFA’s current strategy for assessing state agency survey performance has
Limitations Hinder       limitations that prevent HCFA from developing accurate and reliable
HCFA Oversight           assessments. The number of comparative surveys required to be
Programs’                completed each year is negligible in that only one or two are required in
                         most of the states, and over half of the comparative surveys are started
Effectiveness in         more than a month after the state completes its survey. Observational
Assessing State          surveys are also limited in their effectiveness because these tend to cause
                         state surveyors to perform their survey tasks more attentively than they
Survey Performance       would if federal surveyors were not present (the Hawthorne effect), thus
                         masking a state’s typical performance. Observational surveys have also
                         had other problems during their first year of implementation, such as the
                         fact that federal surveyors are not required to observe state surveyors
                         performing most survey tasks, the lack of an effective data system for
                         recording results, and the slowness of written feedback to state surveyors.
                         Finally, SAQIP does not require independent verification of states’
                         self-reported performance, and its standards do not address all important
                         aspects of the state survey process.

HCFA’s Use of            Although comparative surveys are the only oversight tool that furnishes an
Comparative Surveys Is   independent federal survey where results can be compared with those of
Negligible               the states, HCFA’s use of them is negligible. Conducting a sufficient number
                         of these comparisons is important because of concern that some state
                         survey agencies miss significant problems. For example, HCFA surveyors
                         found deficiencies that were more serious than those found by the state
                         surveyors in about two-thirds of the comparative surveys they conducted
                         between October 1998 and July 1999.

                         Page 8                              GAO/HEHS-00-6 Federal Nursing Home Oversight

As we reported in July 1998, state agency surveyors can miss problems
that affect the health and safety of residents.13 In addition, HCFA data show
significant variation in the extent to which state surveyors identify serious
deficiencies. For example, state survey agencies in Idaho, Kansas, North
Dakota, and Washington identified serious deficiencies in more than half
of their surveys. On the other hand, state surveyors in Maine, Colorado,
Tennessee, and Oklahoma identified such problems in only 8 to 13 percent
of their surveys.14 With such a range in identified serious deficiencies in
nursing homes, HCFA needs to know to what extent such data accurately
portray the quality of care provided or the adequacy of state survey agency

Of the 64 comparative surveys that HCFA completed between October 1998
and July 1999, 44 (69 percent) identified a more serious deficiency than
had the state surveyors. For example, during a comparative survey
conducted at a nursing home in Missouri in November 1998, HCFA found 24
deficiencies that it believes state surveyors should have, but did not,
identify during their review about 6 weeks earlier. One of these
deficiencies identified six residents whose nutritional status was not being
adequately assessed by the nursing home, resulting in significant weight
loss to several of them. One resident lost 19 percent of his weight between
June and October 1998. His weight at the time of HCFA’s survey was 93
pounds, which HCFA indicated was significantly below the resident’s
minimally acceptable body weight of 108 pounds. Less than 4 months after
this resident’s admission to the home, he had also developed two
moderately severe pressure sores, which the nursing home was
inappropriately treating with a cream that its manufacturer stated was not
intended to heal pressure sores but rather to prevent irritation to the skin.

Until 1992, comparative surveys were the sole method HCFA used to carry
out state agency oversight responsibilities. According to HCFA documents,
the agency began to decrease its reliance on comparative surveys in 1992
because (1) it was difficult to adjust for changes in the nursing home that
may have arisen between the dates of the state and the federal surveys,
(2) two separate surveys during a short time period created a strain on the
nursing home, (3) too much time had passed between the completion of
the state survey and the time the state received feedback from federal
surveyors for the state surveyors to recall the details of the survey, and
(4) comparative surveys were resource-intensive.

  GAO/HEHS-98-202, July 27, 1998.
  Based on standard surveys conducted between January 1997 and April 1999.

Page 9                                        GAO/HEHS-00-6 Federal Nursing Home Oversight

                                       Under the revised federal monitoring surveys, started in October 1998,
                                       HCFA acknowledged the need to do more comparative surveys than the 21
                                       that had been done over the previous 2 years. Nevertheless, under this
                                       program, only about 10 percent of the federal monitoring surveys
                                       completed each year must be comparative surveys, and the remaining
                                       90 percent may be observational surveys. Specifically, HCFA now requires a
                                       minimum of one comparative survey in states having fewer than 200
                                       nursing homes, two in states with 200 to 599 nursing homes, and three in
                                       states with 600 or more homes. Table 2 shows the minimum number of
                                       comparative surveys to be completed in each state and the District of

Table 2: Minimum Number of
Comparative Surveys Required Yearly,                              Minimum
by State                                                         number of
                                       Number of homes             surveys
                                       in state (as of May   required each   Number of
                                       1999)                          year      states States
                                       Fewer than 200                   1            20 Alaska, Arizona, Delaware,
                                                                                        District of Columbia, Hawaii,
                                                                                        Idaho, Maine, Montana, Nevada,
                                                                                        New Hampshire, New Mexico,
                                                                                        North Dakota, Oregon, Rhode
                                                                                        Island, South Carolina, South
                                                                                        Dakota, Utah, Vermont, West
                                                                                        Virginia, Wyoming
                                       200 to 599                       2            24 Alabama, Arkansas, Colorado,
                                                                                        Connecticut, Georgia, Indiana,
                                                                                        Iowa, Kansas, Kentucky,
                                                                                        Louisiana, Maryland,
                                                                                        Massachusetts, Michigan,
                                                                                        Minnesota, Mississippi, Missouri,
                                                                                        Nebraska, New Jersey, North
                                                                                        Carolina, Oklahoma, Tennessee,
                                                                                        Virginia, Washington, Wisconsin
                                       600 or more                      3             7 California, Florida, Illinois, New
                                                                                        York, Ohio, Pennsylvania, Texas

                                       While providing important information, the low number of comparative
                                       surveys will not permit HCFA to determine how representative these one to
                                       three surveys per state are of overall state performance.

                                       Page 10                               GAO/HEHS-00-6 Federal Nursing Home Oversight

Gap in Time Between State   HCFA  reestablished the comparative survey as part of its oversight survey
Survey and HCFA             process in 1998, but it has not adequately addressed the earlier concern
Comparative Survey Makes    about the time that elapses between the end of the state’s survey and the
                            start of the federal survey. We found that 33 days on average passed
Assessment of State         between these dates for comparative surveys completed from October
Performance Difficult       1998 through July 1999. Although this meets the requirements of OBRA 87,
                            the gap in time between the two surveys raises the possibility that changes
                            at the nursing home between the two surveys were responsible for
                            differences, thus calling into question the extent to which federal results
                            can be used to assess state performance.

