This is the accessible text file for GAO report number GAO-10-70 entitled 'Nursing Homes: Addressing the Factors Underlying Understatement of Serious Care Problems Requires Sustained CMS and State Commitment' which was released on December 28, 2009. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to Congressional Requesters: United States Government Accountability Office: GAO: November 2009: Nursing Homes: Addressing the Factors Underlying Understatement of Serious Care Problems Requires Sustained CMS and State Commitment: GAO-10-70: GAO Highlights: Highlights of GAO-10-70, a report to congressional requesters. Why GAO Did This Study: Under contract with the CMS, states conduct surveys at nursing homes to help ensure compliance with federal quality standards. Over the past decade, GAO has reported on inconsistencies in states’ assessment of nursing homes’ quality of care, including understatement—that is, when state surveys fail to cite serious deficiencies or cite them at too low a level. In 2008, GAO reported that 9 states had high and 10 had low understatement based on CMS data for fiscal years 2002 through 2007. This report examines the effect on nursing home deficiency understatement of CMS’s survey process, workforce shortages and training, supervisory reviews of surveys, and state agency practices. GAO primarily collected data through two Web-based questionnaires sent to all eligible nursing home surveyors and state agency directors, achieving 61 and 98 percent response rates, respectively. What GAO Found: A substantial percentage of both state surveyors and directors identified general weaknesses in the nursing home survey process, that is, the survey methodology and guidance on identifying deficiencies. On the questionnaires, 46 percent of surveyors and 36 percent of directors reported that weaknesses in the traditional survey methodology, such as too many survey tasks, contributed to understatement. Limited experience with a new data-driven survey methodology indicated possible improvements in consistency; however, an independent evaluation led CMS to conclude that other tools, such as survey guidance clarification and surveyor training and supervision, would help improve survey accuracy. According to questionnaire responses, workforce shortages and greater use of surveyors with less than 2 years’ experience sometimes contributed to understatement. Nearly three-quarters of directors reported that they always or frequently experienced a workforce shortage, while nearly two-thirds reported that surveyor inexperience always, frequently, or sometimes led to understatement. Substantial percentages of directors and surveyors indicated that inadequate training may compromise survey accuracy and lead to understatement. According to about 29 percent of surveyors in 9 high understatement states compared to 16 percent of surveyors in 10 low understatement states, initial surveyor training was not sufficient to cite appropriate scope and severity—a skill critical in preventing understatement. Furthermore, over half of directors identified the need for ongoing training for experienced surveyors on both this skill and on documenting deficiencies, a critical skill to substantiate citations. CMS provides little guidance to states on supervisory review processes. In general, directors reported on our questionnaire that supervisory reviews occurred more often on surveys with higher-level rather than on those with lower-level deficiencies, which were the most frequently understated. Surveyors who reported that survey teams had too many new surveyors also reported frequent changes to or removal of deficiencies, indicating heavier reliance on supervisory reviews by states with inexperienced surveyors. Surveyors and directors in a few states informed us that, in isolated cases, state agency practices or external pressure from stakeholders, such as the nursing home industry, may have led to understatement. Forty percent of surveyors in five states and four directors reported that their state had at least one practice not to cite certain deficiencies. Additionally, over 40 percent of surveyors in four states reported that their states’ informal dispute resolution processes favored concerns of nursing home operators over resident welfare. Furthermore, directors from seven states reported that pressure from the industry or legislators may have compromised the nursing home survey process, and two directors reported that CMS’s support is needed to deal with such pressure. If surveyors perceive that certain deficiencies may not be consistently upheld or enforced, they may choose not to cite them. What GAO Recommends: GAO is making seven recommendations to the CMS Administrator to address state and surveyor issues about CMS’s survey methodology and guidance, workforce shortages and insufficient training, inconsistencies in the focus and frequency of the supervisory review of deficiencies, and external pressure from the nursing home industry. CMS concurred with five of GAO’s seven recommendations and indicated it would explore alternate solutions to the remaining two recommendations. View [hyperlink, http://www.gao.gov/products/GAO-10-70] or key components. For more information, contact John E. Dicken at (202) 512- 7114 or dickenj@gao.gov. Also see [hyperlink, http://www.gao.gov/products/GAO-10-74SP] for summary data from the questionnaires. [End of section] Contents: Letter: Background: Weaknesses in CMS Survey Process Contributed to Understatement, but Long-Term Effect of New Survey Methodology Is Not Yet Known: Workforce Shortages and Training Inadequacies May Contribute to Understatement: State Supervisory Reviews Often Are Not Designed to Identify Understatement: State Agency Practices and External Pressure May Compromise Survey Accuracy and Lead to Understatement in a Few States: Conclusions: Recommendations for Executive Action: Agency and AHFSA Comments and Our Evaluation: Appendix I: Scope and Methodology: Appendix II: Comments from the Department of Health & Human Services: Appendix III: GAO Contact and Staff Acknowledgments: Related GAO Products: Tables: Table 1: Scope and Severity of Deficiencies Identified during Nursing Home Surveys: Table 2: Surveyors' and State Agency Directors' Responses to Questions on CMS's Survey Process: Table 3: Percentage of Surveyors Reporting That Guidance for Certain Federal Quality Standards Was Not Sufficient to Identify Deficiencies: Table 4: State Agency Directors' Responses to Questions about Surveyor Workforce Issues: Table 5: State Survey Agency Vacancy Rates and Percentage of State Surveyors with Less Than 2 Years' Experience: Table 6: Surveyors' and State Agency Directors' Responses to Questions on Workforce Issues: Table 7: Responses from Surveyors and State Agency Directors to Key Questions on Training: Table 8: Percentage of Surveyors Reporting Changes in Deficiency Citations during Supervisory Review: Table 9: Response Rates to GAO's Questionnaire of Nursing Home Surveyors, 2008: Figures: Figure 1: Zero-, Low-, and High-Understatement States, Fiscal Years 2002-2007: Figure 2: Eight State Agency Director Responses on Five Questions Related to the QIS: Figure 3: Number of State Supervisory Reviews at the Potential for More than Minimal Harm (D-F) and Immediate Jeopardy Levels (J-L): Figure 4: Percentage of Surveyors in Each State Reporting at Least One Noncitation Practice: Figure 5: Percentage of Surveyors in Each State Reporting the IDR Process Favored Concerns of Nursing Home Operators over Resident Welfare: Abbreviations: AHFSA: Association of Health Facility Survey Agencies: CMS: Centers for Medicare & Medicaid Services: HHS: Department of Health & Human Services: IDR: Informal Dispute Resolution: OSCAR: On-Line Survey, Certification, and Reporting system: RN: registered nurse: SMQT: Surveyor Minimum Qualifications Test: SOM: State Operations Manual: QIS: Quality Indicator Survey: [End of section] United States Government Accountability Office: Washington, DC 20548: November 24, 2009: The Honorable Herb Kohl: Chairman: Special Committee on Aging: United States Senate: The Honorable Charles E. Grassley: Ranking Member: Committee on Finance: United States Senate: Federal and state governments share responsibility for ensuring that nursing homes provide quality care in a safe environment for the nation's 1.5 million residents dependent on such care. The federal government is responsible for setting quality requirements that nursing homes must meet to participate in the Medicare and Medicaid programs. [Footnote 1] The Centers for Medicare & Medicaid Services (CMS), within the Department of Health & Human Services (HHS), contracts with state survey agencies to conduct periodic inspections, known as surveys, and complaint investigations, both of which assess whether homes meet federal standards.[Footnote 2] State survey agencies are required to follow federal regulations for surveying facilities; however, several survey activities and policies are left largely to the discretion of state survey agencies, including hiring and retaining a surveyor workforce, training surveyors, reviewing deficiency citations, and managing regulatory interactions with the industry and public. In response to congressional requests over the last decade, we have reported significant weaknesses in federal and state activities designed to detect and correct quality and safety problems in nursing homes and the persistence of serious deficiencies, which are those deficiencies that harm residents or place them at risk of death or serious injury.[Footnote 3] In the course of our work, we regularly found significant variation across states in their citations of serious deficiencies--indicating inconsistencies in states' assessment of quality of care. We also found evidence of substantial understatement-- that is, state inspections that failed to cite serious deficiencies or that cited deficiencies at too low a level. In this report, we complete our response to your request to examine the understatement of serious deficiencies in nursing homes by state surveyors nationwide and the factors that contribute to understatement. Our first report, issued in May 2008, identified the extent of nursing home understatement nationwide.[Footnote 4] It found that 15 percent of federal nursing home surveys nationwide and 25 percent of these surveys in nine states identified state surveys that failed to cite serious deficiencies. This report examines how the following factors affect the understatement of nursing home deficiencies: (1) the CMS survey process, (2) workforce shortages and training, (3) supervisory reviews, and (4) state agency practices. To do this work, we analyzed data collected from two GAO-administered Web-based questionnaires, one to nursing home surveyors and the other to state agency directors; analyzed federal and state nursing home survey results; interviewed CMS officials from the Survey and Certification Group and selected Regional Offices; reviewed federal regulations and guidance, and our prior work; and conducted follow-up interviews with state agency directors, as needed, to clarify and better understand their unique state circumstances.[Footnote 5] Our prior work documented the prevalence of understatement nationwide and described several factors that may contribute to survey inconsistency and the understatement of deficiencies by state survey teams: (1) weaknesses in CMS's survey methodology, including poor documentation of deficiencies,[Footnote 6] (2) confusion among surveyors about the definition of actual harm,[Footnote 7] (3) predictability of surveys, which allows homes to conceal problems if they so desire,[Footnote 8] (4) inadequate quality assurance processes at the state level to help detect understatement in the scope and severity of deficiencies,[Footnote 9] and (5) inexperienced state surveyors as a result of retention problems.