                            OBRA 87 requires HCFA to start a comparative survey in a nursing home
                            within 2 months of the completion of the state’s survey. HCFA’s conclusions
                            about the state’s survey effectiveness are supposed to take into account
                            the difference in time. Because aspects of a nursing home, such as
                            residents, staffing, and ownership, can change in a short period of time,
                            developing a fair and accurate assessment of state surveyor performance
                            after a lapse in time can be difficult. Several state and HCFA surveyors we
                            interviewed told us that the time lag between the surveys continues to be a
                            problem. Our analysis shows that for the 64 comparative surveys that HCFA
                            completed between October 1998 and July 1999, 33 days, on average, had
                            passed from the time the state completed its survey until HCFA started its
                            comparative survey. Sixty-three of these comparative surveys started
                            within the 2-month time frame mandated by OBRA 87, while only one, which
                            started 68 days after the state’s survey, did not.15 Four regions averaged
                            fewer than 30 days, with the Dallas and Denver regions averaging 17 days.
                            At the other end of the scale, the Atlanta and New York regions averaged
                            43 days, while the Philadelphia region averaged 47 days. Table 3 shows the
                            number and time frames of comparative surveys conducted between
                            October 1998 and July 1999.

                              The median time for the start of the 64 comparative surveys was 32 days.

                            Page 11                                         GAO/HEHS-00-6 Federal Nursing Home Oversight

Table 3: Time Frames for the Start of
the 64 Comparative Surveys HCFA         Days from end of state survey to start                       Number of comparative
Conducted Between October 1998 and      of comparative survey                                                     surveys
July 1999                               0 to 10                                                                            7
                                        11 to 20                                                                          10
                                        21 to 30                                                                          12
                                        31 to 40                                                                          12
                                        41 to 50                                                                          10
                                        51 to 60                                                                           8
                                        Over 60                                                                            5

                                        To avoid the problems presented by gaps in time between the state and
                                        federal surveys, in August 1999 HCFA instructed its regions to start
                                        comparative surveys within 2 to 4 weeks after the state’s survey, because
                                        the less time that elapses between the two surveys, the less likely it will be
                                        that a home’s environment, staff, or residents will have changed. While this
                                        is an improvement, a delay of several weeks in the comparative survey
                                        could still result in problems. State and federal surveyors told us that
                                        comparative surveys are more effective and reliable in assessing state
                                        performance if they start immediately after the state has completed its
                                        survey, even as soon as the day after the state’s exit from the home.

Observational Surveys’                  HCFA  relies on observational surveys as its primary federal monitoring
Effectiveness Hindered by               technique. Observational surveys may help HCFA to identify state agency
Systemic and Transitional               training needs on a real-time basis, but several problems inhibit these
                                        surveys from getting a clear and accurate picture of a state’s survey
Problems                                capability. As designed, the surveys have a systemic weakness in that they
                                        require federal surveyors to observe state surveyors as they conduct a
                                        survey. Because they know they are being observed, state surveyors may
                                        be more attentive to survey tasks than they would normally be. In addition
                                        to this weakness, HCFA has encountered several specific problems as it
                                        transitions from its previous types of observational surveys to those now
                                        being conducted. For example, the surveys (1) cannot identify all
                                        significant deficiencies that state surveyors miss, in part because HCFA
                                        surveyors are not required to observe most of the tasks state surveyors
                                        perform; (2) necessitate that one HCFA surveyor must oversee the efforts of
                                        as many as three state surveyors at one time; (3) rate state surveyors for
                                        some survey activities that are not required by federal survey regulations;
                                        and (4) have had serious data system problems that prevent HCFA from
                                        assessing the results of observational surveys conducted since
                                        October 1998. In addition, HCFA surveyors have not always given timely

                                        Page 12                                  GAO/HEHS-00-6 Federal Nursing Home Oversight

written feedback to state agencies with respect to state surveyors’
performance so that corrective action can be implemented promptly.

Unlike comparative surveys, observational surveys do not require HCFA
surveyors to perform an independent review of a nursing home. Instead,
HCFA surveyors observe state surveyors as they perform portions of a
survey and rate them on one or more of eight possible survey tasks.16 As a
result of observing only a portion of the survey, HCFA surveyors cannot
determine whether state surveyors identify all significant deficiencies.
HCFA officials told us that observational surveys were not designed to
identify all deficiencies. They also said that of the 631 observational
surveys completed between October 1998 and August 1999, only 8
(1 percent) identified deficiencies that were more serious than those
identified by the state.17 During our interviews, nine of the regions
indicated that observational surveys allow them to help state surveyors
identify deficiencies that may otherwise be missed but that the surveys do
not ensure that HCFA surveyors identify all serious deficiencies. The Kansas
City region agreed with the other regions but also indicated that federal
surveyors can identify deficiencies missed by state surveyors if they
perform all eight survey tasks, as this region says it does, during an
observational survey.

During an observational survey, federal surveyors are generally required to
observe only two of the eight tasks—the preparation of the statement of
deficiencies and the resident review and quality-of-life assessments.18 As a
result, during any given survey, federal surveyors are not required to
observe most of the survey tasks that state surveyors perform.
Furthermore, some federal surveyors told us that even observing only the
required tasks can be a problem because a single federal surveyor has
sometimes been required to observe as many as three state surveyors at a
time. Although HCFA guidance to the regions suggests that one federal
surveyor should be able to observe the work of two or three state

  The tasks are to determine how well state surveyors perform (1) the off-site preparation activities,
(2) the entrance conference, (3) the initial tour of the home, (4) the selection of a sample of residents
for review, (5) the analysis of information they developed to determine deficiencies, (6) the exit
conference, and (7) the preparation of the statement of deficiencies. An eighth task, with six subtasks,
includes how well state surveyors observe general conditions of the home, kitchen and food service
activities, care and treatment of several sampled residents, quality of life for selected sampled
residents, administration of medications, and the adequacy of the home’s internal quality assurance
  These eight surveys were all from the Kansas City region.
 Each region must address three of the eight tasks in a given percentage of yearly observational
surveys. For instance, the sample selection task must be reviewed in at least 40 percent of the
observational surveys completed in a state during the year.

Page 13                                          GAO/HEHS-00-6 Federal Nursing Home Oversight

surveyors, federal surveyors we interviewed indicated that, in their
experience, it is generally not possible for one federal surveyor to
adequately observe more than two state surveyors at a time.

Several state surveyors expressed concern to us that some of the criteria
HCFA used to assess their performance are inconsistent with the survey
tasks required in federal guidelines. For example, although federal
guidelines do not require an employee of the nursing home to accompany
the state surveyors during the required tour of the home, a state surveyor
in Florida told us that he was criticized for conducting the tour without an
employee present. Officials from one HCFA region confirmed that such
situations have occurred and attributed them to a flaw in the federal
survey’s protocol that does not distinguish between optional and required
survey activities.

Inconsistencies between the federal guidelines and survey protocol
increased when HCFA issued a revision to the guidelines on July 1, 1999.
Some of these changes introduced significantly different survey
requirements, such as a new methodology for state surveyors to use in
selecting a sample of nursing home residents to review. A HCFA official
acknowledged that inconsistencies have existed since revised
observational surveys started in October 1998 and that the revision to the
federal guidelines has added to the inconsistencies. She told us that one of
several HCFA work groups now reviewing observational surveys is working
to identify and eliminate the inconsistencies.

In addition, the data system developed to support the observational
surveys has not been able to produce usable management reports.
Because of technical problems with the system, HCFA regions were not able
to enter survey results in the database for several months. As a result, HCFA
has been unable to use the database to identify poorly performing state
agencies or to determine needed corrective actions. The Dallas regional
office found these problems particularly troublesome and told us that its
surveyors are extremely frustrated with the database because it requires
too much time to input data and does not allow easy access to the data
that have been entered. As a result, the region has not been able to use the
database to analyze the results of its surveys.

HCFA officials told us that problems with the data system occurred because
HCFA rushed its development in order to meet the program implementation
date of October 1998 and thus did not follow standard systems
development practices. For example, a requirements analysis was not

Page 14                             GAO/HEHS-00-6 Federal Nursing Home Oversight

completed to ensure that the system would provide all the information
HCFA would need to assess state performance. HCFA entered into a contract
in June 1999 to address these problems and to add additional capabilities,
including the creation of a new database for surveys conducted after
October 1, 1999. However, the results of surveys conducted in fiscal year
1999 will not be included in the new database. Although a HCFA official told
us that it would be unacceptable for regions not to consider survey results
from the first year of the revised surveys in determining a state survey
agency’s performance, there has been no guidance from HCFA requiring the
regions to consider information in both databases when assessing state
performance. Moreover, the observational survey database is now
maintained separately from the database that HCFA uses to track identified
deficiencies in nursing homes. Although HCFA plans to include the results
of observational surveys in its redesigned central database, this redesign
will not take place for several more years, according to HCFA officials.

Finally, although formal written feedback is not required as part of
observational surveys, the regions we visited provide feedback in this way
to state agencies. To be useful, this feedback should be provided in a
timely manner to both the state surveyors who performed the survey and
their managers so that any needed corrective action can be taken.
Nevertheless, in three of the four states we visited, the surveyors and their
managers sometimes did not receive the written feedback for 3 to 5
months after the survey was completed. Furthermore, in some cases, state
surveyors were surprised at the content of the written feedback because it
was much more critical of their performance than the initial verbal
feedback they received from HCFA surveyors during the survey. Although
the revised surveys started in most states in October 1998, an official of
the Association of Health Facility Survey Agencies testified that as of late
June 1999, state agency managers from a majority of states had received
no formal feedback. She further indicated that oversight without feedback
is not effective in improving quality.19

Some state agency officials told us that HCFA’s delays in providing written
feedback have prevented the agencies from initiating important corrective
actions when problems with state surveyor performance were identified.
For example, for an observational survey conducted in Florida in October
1998, HCFA did not give feedback to the state until February 1999. This
survey found, among other things, that the state surveyors did not
correctly select the resident sample and that HCFA surveyors had to
intervene to ensure that the sample was appropriate. For another survey

  Testimony before the Senate Special Committee on Aging, June 30, 1999.

Page 15                                        GAO/HEHS-00-6 Federal Nursing Home Oversight

                           conducted in Florida in March 1999, HCFA gave feedback to the director of
                           the state’s survey program in early April 1999, but the state surveyors who
                           participated in the review and their manager did not receive the written
                           results until May 1999. In this survey, HCFA found that the state surveyors
                           were too quick to accept explanations of apparent deficient practices by
                           the nursing home’s staff, thus permitting potentially deficient practices to

SAQIP Does Not Provide     In addition to the federal monitoring surveys that are required by statute,
HCFA With Complete State   HCFA instructs its regions to use SAQIP to improve and monitor certain state

Agency Performance Data    agency survey-related activities. When a state is not in compliance with a
                           SAQIP standard, HCFA is expected to help the state develop a corrective
                           action plan and to work with the state as a partner to correct the problem.
                           However, SAQIP cannot ensure that HCFA knows of significant problems in a
                           state agency’s activities because SAQIP relies on the state agency to self
                           certify to HCFA whether it is in compliance with particular requirements
                           associated with its performance. Because HCFA does not independently
                           verify the information the states provide, it has no assurance that states
                           surface all serious problems or that they correct the problems they have
                           identified. Although SAQIP also includes some indicators of state
                           performance that HCFA itself assesses, these indicators do not address all
                           the important aspects of a state agency’s activities.

                           SAQIP replaced HCFA’s State Agency Evaluation Program in 1996.20 Under
                           the previous program, HCFA’s regions analyzed data maintained in HCFA’s
                           databases to evaluate some indicators of state performance and conducted
                           on-site reviews at state survey agencies of state documentation to assess
                           compliance with other indicators. For example, to assess a state’s
                           complaint process, regional surveyors visited the state to review state
                           documentation from a sample of complaints filed against nursing homes
                           and determine whether the state responded appropriately to the
                           complaints. Each region was required to follow the same procedures in
                           reviewing all state agencies, the states were rated on specific activities,
                           and the results of each state’s rating were compiled in a single report.

                           In essence, SAQIP includes nearly all of the same broad standards that the
                           previous program included, but it shifts the responsibility for assessing
                           compliance with seven standards from the regions to the states. SAQIP does
                           not require that HCFA independently verify the state’s assertion.

                            Both the State Agency Evaluation Program and SAQIP apply to all types of providers, including home
                           health agencies, kidney dialysis facilities, nursing homes, and others.

                           Page 16                                       GAO/HEHS-00-6 Federal Nursing Home Oversight

                                  Compliance with the other four standards for nursing homes is determined
                                  by HCFA (see table 4).

Table 4: SAQIP Nursing Home
Performance Standards for State   Responsible entity      Performance standards
Agencies                          State                   — Properly document deficiencies discovered in a nursing home
                                                          — Ensure that nursing homes’ plans of correction accepted by the
                                                          state agency reflect appropriate actions and time frames to correct
                                                          cited deficiencies
                                                          — Conduct all surveys with qualified individuals
                                                          — Ensure consistency in survey performance
                                                          — Measure the accuracy and improve the consistency in applying
                                                          enforcement actions against nursing homes
                                                          — Effectively investigate and process complaints filed against
                                                          nursing homes
                                                          — Monitor expenditures and support charges to federal programs
                                                          in accordance with regulations
                                  HCFA                    — Ensure that each nursing home is subject to a standard survey
                                                          not more than 15 months after its previous survey and that the
                                                          statewide average between standard surveys does not exceed 12
                                                          — Ensure that all surveys, including complaint investigations, are
                                                          conducted unannounced or are announced consistent with HCFA
                                                          — Ensure that the state agency’s annual budget request, activity
                                                          plan, and expenditure reports are prepared and submitted in
                                                          accordance with federal instructions and accurately reflect the
                                                          allocation of costs between state and federal programs
                                                          — Ensure that the state agency effectively maintains the database
                                                          HCFA uses to record survey results

                                  As a result of this change in SAQIP’s design, HCFA has no assurance that a
                                  state is in compliance with the first seven standards. For instance, for the
                                  standard that requires states to “effectively investigate and process
                                  complaints filed against nursing homes,” our March 22, 1999, report noted
                                  that Michigan’s SAQIP acknowledged the state had not determined whether
                                  it was investigating and processing complaints in accordance with state
                                  time frames; yet, the state indicated that it believed it was doing so. When
                                  we reviewed the state documentation, we found that more than 100
                                  pending complaints filed against Michigan nursing homes remained
                                  uninvestigated weeks and even months after their receipt, and that
                                  complaints that were investigated had not been investigated within
                                  Michigan’s required time frames.21 The Atlanta regional office staff also
                                  told us that it identified more significant problems in a state when it used
                                  the previous evaluation program than it does using SAQIP. From this
                                  region’s perspective, SAQIP does not provide for an in-depth evaluation of a

                                    GAO/HEHS-99-80, Mar. 22, 1999.