[Footnote 10] We relied on this information and feedback from pretests with six surveyors from a local state and five current or former state agency directors to develop our questionnaires on the nursing home survey process and factors that contribute to the understatement of deficiencies. Our Web-based questionnaires of nursing home surveyors and state agency directors achieved response rates of 61 percent and 98 percent, respectively. The first questionnaire collected responses from 2,340 of the total 3,819 eligible nursing home surveyors in 49 states and the District of Columbia.[Footnote 11] The resulting sample of surveyors who responded to our questionnaire between May and July 2008 was representative of surveyors nationally, with the exception of Pennsylvania.[Footnote 12] Fifty state agency directors responded to the second questionnaire from September to November 2008.[Footnote 13] Many questions on our questionnaires asked respondents to identify the frequency that an event occurred using the following scale--always, frequently, sometimes, infrequently, or never; however, for reporting purposes, we grouped responses into three categories--always/ frequently, sometimes, and infrequently/never. In addition, our questionnaire to state agency directors asked them to rank the degree to which several factors, derived from our previous work, contributed to understatement.[Footnote 14] Summary results from the GAO questionnaires are available as an e-supplement to this report. See Nursing Homes: Responses from Two Web-Based Questionnaires to Nursing Home Surveyors and State Agency Directors (GAO-10-74SP), an e- supplement to GAO-10-70. We analyzed the data collected from these questionnaires as stand-alone datasets and in relationship to state performance on federal comparative and observational surveys as captured in the federal monitoring survey database, which we reported on in 2008.[Footnote 15] In addition, to inform our understanding of the extent to which each factor contributed to understatement, we examined relationships among the responses to both questionnaires and the results of the federal comparative and observational surveys for fiscal years 2002 through 2007. We used the results of the federal comparative surveys for these years to identify states with high and low percentages of serious missed deficiencies. We report results for tests of association and differences between group averages. We also interviewed directors and other state agency officials in eight states to better understand unusual or interesting circumstances related to surveyor workforce and training, supervisory review, or state policies and practices. We selected these eight state agencies based on our analysis of questionnaire responses from the state agency directors and nursing home surveyors. To compare average facility citations on state survey records with the average citations on federal survey records, we collected information from the On-Line Survey, Certification, and Reporting (OSCAR) system for those facilities where federal teams assessed state surveyor performance for fiscal years 2002 through 2007.[Footnote 16] Except where otherwise noted, we used data from fiscal year 2007 because they were the most recently available data at the time of our analysis (see appendix I for more on our scope and methodology). We conducted this performance audit from April 2008 through November 2009 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. Background: Oversight of nursing homes is a shared federal-state responsibility. As part of this responsibility, CMS (1) sets federal quality standards, (2) establishes state responsibilities for ensuring federal quality standards are met, (3) issues guidance on determining compliance with these standards, and (4) performs oversight of state survey activities. It communicates these federal standards and state responsibilities in the State Operations Manual (SOM) and through special communications such as program memorandums and survey and certification letters. CMS provides less guidance on how states should manage the administration of their survey programs. CMS uses staff in its 10 regional offices to oversee states' performance on surveys that ensure that facilities participating in Medicare and Medicaid provide high-quality care in a safe environment. Yet, the persistent understatement of serious nursing home deficiencies that we have reported and survey quality weaknesses that we and the HHS Office of Inspector General identified serve as indicators of weaknesses in the federal, state, or shared components of oversight. Survey Process: Every nursing home receiving Medicare or Medicaid payment must undergo a standard state survey not less than once every 15 months, and the statewide average interval for these surveys must not exceed 12 months. During a standard survey, teams of state surveyors--generally consisting of registered nurses, social workers, dieticians, or other specialists--evaluate compliance with federal quality standards. The survey team determines whether the care and services provided meet the assessed needs of the residents and measure resident outcomes, such as the incidence of preventable pressure sores, weight loss, and accidents. In contrast to a standard survey, a complaint investigation generally focuses on a specific allegation regarding a resident's care or safety and provides an opportunity for state surveyors to intervene promptly if problems arise between standard surveys. Surveyors assess facilities using federal nursing home quality standards that focus on the delivery of care, resident outcomes, and facility conditions. These standards total approximately 200 and are grouped into 15 categories, such as Quality of Life, Resident Assessment, Quality of Care, and Administration.