                                  Page 17                                    GAO/HEHS-00-6 Federal Nursing Home Oversight

                           state agency’s survey or operational performance because it depends on a
                           state’s willingness to devote adequate resources for assessment and to be
                           candid in identifying problems to HCFA.

                           In addition to removing HCFA’s direct responsibility for verifying state
                           compliance with the seven standards, SAQIP does not set national
                           performance thresholds. For instance, under the previous evaluation
                           program, HCFA reviewed a sample of deficiencies identified by the state
                           survey agency to determine whether at least 90 percent of them were
                           properly documented. If a state met this 90-percent threshold, its
                           performance for this standard was acceptable. SAQIP leaves the
                           determination of such thresholds to each state, and we found wide
                           differences among states in the percentages they use. For example, in
                           1998, Oklahoma’s goal was 75 percent, Louisiana’s goal was 90 percent,
                           and Wyoming’s goal was 100 percent.

                           One of HCFA’s goals in revising its federal monitoring surveys was to
Regions Are                establish consistency among the regions in the process used to assess
Inconsistent in How        state performance. However, differences still exist among the regions in
They Conduct               how they select and conduct oversight surveys. Although regions may
                           need some flexibility in selecting surveys for review, conducting them
Oversight Activities       requires a high level of consistency among the regions to ensure that states
                           are being held equally accountable to federal standards. Additionally, HCFA
                           regions differ in their view of SAQIP’s effectiveness as an oversight tool, and
                           some supplement HCFA’s federal monitoring surveys and SAQIP by analyzing
                           other available survey data to assess state agency performance. Finally,
                           the regions spend very different amounts of time, on average, conducting
                           observational surveys, which raises questions about the scope and quality
                           of their reviews.

Regions Use Different      When HCFA established the new federal monitoring survey in October 1998,
Criteria and Methods for   it suggested that, for comparative surveys, the regions select homes for
Selecting and Conducting   which the state did not find deficiencies categorized as “immediate
                           jeopardy,” “actual harm,” or “substandard quality of care” (see app. I for
Surveys for Review         definitions of these deficiency categories).22 However, only the Dallas,
                           Denver, San Francisco, and Seattle regions indicated that they routinely
                           follow this guidance. Other regions generally used a variety of other
                           criteria, including selecting homes for which the state has identified

                             In commenting on a draft of this report, HCFA indicated that the focus of its comparative surveys for
                           the next year will be on deficiency-free nursing homes.

                           Page 18                                         GAO/HEHS-00-6 Federal Nursing Home Oversight

continuing serious problems or homes located in diverse geographic areas.
Some regions also indicated that when conducting comparative surveys,
they specifically select homes designated by HCFA for special focus
because of the homes’ poor performance histories.23 By focusing on homes
that were identified by state surveyors as having serious problems, HCFA’s
surveyors are less likely to identify situations in which state surveyors
underreported serious deficiencies.

Another significant difference we found among the regions in conducting
comparative surveys is the way they select a sample of nursing home
residents for review. State surveyors select a sample of a home’s residents
as part of their procedures for evaluating the care and treatment given to
the residents. For this sample, the surveyors may review a resident’s
medical record, interview residents and their relatives, and observe the
environment and care practices of the home. Sample selection is a key
survey task for determining whether a nursing home is in compliance with
federal survey regulations and is providing appropriate care to its
residents. In conducting a comparative survey, federal surveyors
determine whether the state surveyors selected an appropriate sample of
residents, and then also select a sample of residents as part of the
comparative survey.

The regions vary in how they select resident samples, with some regions
selecting a sample that includes some overlap with the state sample and
other regions making no attempt to do so. For example, the Kansas City
region tries to include in its sample one-half of the residents who were
included in the state’s sample, believing that this practice allows it to more
closely duplicate the state’s survey and thus obtain a more valid
assessment of the state’s performance. On the other hand, the Atlanta
region makes no attempt to include any of the same residents in its
samples, believing that if systemic care problems exist in a home, any
sample will disclose these problems. We believe that, to better determine
the reasons for discrepancies between comparative and state surveys,
federal surveyors should sample as many of the same residents as the state
sampled in cases in which federal surveyors determine that the state
sample was appropriate. Federal and state surveyors we interviewed
agreed that reviewing the same sample would improve the consistency of
comparative surveys among HCFA’s regions. However, they also noted that
reviewing the same sample would require the comparative surveys to start
much sooner than 2 months after the state’s survey.

  Homes selected by HCFA for special focus are those that continually have serious problems and
require intense state agency monitoring as part of HCFA’s nursing home initiatives. HCFA has
designated two special-focus nursing homes in each state.

Page 19                                        GAO/HEHS-00-6 Federal Nursing Home Oversight

                           Each of the regions also uses a variety of criteria to select nursing homes
                           for observational surveys. HCFA’s guidance for selecting homes instructs
                           the regions to use type, size, and geographic location of a nursing home, as
                           well as to consider other criteria, including the performance differences
                           among state survey offices in the region. Although observational surveys
                           give HCFA an opportunity to provide on-site training to state surveyors,
                           HCFA’s guidance does not ensure that the regions will observe as many
                           state surveyors as possible. For example, the Atlanta and Seattle regions
                           indicated that surveys are often selected primarily on the basis of the
                           characteristics of the nursing home and not in order to assess a broad
                           range of state surveyors. At the time of our visit to the Atlanta region, it
                           had completed seven observational surveys in two of Tennessee’s three
                           districts, but none in the third district. Furthermore, 7 of the 20 state
                           surveyors reviewed during these surveys had been reviewed two or three
                           times, while over two-thirds of Tennessee’s surveyors had not been
                           reviewed at all. The Atlanta region did not indicate any particular problem
                           with the performance of these surveyors that would require repeated
                           review of their performance.

Regions Differ in Their    HCFA regions also vary in how valuable they believe SAQIP is as an oversight
View and Use of SAQIP      program. Some regions believe that SAQIP is effective, while others believe
and Other Available Data   some SAQIP standards encourage states to improve the quality of their
                           survey programs but other standards do not surface all serious problems.
                           The Atlanta regional office told us that it does not believe SAQIP is an
                           effective method for tracking and reporting operational problems in a
                           state agency. Believing that SAQIP has limitations, some regions supplement
                           it by analyzing data included in HCFA’s databases to evaluate state agency
                           performance in areas such as survey predictability.

                           The Seattle region, which played a major role in developing SAQIP, believes
                           SAQIP is an effective oversight tool because it allows state agencies to
                           identify survey performance problems unique to a state and permits the
                           state agency to develop corrective action plans. The Philadelphia, Dallas,
                           and Denver regions believe that some of the seven SAQIP quality
                           improvement standards the states assess themselves are effective in
                           encouraging states to improve their survey processes, but they also believe
                           that the four remaining standards for which HCFA assesses performance
                           are incomplete. For example, although one standard requires regions to
                           review data to determine whether a state performs its surveys within
                           statutory time frames, it does not require the regions to assess the same
                           data to determine whether a state’s survey schedule allows nursing homes

                           Page 20                            GAO/HEHS-00-6 Federal Nursing Home Oversight

                                    to predict the date of their next survey. Avoiding predictability is
                                    important because the extent of care problems in a nursing home can be
                                    hidden if nursing homes can predict when their next survey will occur.

                                    The Atlanta regional office recently started a program to conduct in-depth
                                    analyses of each state agency in its region by using available survey data.
                                    Through these analyses, the region recently determined that the annual
                                    state surveys of nursing homes in four of the region’s eight states (South
                                    Carolina, Kentucky, Tennessee, and Mississippi) are highly predictable,
                                    contrary to HCFA policy. It also found that in most of the states where it has
                                    completed reviews, state surveyors were not conducting revisits of nursing
                                    homes to determine whether identified deficiencies have been corrected
                                    within HCFA’s recommended time frame. Although HCFA recommends that a
                                    state wait no longer than 55 days to conduct its first revisit to a nursing
                                    home, four of the six states for which the Atlanta region completed
                                    reviews did not meet this time frame more than half of the time.24 Table 5
                                    presents the results of the Atlanta region’s analysis of state time frames
                                    related to this requirement. The Atlanta region also found several
                                    problems with Florida’s timeliness in conducting revisits, including one
                                    case in which state surveyors did not conduct a revisit until about 10
                                    months after the home’s original survey.

Table 5: Percentage of Surveys in
Which First Revisit Was Not Made                                                                                        Percentage of
Within HCFA’s Recommended Time                                                                                      surveys in which
Frame                                                                                                                 first revisit was
                                                                                                                     not made within
                                                                                                   Number of           recommended
                                    State                                                 revisits conducted                time frame
                                    Alabama (through Jan. 20, 1999)                                           21                     76
                                    Georgia (through Mar. 31. 1999)                                         105                      77
                                    Kentucky (through June 1, 1999)                                         108                      22
                                    Mississippi (through June 16, 1999)                                       69                     61
                                    North Carolina (through Apr. 20, 1999)                                    60                     53
                                    Tennessee (through Mar. 3, 1999)                                          84                     10

                                    After identifying problems with state performance, Atlanta officials meet
                                    with top-level survey and certification management officials in each state
                                    to discuss the results and present the data used to develop the findings.
                                    They then follow up this meeting with a formal letter to the state

                                     The Atlanta region had not completed comparable analyses of Florida or South Carolina at the time
                                    of our visit.

                                    Page 21                                        GAO/HEHS-00-6 Federal Nursing Home Oversight

                                     summarizing the results of the meeting and identifying actions the region
                                     expects the state to take to correct identified problems. These
                                     management reports, which the Atlanta region plans to begin issuing
                                     quarterly, are expected to provide the region with a documented history of
                                     a state agency’s performance over time, making it possible to more readily
                                     identify patterns of poor state performance.

Variation in Resources,              In addition to the variations among regions in the methods they use to
Survey Time Across                   conduct reviews, variations exist in the level of oversight resources
Regions                              available to them. Table 6 contains the ratio of state to federal surveyors,
                                     the ratio of observational surveys to be completed in fiscal year 1999, and
                                     the number of federal surveyors available to complete them. It also shows
                                     the average number of hours each region spent conducting surveys from
                                     October 1998 through July 1999.

Table 6: Variation in Resources
Available for and Time to Complete                                                            Ratio of   Average no. of
Observational Surveys                                       Number of                   observational        hours per
                                                       federal nursing      Ratio of          surveys     observational
                                                                 home       state to      required in      survey (Oct.
                                                         surveyors (as       federal   1999 to federal      1998 - July
                                     Region            of August 1999)    surveyors        surveyors              1999
                                     Boston                         12     14.4 to 1          4.6 to 1             26.9
                                     New York                        7     33.3 to 1          7.1 to 1             31.0
                                     Philadelphia                   12     15.8 to 1          6.2 to 1             48.7
                                     Atlanta                        18     32.8 to 1          6.8 to 1             60.9
                                     Chicago                        22     30.7 to 1          8.0 to 1             70.6
                                     Dallas                         11     59.8 to 1         10.2 to 1             37.5
                                     Kansas City                    12     29.8 to 1          6.3 to 1             50.6
                                     Denver                          8     17.6 to 1          3.8 to 1             58.8
                                     San Francisco                  11     26.5 to 1          7.5 to 1             53.6
                                     Seattle                         9     15.7 to 1          3.0 to 1             51.6
                                     Nationwide                    122     28.3 to 1          6.6 to 1             52.2

                                     Table 6 illustrates that federal surveyors in some regions must conduct
                                     significantly more observational surveys than surveyors in other regions.
                                     For example, at one extreme, Seattle surveyors are required to complete
                                     an average of 3 observational surveys per surveyor per year, while at the
                                     other extreme, Dallas surveyors must perform more than 10. Officials from
                                     the Dallas region told us that they did not have the resources available to
                                     complete the required review of 5 percent of state surveys in fiscal year

                                     Page 22                             GAO/HEHS-00-6 Federal Nursing Home Oversight

                             1998 and indicated that the shortage would likely prevent them from
                             meeting their survey requirements again in fiscal year 1999 unless they
                             received help from other regions. The region attributed this shortage to
                             the retirement of several surveyors at the end of the prior year, who had
                             not yet been replaced, and the need to focus on other, higher-profile
                             projects. A Dallas region official told us that the region plans to hire two
                             more surveyors this year, but it will nevertheless need an additional four
                             or five surveyors to meet its oversight requirements.