[Footnote 17] For example, there are 23 standards (known as F-tags) within the Quality of Care category ranging from the prevention of pressure sore development (F-314) to keeping the resident environment as free of accident hazards (F-323) as is possible. Surveyors categorize deficient practices identified on standard surveys and complaint investigations--facilities' failures to meet federal standards--according to scope (i.e., the number of residents potentially or actually affected) and severity (i.e., the degree of relative harm involved)--using a scope and severity grid (see table 1). Homes with deficiencies at the A through C levels are considered to be in substantial compliance, while those with deficiencies at the D through L levels are considered out of compliance. Throughout this report, we refer to deficiencies at the actual harm and immediate jeopardy levels--G through L--as serious deficiencies. CMS guidance requires state survey teams to revisit a home to verify that serious deficiencies have actually been corrected.[Footnote 18] Table 1: Scope and Severity of Deficiencies Identified during Nursing Home Surveys: Severity: Immediate jeopardy[A]; Scope: Isolated: J; Scope: Pattern: K; Scope: Widespread: L. Severity: Actual harm; Scope: Isolated: G; Scope: Pattern: H; Scope: Widespread: I. Severity: Potential for more than minimal harm; Scope: Isolated: D; Scope: Pattern: E; Scope: Widespread: F. Severity: Potential for minimal harm[B]; Scope: Isolated: A; Scope: Pattern: B; Scope: Widespread: C. Source: CMS. [A] Actual or potential for death/serious injury. [B] Nursing home is considered to be in "substantial compliance." [End of table] In addition, when serious deficiencies are identified, sanctions can be imposed to encourage facilities to correct the deficiencies and enforce federal quality standards. Sanctions include fines known as civil money penalties, denial of payment for new Medicare or Medicaid admissions, or termination from the Medicare and Medicaid programs. For example, facilities that receive at least one G through L level deficiency on successive standard surveys or complaint investigations must be referred for immediate sanctions. Facilities may appeal cited deficiencies and if the appeal is successful, the severity of the sanction could be reduced or the sanction could be rescinded. Facilities have several avenues of appeal, including informal dispute resolution (IDR) at the state survey agency level.[Footnote 19] The IDR gives providers one opportunity to informally refute cited deficiencies after any survey. While CMS requires that states have an IDR policy in place, it does not specify how IDR processes should be structured. Survey Methodology: To conduct nursing home surveys, CMS has traditionally used a methodology that requires surveyors to select a sample of residents and (1) review data derived from the residents' assessments and medical records; (2) interview nursing home staff, residents, and family members; and (3) observe care provided to residents during the course of the survey. When conducting a survey, surveyors have discretion in: selecting a sample of residents to evaluate; allocating survey time and emphasis within a framework prescribed by CMS; investigating potentially deficient practices observed during the survey; and determining what evidence is needed to identify a deficient practice. CMS has developed detailed investigative protocols to assist state survey agencies in determining whether nursing homes are in compliance with federal quality standards. These protocols are intended to ensure the thoroughness and consistency of state surveys and complaint investigations. In 1998, CMS awarded a contract to revise the survey methodology. The new Quality Indicator Survey (QIS) was developed to improve the consistency and efficiency of state surveys and provide a more reliable assessment of quality. The QIS uses an expanded sample of residents and structured interviews with residents and family members in a two-stage process. Surveyors are guided through the QIS process using customized software on tablet personal computers. In stage 1, a large resident sample is drawn and relevant data from on-and off-site sources is analyzed to develop a set of quality-of-care indicators, which will be compared to national benchmarks.[Footnote 20] Stage 2 systematically investigates potential quality-of-care concerns identified in stage 1. Because of delays in implementing the QIS, we recommended in 2003 that CMS finalize the development, testing, and implementation of a more rigorous survey methodology, including investigative protocols that provide guidance to surveyors in documenting deficiencies at the appropriate scope and severity level.[Footnote 21] CMS concluded a five- state demonstration process of the QIS in 2007 and is currently expanding the implementation of the QIS. As of 2008, only Connecticut had implemented the QIS statewide, and CMS projected that the QIS would not be fully implemented in every state until 2014. State Administration: States are largely responsible for the administration of the survey program. State survey agencies administer and have discretion over many survey activities and policies, including hiring and retaining a surveyor workforce, training surveyors, conducting supervisory reviews of surveys, and other activities. * Hiring and Retaining a Surveyor Workforce: State survey agencies hire the staff to conduct surveys of nursing homes and determine the salaries of these personnel according to the workforce practices and restrictions of the state. Salaries, particularly surveyor salaries, are the most significant cost component of state survey activities, which are supported through a combination of Medicare, Medicaid, and non-Medicaid state funds.[Footnote 22] CMS has some requirements for the make-up of nursing home survey teams, including the involvement of at least one registered nurse (RN) in each nursing home survey. In February 2009, we reported that officials from the Association of Health Facility Survey Agencies (AHFSA) and other state officials told us they have had difficulty recruiting and retaining the survey workforce for several years. In our report, we recommended that CMS undertake a broad-based reexamination to ensure, among other aspects, an adequate survey workforce with sufficient compensation to attract and retain qualified staff.