                             As shown in the last column of table 6, there are also large differences in
                             the average time that regions have invested to complete observational
                             surveys since October 1998, ranging from about 27 hours in the Boston
                             region to nearly 71 in the Chicago region. HCFA officials could not explain
                             the reasons for such variation. One official theorized that the variations
                             might be due simply to differences in how the regions account for the time
                             they spend conducting these surveys but also said that the variations could
                             signify a difference in the content and quality of the surveys being

                             Although HCFA has authority under the Social Security Act to take
HCFA’s Options for           corrective action against a state agency that performs inadequately in
Addressing Poorly            conducting surveys, HCFA does not now have an adequate array of effective
Performing State             sanctions. HCFA may use several remedies to encourage a state to improve
                             performance. When remedies fail, HCFA may impose either of two
Agencies Are                 sanctions—reducing the state’s funding for survey and certification
Inadequate                   activities or terminating the state’s survey contract. However, HCFA has
                             only once reduced a state’s survey and certification funding for failure to
                             conduct surveys in accordance with HCFA regulations, and it has never
                             terminated a state’s contract.25 HCFA is considering regulations to authorize
                             two additional actions that it classifies as sanctions, but, on the basis of
                             our review of the proposed regulations, we believe their effectiveness is

HCFA’s Criteria for          HCFA has the authority to take certain actions against a state agency that
Applying Present             performs poorly.26 Currently, HCFA defines inadequate state agency
Sanctions Limits Their Use   performance to include a state’s failure to identify an instance in which

                              HCFA reduced survey and certification funding to the California state agency for its refusal to
                             conduct surveys of nursing homes in 1990 and 1991 in accordance with OBRA 87.
                               See, for example, sections 1819 (g)(3)(C) and 1919 (g)(3)(C) of the Social Security Act.

                             Page 23                                          GAO/HEHS-00-6 Federal Nursing Home Oversight

                                       nursing home residents are placed in immediate jeopardy, or if a state
                                       agency demonstrates a pattern of other problems, including failure to

                                   •   identify deficiencies when the failure cannot be explained by changed
                                       conditions in the home;
                                   •   cite only valid deficiencies; and
                                   •   conduct surveys in accordance with federal requirements.

                                       When HCFA identifies inadequate state performance, its first step is to
                                       impose one or more of several remedies against the state, with the
                                       objective of improving the state’s performance. The most commonly used
                                       remedy has been to provide training for individual state surveyors or
                                       survey teams. A HCFA official indicated that other remedies are
                                       occasionally used but that training is viewed as the preferred approach.
                                       When remedies fail to improve state performance, HCFA can impose
                                       sanctions, which are intended to penalize a state agency for failing to
                                       improve performance. See table 7 for the remedies and sanctions now
                                       available to HCFA.

Table 7: Remedies and Sanctions
Available to HCFA in Response to       Tool                       Action
Poor State Agency Performance          Remedies                   — Provide training to survey teams
                                                                  — Provide technical assistance to the state with respect to
                                                                  scheduling surveys and following survey procedures
                                                                  — Require the state agency to implement improvements identified
                                                                  in the state’s plan of correction
                                                                  — Assume responsibility for developing survey schedules for the
                                                                  state agency
                                       Sanctions                  — Reduce the state’s Medicaid funding for survey and certification
                                                                  — Terminate the state’s survey contract
                                       Note: Remedies and sanctions are shown in the order they appear in HCFA documentation.

                                       To reduce a state’s survey and certification funding, HCFA must
                                       demonstrate that the state displays a pattern of failure to identify
                                       deficiencies in nursing homes.27 To develop this information, HCFA
                                       compares the deficiencies it identifies through its surveys of nursing
                                       homes with those found by state surveyors. When HCFA determines that the
                                       quarterly disparity rate between the deficiencies it identified and those the
                                       state agency identified is greater than 20 percent in at least three of the

                                         HCFA cannot reduce survey and certification funding to a state or terminate the state’s contract
                                       based on a single instance of a state’s failure to identify a situation that places the health or safety of a
                                       resident in immediate jeopardy.

                                       Page 24                                            GAO/HEHS-00-6 Federal Nursing Home Oversight

last four quarters and the state fails to correct the disparity within the next
two quarters, it may impose this sanction.

Currently, the only way HCFA can develop the specific quantitative
evidence it needs to identify disparity rates is through comparative
surveys. However, as previously discussed, the regions conduct too few
such surveys to establish a pattern of inadequate state agency
performance. For instance, 43 states and the District of Columbia will have
only one or two comparative surveys per year. A HCFA official
acknowledged that conducting so few comparative surveys is insufficient
to establish a pattern of poor performance in any state. Therefore, this
official told us, HCFA attempts to rely on observational surveys to obtain
the needed quantitative evidence, but the official agreed that the present
observational surveys are not designed to provide this information. As
noted, 69 percent of the comparative surveys completed as of August 1999
identified deficiencies that were more serious than those found by state
surveyors, but only in 1 percent of their observational surveys did the
regions identify deficiencies that were more serious. This is largely
because the goal of observational surveys is not to identify all deficiencies
but rather to observe state survey performance. Thus, it does not appear
that either the comparative or the observational surveys can provide the
basis HCFA needs to use the sanction of reducing a state’s Medicaid survey
and certification funding for failing to properly identify deficiencies in
nursing homes.

Terminating a contract with a state agency would require HCFA to find a
replacement for the state surveyors. HCFA officials indicated to us that
three alternatives exist: (1) use federal surveyors to conduct surveys in a
state, (2) contract with another state agency in the same state, or
(3) contract with a state agency from another state. According to HCFA and
state survey agency officials, HCFA does not have a sufficient number of
surveyors to conduct surveys in states that have a large number of nursing
homes. Texas alone has 402 state surveyors, while HCFA has 122 surveyors
nationwide. Six other states also have more surveyors than HCFA.
Furthermore, even in smaller states, the use of federal surveyors for an
extended period could become a problem because they would not be able
to perform their normal monitoring duties and other responsibilities. With
regard to the second alternative, the President of the Association of Health
Facility Survey Agencies told us that contracting with another agency
within the state is unrealistic given the negative atmosphere created
between HCFA and the state if the state agency’s contract has been
terminated. The third alternative, contracting with another state, would be

Page 25                              GAO/HEHS-00-6 Federal Nursing Home Oversight

                                 feasible only if the contracting state had or could hire a sufficient number
                                 of surveyors to conduct surveys in both states.

Potential Effectiveness of       As part of its nursing home initiatives, HCFA established a task force in late
Proposed Sanctions Is            1998 to expand and clarify the definition of inadequate state survey
Doubtful                         performance and to suggest additional remedies and sanctions that HCFA
                                 could take against state agencies that perform poorly. HCFA indicated to us
                                 that new instructions should be issued sometime in the fall of 1999.
                                 However, on the basis of our review of the proposed changes, it is
                                 uncertain whether the additional sanctions HCFA is considering will be
                                 strong enough to compel a state agency to improve its performance.

                                 Under its proposed guidelines, HCFA would add seven situations to its
                                 definition of inadequate state performance. These seven include situations
                                 in which a state agency has a pattern of failure to

                             •   conduct surveys within required time frames;
                             •   use proper enforcement actions against a nursing home;
                             •   respond to complaints in accordance with requirements;
                             •   enter nursing home deficiency data into HCFA’s database timely and
                             •   follow federal standards, protocols, forms, methods, procedures, policies,
                                 and systems specified in HCFA’s instructions;
                             •   ensure that nursing homes maintain specific resident information; and
                             •   enter nursing home resident assessment data into federal data systems.