[Footnote 23] * Training: States are responsible for training new surveyors through participating in actual surveys under direct supervision. Within their first year of employment, surveyors must complete two CMS online training courses--the Basic Health Facility Surveyor Course and Principles of Documentation--and a week-long CMS-led Basic Long-Term Care Health Facility Surveyor Training Course; at the conclusion of the course surveyors must pass the Surveyor Minimum Qualifications Test (SMQT) to survey independently. In addition, state survey agencies are required to have their own programs for staff development that respond to the need for continuing development and education of both new and experienced employees. Such staff development programs must include training for surveyors on all regulatory requirements and the skills necessary to conduct surveys. To assist in continuing education, CMS develops a limited number of courses for ongoing training and provides other training materials. * Supervisory Reviews: States may design a supervisory review process for deficiencies cited during surveys, although CMS does not require them to do so. In July 2003, we recommended that CMS require states to have a minimum quality-assurance process that includes a review of a sample of survey reports below the level of actual harm to assess the appropriateness of scope and severity levels cited and help reduce instances of understated quality-of-care problems.[Footnote 24] CMS did not implement this recommendation.[Footnote 25] * State Agency Practices and Policies: State survey agencies' practices, including those on citing deficiencies and addressing pressure from the industry or others, are largely left to the discretion of state survey agencies. In the past, we reported that in one state, CMS officials had found surveyors were not citing all deficiencies.[Footnote 26] If a state agency fails to cite all deficiencies associated with noncompliance, nursing home deficiencies are understated on the survey record. CMS can identify or monitor states for systematic noncitation practices through reviews of citation patterns, informal feedback from state surveyors, state performance reviews, and federal monitoring surveys (discussed below).[Footnote 27] CMS also gives states latitude in defining their IDR process. Federal Monitoring Surveys and Evidence of Understatement: Federal law requires federal surveyors to conduct federal monitoring surveys in at least 5 percent of state-surveyed Medicare and Medicaid nursing homes in each state each year. CMS indicates it meets the statutory requirement by conducting a mix of on-site reviews: comparative and observational surveys.[Footnote 28] Comparative surveys. A federal survey team conducts an independent survey of a home recently surveyed by a state survey agency in order to compare and contrast its findings with those of the state survey team. This comparison takes place after completion of the federal survey. When federal surveyors identify a deficiency not cited by state surveyors, they assess whether the deficiency existed at the time of the state survey and should have been cited.[Footnote 29] This assessment is critical in determining whether understatement occurred, because some deficiencies cited by federal surveyors may not have existed at the time of the state survey. Our May 2008 report stated that comparative surveys found problems at the most serious levels of noncompliance--the actual harm and immediate jeopardy levels (G through L).[Footnote 30] About 15 percent of federal comparative surveys nationwide identified at least one deficiency at the G through L level that state surveyors failed to cite. While this proportion is small, CMS maintains that any missed serious deficiencies are unacceptable. Further, state surveys with understated deficiencies may allow the surveyed facilities to escape sanctions intended to discourage repeated noncompliance. In our May 2008 report we found that for nine states federal surveyors identified missed serious deficiencies in 25 percent or more comparative surveys for fiscal years 2002 through 2007; we defined these states as high-understatement states (see figure 1). Zero- understatement states were states that had no federal comparative surveys identifying missed deficiencies at the actual harm or immediate jeopardy levels; and low-understatement states were the 10 states with the lowest percentage of missed serious deficiencies (less than 6 percent), including all 7 zero-understatement states. Figure 1: Zero-, Low-, and High-Understatement States, Fiscal Years 2002-2007: [Refer to PDF for image: map of the United States] Zero: Alaska: Idaho: Maine: North Dakota: Oregon: Vermont: West Virginia: Low: Arkansas: Nebraska: Ohio: Mid-range: California: Colorado: Connecticut: Delaware: District of Columbia: Florida: Georgia: Hawaii: Illinois: Indiana: Iowa: Kansas: Kentucky: Louisiana: Maryland: Massachusetts: Michigan: Minnesota: Mississippi: Montana: Nevada: New Hampshire: New Jersey: New York: North Carolina: Pennsylvania: Rhode Island: Texas: Utah: Virginia: Washington: Wisconsin: High: Alabama: Arizona: Missouri: New Mexico: Oklahoma: South Carolina: South Dakota: Tennessee: Wyoming: Source: GAO analysis of CMS data. Map: Copyright © Corel Corp. All rights reserved. Note: Zero-understatement states were those that had no missed serious deficiencies on federal comparative surveys. Low-understatement states were the 10 states with the lowest percentage of missed serious deficiencies on federal comparative surveys (less than 6 percent), including all zero-understatement states. High-understatement states were the 9 states with the highest percentage of serious missed deficiencies (25 percent or more) on federal comparative surveys. [End of figure] Our May 2008 report also found that missed deficiencies at the potential for more than minimal harm level (D through F) were considerably more widespread than those at the G through L level on comparative surveys, with approximately 70 percent of comparative surveys nationwide identifying at least one missed deficiency at this level. Undetected care problems at this level are of concern because they could become more serious over time if nursing homes are not required to take corrective actions.[Footnote 31] Observational surveys. Federal surveyors accompany a state survey team to evaluate the team's performance and ability to document survey deficiencies. State teams are evaluated in six areas, including two-- General Investigation and Deficiency Determination--that affect the appropriate identification and citation of deficiencies. The General Investigation segment assesses the effectiveness of state survey team actions such as collection of information, discussion of survey observations, interviews with nursing home residents, and implementation of CMS investigative protocols. The Deficiency Determination segment evaluates the skill with which the state survey teams (1) analyze and integrate all information collected, (2) use the guidance for surveyors, and (3) assess compliance with regulatory requirements. Federal observational surveys are not independent evaluations of the state survey because state surveyors may perform their survey tasks more attentively than they would if federal surveyors were not present; however, they provide more immediate feedback to state surveyors and may help identify state surveyor training needs. We previously reported that state survey teams' poor performance on federal observational surveys in the areas of General Investigation and Deficiency Determination may contribute to the understatement of deficiencies.[Footnote 32] Further, poor state performance in these two areas supported the finding of understatement as identified through the federal comparative surveys. We found that about 8 percent of state survey teams observed by federal surveyors nationwide received below- satisfactory ratings on General Investigation and Deficiency Determination from fiscal years 2002 through 2007. However, surveyors in high-understatement states performed worse in these two areas of the federal observational surveys than surveyors in the low-understatement states. For example, an average of 12 and 17 percent of state survey teams observed by federal surveyors in high-understatement states received below satisfactory ratings for these two areas, respectively. In contrast, an average of 4 percent of survey teams in low- understatement states received the same below-satisfactory scores for both deficiency determination and investigative skills. Nationwide, one-third of nursing homes had a greater average number of serious deficiencies on federal observational surveys than on state standard surveys during fiscal years 2002 through 2007, but in eight states, it was more than half of homes. Of the one-third of homes nationwide, state standard surveys cited 83 percent fewer serious deficiencies than federal surveys during this same time period. Weaknesses in CMS Survey Process Contributed to Understatement, but Long-Term Effect of New Survey Methodology Is Not Yet Known: Over a third of both surveyors and state agency directors responding to our questionnaire identified weaknesses in the federal government's nursing home survey process that contributed to the understatement of deficiencies.[Footnote 33] The weaknesses included problems with the current survey methodology; written guidance that is too long or complex; and to a lesser extent, survey predictability or other advance notice of inspections, which may allow nursing homes to conceal deficiencies. At the time our questionnaires were fielded, eight states had started implementing CMS's new survey methodology. The limited experience among these states suggests that the new methodology may improve consistency of surveys, but information is limited, and the long-term ability of the new methodology to reduce understatement is not yet known. Weaknesses in CMS's Survey Process Contributed to Understatement: Both surveyors and state agency directors reported weaknesses in the survey process, and on our questionnaire linked these weaknesses to understatement of deficiencies. Nationally, 46 percent of nursing home surveyors responded that weaknesses in the current survey methodology resulted in missed or incorrectly identified deficiencies, with this number ranging by state from 0 to 74 percent (see table 2).[Footnote 34] Thirty-six percent of state agency directors responded that weaknesses in the current survey methodology at least sometimes contributed to understatement of deficiencies in their states. One such weakness identified by both surveyors and directors was the number of survey tasks that need to be completed. Table 2: Surveyors' and State Agency Directors' Responses to Questions on CMS's Survey Process: Questions related to CMS's survey process: Weaknesses in the current survey methodology at least sometimes result in missed or incorrectly identified deficiencies at the facility; Percentage of surveyors' responses: 46%; Percentage of directors' responses: 36%. Questions related to CMS's survey process: Additional training is needed to apply CMS guidance; Percentage of surveyors' responses: 40%; Percentage of directors' responses: 58%. Source: GAO. [End of table] [Sidebar: Surveyor Quotation about CMS Written Guidance: “Appreciate the guidances and protocols. However, making Appendix PP [guidance for investigating federal quality standards] into a tome is not helping us out in the field. They are too cumbersome and voluminous. Please find a way to be more concise in these guidances.” End of sidebar] According to surveyors and agency directors responding to our questionnaire, another weakness with the federal survey process involved CMS's written guidance to help state agencies follow federal regulations for surveying long-term care facilities.