                                 The proposed changes include a new remedy that would require HCFA to
                                 develop a plan of correction for the state to implement. In addition, two
                                 new sanctions are proposed to penalize a poorly performing state agency:
                                 (1) placing the state agency on notice that it is not in compliance with its
                                 Medicaid plan regarding nursing home survey performance and
                                 (2) requiring HCFA officials to meet with the state governor and other
                                 high-level state officials.

                                 Although HCFA refers to the proposed two new actions as sanctions, they
                                 are not as severe as what is normally thought of as sanctions and may not
                                 be enough to compel a state to improve its performance. When we
                                 discussed placing the state agency on notice, we were told by a HCFA
                                 official that, under this sanction, HCFA expects the regions to work
                                 collaboratively with state agencies to comply with the requirements in
                                 their state Medicaid plan. The proposed sanction requiring HCFA officials to

                                 Page 26                             GAO/HEHS-00-6 Federal Nursing Home Oversight

                  meet with the governor or other state officials can raise problems to a
                  higher level in state government and possibly secure greater state support
                  to improve performance. Nevertheless, it is not clear what impact either of
                  these sanctions would have on a state agency to effect the desired

                  HCFA  has recently demonstrated the desire and initiative to confront and
Conclusions       respond to various quality problems facing the nation’s nursing homes and
                  their residents. Some of the methods HCFA currently uses and is
                  developing, to ensure that state agencies develop effective survey
                  programs capable of identifying survey deficiencies in nursing homes, can
                  contribute useful information for assessing the overall effectiveness of its
                  many nursing home quality improvement initiatives. However, the limited
                  scope and rigor of its various state performance monitoring mechanisms,
                  and the uneven application of the mechanisms across the regions, do not
                  provide HCFA with a systematic, consistent means of assessing the
                  sufficiency of state survey performance. Additionally, HCFA’s approach to
                  conducting federal monitoring surveys is not adequate to establish a
                  pattern of inadequate state performance, which is needed before imposing
                  any sanction. Specifically,

              •   the negligible use of comparative surveys, combined with delays in
                  scheduling them to closely follow state surveys, does not provide HCFA
                  with sufficient evidence to establish whether states are appropriately
                  assessing nursing homes’ compliance with federal standards.
              •   systems development problems for reporting results of observational
                  surveys and delays in giving states written feedback have hindered HCFA’s
                  ability to effectively use survey results as a management tool.
              •   inconsistencies among HCFA regional offices in how they target their
                  federal monitoring surveys within each state and select resident samples
                  for comparative surveys further hamper HCFA’s ability to ensure that these
                  reviews effectively and equitably assess state survey performance.

                  Even if HCFA strengthens its oversight programs to be able to establish any
                  pattern of unacceptable state survey performance, it has yet to develop
                  effective alternatives for compelling the state to come into compliance
                  with nursing home survey standards, short of cutting off federal funds.
                  Being able to accurately and consistently assess state agency performance
                  and hold states accountable for meeting HCFA standards is essential to the
                  success of HCFA’s recent initiatives to improve the quality of care for the
                  nation’s nursing home residents.

                  Page 27                            GAO/HEHS-00-6 Federal Nursing Home Oversight

                      To ensure that states effectively meet federal standards for certifying
Recommendations       nursing homes and to consistently implement the more stringent
                      requirements of HCFA’s recent initiatives, HCFA needs to strengthen its
                      oversight of state survey agencies’ activities. To accomplish this, we
                      recommend that the Administrator of HCFA take the following actions:

                      1. Improve the scope and rigor of HCFA’s oversight process by

                  •   increasing the proportion of federal monitoring surveys conducted as
                      comparative surveys to ensure that a sufficient number are completed in
                      each state to assess whether the state appropriately identifies serious
                  •   ensuring that comparative surveys are initiated closer to the time the state
                      agency completes the home’s annual standard survey,
                  •   requiring regions to provide more timely written feedback to the states
                      after the completion of federal monitoring surveys, and
                  •   improving the data system for observational surveys so that it is an
                      effective management tool for HCFA to properly assess the findings of
                      observational surveys.

                      2. Improve the consistency in how HCFA holds state survey agencies
                      accountable by standardizing procedures for selecting state surveys and
                      conducting federal monitoring surveys, including

                  •   ensuring that the regions target surveys for review that will provide a
                      comprehensive assessment of state surveyor performance, and
                  •   requiring federal surveyors to include as many of the same residents as
                      possible in their comparative survey sample as the state included in its
                      sample (where HCFA surveyors have determined that the state sample
                      selection process was appropriate).

                      3. Further explore the feasibility of appropriate alternative remedies or
                      sanctions for those states that prove unable or unwilling to meet HCFA’s
                      performance standards.

                      In its comments on our draft report, HCFA noted that enhanced oversight of
Agency Comments       state programs is critical to improving the quality of care in nursing homes
                      and generally agreed with our recommendations.

                      HCFAcommented, and we agree, that a mix of comparative and
                      observational surveys is a prudent approach to overseeing state survey

                      Page 28                             GAO/HEHS-00-6 Federal Nursing Home Oversight

agencies’ performance. It did not agree to immediately increase the
number of comparative surveys it requires the regions to undertake but
stated that it is reviewing the issue of the appropriate balance between the
two types of monitoring surveys. While we agree that a mix is beneficial,
an adequate number of comparative surveys is critical to assess the
performance of state agencies. Furthermore, these surveys can help
improve the effectiveness of resources devoted to observational surveys
by focusing more of them on areas of poorer performance. We believe that
the results of the comparative surveys conducted in the past year indicate
that these surveys have been more effective than observational surveys in
identifying serious deficiencies affecting the health and safety of residents
that state agencies’ survey processes miss. While we agree that
observational surveys can serve as an effective training tool for state
surveyors, in our view, they do not provide an accurate representation of
typical state surveyor performance because of the likelihood that state
surveyors modify their performance when they are aware that they are
being observed.

HCFA  agreed that its comparative surveys should be initiated in closer
proximity to the state agency’s completion of the nursing home’s annual
standard survey. However, HCFA indicated that because it will focus its
comparative surveys on deficiency-free homes in the future, it does not
believe it would be realistic to start these surveys earlier than 2 weeks
after the state has completed its survey because the results of the state
survey are not often known for 2 weeks. But HCFA’s decision to focus its
comparative surveys on deficiency-free homes would exclude more than
two-thirds of homes nationwide from this review. In addition to
reconsidering the universe of nursing homes for which comparative
surveys are conducted, we believe that HCFA can initiate a comparative
survey more promptly, even if it means that HCFA does not know the
results of the state’s most recent survey before beginning its comparative
survey. HCFA can still achieve its monitoring objective by choosing its
sample on the basis of nursing homes’ prior survey history and could
improve results by having the two surveys more closely coincide. In fact,
not having the results could eliminate any bias associated with having
state survey results.