[Footnote 35] Both surveyors and state agency directors mentioned concerns about the length, complexity, and subjectivity of the written guidance. One state agency director we interviewed told us that the size of the SOM made it difficult for surveyors to carry the guidance and consult it during surveys. Although the SOM is available in an electronic format, surveyors in this state did not use laptops. In addition, a small percentage of surveyors commented on our questionnaire that CMS guidance was inconsistently applied in the field. A common complaint from these surveyors was that different supervisors required different levels of evidence in order to cite a deficiency at the actual harm or immediate jeopardy level. Forty percent of surveyors and 58 percent of state agency directors reported that additional training on how to apply CMS guidance was needed. A specific concern raised about the current survey guidance was determining the severity level for an observed deficiency. Forty-four percent of state agency directors reported on our questionnaire that confusion about CMS's definition of the actual-harm level severity requirements at least sometimes contributed to understatement in their states. CMS's guidance for determining actual harm states, "this does not include a deficient practice that only could or has caused limited consequence to the resident."[Footnote 36] State agency directors from several states found this language confusing, including one director who said it is unclear whether conditions like dehydration that are reversed in the hospital should be cited as actual harm. As we reported in 2003, CMS officials acknowledged that the language linking actual harm to practices that have "limited consequences" for a resident has created confusion; however, the agency has not changed or revised this language.[Footnote 37] State agency directors and surveyors indicated that CMS's written guidance for certain federal nursing home quality standards could be improved and that revised investigative protocols were helpful. [Footnote 38] Specifically, 11 state agency directors reported that CMS guidance on quality standards related to abuse could be improved. State agency directors commented that the guidance for certain quality standards was too long, with the guidance for two standards being over 50 pages long. One state agency director also noted that overly complex guidance will lead to an unmanageable survey process. Surveyors' concerns about the sufficiency of CMS's guidance varied for different quality standards (see table 3). For instance, 21 percent of surveyors nationwide reported that CMS guidance on pain management was not sufficient to identify deficiencies, whereas only 5 percent reported that guidance on pressure ulcers was not sufficient. Our analysis found that fewer surveyors had concerns with the guidance on quality standards revised through CMS's guidance update initiative.[Footnote 39] For example, the guidance on pressure ulcers was revised in 2004 and the guidance on accidents was revised in 2007; these topics ranked last among the areas of concern.[Footnote 40] Furthermore, state agency directors from several states commented on the usefulness of CMS's revised investigative protocols for federal quality standards. Table 3: Percentage of Surveyors Reporting That Guidance for Certain Federal Quality Standards Was Not Sufficient to Identify Deficiencies: Federal quality standard (number): Pain Management (multiple F-tags)[A, B]; Percentage reporting guidance on quality standard was not sufficient: 21. Federal quality standard (number): Quality of Care/Provide Necessary Care and Services for Highest Practicable Well-Being (F-309)[C]; Percentage reporting guidance on quality standard was not sufficient: 20. Federal quality standard (number): Range of Motion Mobility Treatment (F-318); Percentage reporting guidance on quality standard was not sufficient: 14. Federal quality standard (number): Accuracy of Resident Assessment (F- 278); Percentage reporting guidance on quality standard was not sufficient: 13. Federal quality standard (number): Comprehensive Care Plans (F-279); Percentage reporting guidance on quality standard was not sufficient: 12. Federal quality standard (number): Sanitary Conditions for Food (F- 371)[B]; Percentage reporting guidance on quality standard was not sufficient: 12. Federal quality standard (number): Abuse (F-223 through F-226)[D]; Percentage reporting guidance on quality standard was not sufficient: 11. Federal quality standard (number): Maintains Body Weight (F-325)[B]; Percentage reporting guidance on quality standard was not sufficient: 11. Federal quality standard (number): Physical Restraints (F-221); Percentage reporting guidance on quality standard was not sufficient: 11. Federal quality standard (number): Unnecessary Drugs (F-329)[E]; Percentage reporting guidance on quality standard was not sufficient: 11. Federal quality standard (number): Resident Participation in Planning Care and Treatment (F-280); Percentage reporting guidance on quality standard was not sufficient: 10. Federal quality standard (number): Accidents (F-323)[E]; Percentage reporting guidance on quality standard was not sufficient: 8. Federal quality standard (number): Pressure Ulcers (F-314)[E]; Percentage reporting guidance on quality standard was not sufficient: 5. Source: GAO. [A] CMS consolidated guidance on pain management into F-309 on March 31, 2009. [B] CMS revised guidance after our questionnaire of nursing home surveyors was administered in May 2008. [C] CMS added guidance to F-309 for residents receiving hospice or dialysis services on April 10, 2009. [D] CMS plans to begin revising guidance in Fall 2009. [E] CMS revised guidance before our questionnaire of nursing home surveyors was administered in May 2008. [End of table] Another weakness associated with the federal survey process was the potential for surveys to be predictable based solely on their timing. [Footnote 41] Eighteen percent of state agency directors reported that survey predictability or other advance notice of inspections at least sometimes contributed to understatement in their states. We analyzed state agencies' most-recent nursing home surveys and found that 29 percent of these surveys could be considered predictable due to their timing. We previously reported that survey predictability could contribute to understatement because it gives nursing homes the opportunity to conceal deficiencies if they choose to do so.