HCFA also agreed with the importance of ensuring national consistency in
overseeing state survey agencies and establishing definitive and
measurable performance standards to hold states accountable for survey
performance. However, HCFA indicates that it will establish additional
performance standards by the end of 1999. It will evaluate the use of these

Page 29                             GAO/HEHS-00-6 Federal Nursing Home Oversight

standards and its process for working with state agencies to improve their
performance for about 18 months before determining whether other
additional remedies or sanctions are needed. In our view, HCFA currently
does not have an adequate array of effective remedies or sanctions at its
disposal to ensure corrections, and it should more immediately consider
expanding the available remedies and sanctions that can be applied to
compel states to improve their performance.

HCFA’s    comments are included as appendix II.

As agreed with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days after its
issue date. At that time, we will send copies to the Honorable Nancy-Ann
Min DeParle, Administrator of HCFA; other interested congressional
committees; and other interested parties upon request.

Please contact me or Kathryn G. Allen, Associate Director, at
(202) 512-7114 if you or your staffs have questions about this report. Jack
Brennan and Mary Ann Curran prepared this report under the direction of
John Dicken.

William J. Scanlon
Director, Health Financing and
  Public Health Issues

Page 30                              GAO/HEHS-00-6 Federal Nursing Home Oversight
Page 31   GAO/HEHS-00-6 Federal Nursing Home Oversight

Letter                                                                                              1

Appendix I                                                                                         34

HCFA’s Definitions of
Severity of
Appendix II                                                                                        36

Comments From the
Health Care Financing
Related GAO Products                                                                               45

Tables                  Table 1: Nursing Homes, Required Federal Surveys, and Federal               6
                          Surveyors Available, by HCFA Region
                        Table 2: Minimum Number of Comparative Surveys Required                    10
                          Yearly, by State
                        Table 3: Time Frames for the Start of the 64 Comparative Surveys           12
                          HCFA Conducted Between October 1998 and July 1999
                        Table 4: SAQIP Nursing Home Performance Standards for State                17
                        Table 5: Percentage of Surveys in Which First Revisit Was Not              21
                          Made Within HCFA’s Recommended Time Frame
                        Table 6: Variation in Resources Available for and Time to                  22
                          Complete Observational Surveys
                        Table 7: Remedies and Sanctions Available to HCFA in Response              24
                          to Poor State Agency Performance
                        Table I.1: HCFA Scope and Severity Ratings for Nursing Home                34


                        HCFA       Health Care Financing Administration
                        OBRA 87    Omnibus Budget Reconciliation Act of 1987
                        SAQIP      State Agency Quality Improvement Program

                        Page 32                           GAO/HEHS-00-6 Federal Nursing Home Oversight
Page 33   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix I

HCFA’s Definitions of Severity of

                                        HCFA’s nursing home regulations established several categories of
                                        deficiencies that state survey agencies may find and record during their
                                        surveys. Each identified deficiency is placed into one of the 12 categories
                                        ranging from “A” to “L,” depending on the extent of resident harm
                                        (severity) and the number of residents adversely affected (scope). The
                                        following table identifies the scope and severity HCFA has assigned to
                                        Medicare and Medicaid compliance deficiencies.

Table I.1: HCFA Scope and Severity
Ratings for Nursing Home Deficiencies                                                Scope of deficiency
                                        Severity category            Isolated         Pattern              Widespread
                                        Actual or potential for      J                K                    L
                                        death/serious injury (also
                                        referred to as immediate
                                        Other actual harm            G                H                    I
                                        Potential for more than      D                E                    F
                                        minimal harm
                                        Potential for minimal harm   A                B                    C
                                        (substantial compliance)

                                        A home is considered to be in “substantial compliance” if any identified
                                        deficiencies are those in which the potential exists for only minimal harm
                                        to occur to residents (levels A, B, and C). Any nursing home with a
                                        deficiency categorized as D through L is considered to be not in
                                        compliance. “Potential for more than minimal harm” is a deficiency for
                                        which no actual harm has occurred to residents but that presents the
                                        potential for more than minimal harm to occur. “Other actual harm”
                                        includes deficiencies that cause actual harm to residents but do not
                                        immediately jeopardize their health or safety. “Immediate jeopardy,” the
                                        most serious deficiency, includes situations that immediately jeopardize
                                        the health or safety of residents.

                                        In addition to the four severity categories shown in the table, HCFA also
                                        uses a fifth deficiency category referred to as “substandard quality of
                                        care.” Deficiencies in this category are those that affect nursing home
                                        residents in the areas of resident behavior and facility practices, quality of
                                        life, and quality of care, and that are in the F, H, I, J, K, or L categories in
                                        the table.

                                        HCFA  also classifies deficiencies by their scope or extent as follows:
                                        (1) isolated, defined as affecting a limited number of residents; (2) pattern,

                                        Page 34                                 GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix I
HCFA’s Definitions of Severity of

defined as affecting more than a limited number of residents; and
(3) widespread, defined as affecting all or almost all residents.

Page 35                             GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II

Comments From the Health Care Financing

              Page 36     GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 37                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 38                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 39                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 40                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 41                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 42                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 43                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Appendix II
Comments From the Health Care Financing

Page 44                                   GAO/HEHS-00-6 Federal Nursing Home Oversight
Related GAO Products

              Nursing Home Oversight: Industry Examples Do Not Demonstrate That
              Regulatory Actions Were Unreasonable (GAO/HEHS-99-154R, Aug. 13, 1999).

              Nursing Homes: HCFA Initiatives to Improve Care Are Under Way but Will
              Require Continued Commitment (GAO/T-HEHS-99-155, June 30, 1999).

              Nursing Homes: Proposal to Enhance Oversight of Poorly Performing
              Homes Has Merit (GAO/HEHS-99-157, June 30, 1999).

              Nursing Homes: Complaint Investigation Processes in Maryland
              (GAO/T-HEHS-99-146, June 15, 1999).

              Nursing Homes: Complaint Investigation Processes Often Inadequate to
              Protect Residents (GAO/HEHS-99-80, Mar. 22, 1999).

              Nursing Homes: Additional Steps Needed to Strengthen Enforcement of
              Federal Quality Standards (GAO/HEHS-99-46, Mar. 18, 1999).

              California Nursing Homes: Care Problems Persist Despite Federal and
              State Oversight (GAO/HEHS-98-202, July 27, 1998).

(101779)      Page 45                           GAO/HEHS-00-6 Federal Nursing Home Oversight
Ordering Information

The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
following address, accompanied by a check or money order
made out to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.

Orders by mail:

U.S. General Accounting Office
P.O. Box 37050
Washington, DC 20013

or visit:

Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC

Orders may also be placed by calling (202) 512-6000
or by using fax number (202) 512-6061, or TDD (202) 512-2537.

Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information on
how to obtain these lists.

For information on how to access GAO reports on the INTERNET,
send an e-mail message with "info" in the body to:


or visit GAO’s World Wide Web Home Page at:


United States                       Bulk Rate
General Accounting Office      Postage & Fees Paid
Washington, D.C. 20548-0001           GAO
                                 Permit No. G100
Official Business
Penalty for Private Use $300

Address Correction Requested