[Footnote 42] CMS officials previously stated that reducing survey predictability could require increased funding because more surveys would need to be conducted within 9 months of the previous survey.[Footnote 43] However, CMS noted that state agencies are not funded to conduct any surveys within 9 months of the last standard survey. New Survey Methodology's Effect on Understatement Inconclusive: There was no consensus among the eight state agency directors who had started implementing the QIS as of November 2008 about how the new survey methodology would affect understatement.[Footnote 44] Three directors reported that the QIS was likely to reduce understatement; three directors reported that it was not likely to reduce understatement; and two directors were unsure or had no opinion (see figure 2). However, all eight directors reported that the new QIS methodology was likely to improve survey consistency both within and across states. In addition, five of these directors reported that the new QIS methodology was likely to improve survey quality. Five of the eight directors also indicated that the QIS required more time than the traditional survey methodology. Figure 2: Eight State Agency Director Responses on Five Questions Related to the QIS: [Refer to PDF for image: stacked vertical bar graph] Question: Improve consistency within this state; Yes: 8; No: 0; Not sure/no opinion: 0. Question: Improve consistency across states; Yes: 8; No: 0; Not sure/no opinion: 0. Question: Improve quality of nursing home surveys; Yes: 5; No: 1; Not sure/no opinion: 2. Question: Require more time to complete; Yes: 5; No: 3; Not sure/no opinion: 0. Question: Reduce understatement; Yes: 3; No: 3; Not sure/no opinion: 2. Source: GAO. [End of figure] CMS funded an independent evaluation of the QIS, which was completed by a contractor in December 2007.[Footnote 45] The evaluation assessed the effectiveness of the new methodology by studying (1) its effect on accuracy of surveys, (2) documentation of deficiencies, (3) time required to complete survey activities, (4) number of deficiencies cited, and (5) surveyor efficiency. The evaluation did not draw a firm conclusion about the overall effectiveness of the QIS as measured through these five areas. For instance, the QIS methodology was associated with an increase in the total number of deficiencies cited, including an increase in the number of G-level deficiencies and the number of quality standard areas cited. However, the evaluation did not find that the QIS methodology increased survey accuracy, noting that QIS and traditional survey samples were comparable in overall quality and in the frequency of standards cited for deficiencies with either a pattern or widespread scope.[Footnote 46] The results suggested that more deficiencies with higher scope could have been cited for both the QIS and traditional surveys. Similarly, there was no evidence that the QIS resulted in higher-quality documentation or improved surveyor efficiency. Although five state agency directors reported that the QIS required more time to complete than the traditional methodology, the evaluation found some evidence of a learning curve, suggesting that surveyors were able to complete surveys faster as they became familiar with the new process. The evaluation generated a number of recommendations for improving the QIS that are consistent with reducing understatement, such as improving the specificity and usability of investigative protocols and evaluating how well the new methodology accurately identifies the areas in which there are potential quality problems. Since the evaluation did not find improved accuracy, CMS concluded that non-QIS factors, including survey guidance clarification and surveyor training and supervision, would help improve survey accuracy. Additionally, CMS concluded that future QIS development efforts should concentrate on improving survey consistency and giving supervisors more tools to assess the performance of surveyor teams. Ten state agency directors that had not yet started implementing the QIS responded to our questionnaire with concerns about the cost associated with implementing the new methodology, including the resources required to train staff and obtain new equipment.[Footnote 47] Of these 10 directors, 3 also expressed concerns that allotting staff time for QIS implementation would prevent the agency from completing mandatory survey activities. Workforce Shortages and Training Inadequacies May Contribute to Understatement: Workforce shortages and training inadequacies affected states' ability to complete thorough surveys, contributing to understatement of nursing home deficiencies. Responses to our questionnaires indicated that states experienced workforce shortages or were attempting to accomplish their workload with a high percentage of inexperienced surveyors. In states with fewer staff to do the work, time frames were compressed. The increased workload burden may have had an effect on the thoroughness of surveys in those states and surveyors' ability to attend training. The frequent hiring of new surveyors to address workforce shortages also burdened states' surveyor training programs. Surveyors, state agency directors, and state performance on federal observational surveys indicated that inadequacies in initial and ongoing training may have compromised survey accuracy in high- understatement states. Workforce Shortages Sometimes Contributed to Understatement: Although a small percentage of state agency directors reported that workforce shortages always or frequently contributed to the understatement of nursing home deficiencies in their states, 36 percent indicated that workforce shortages sometimes contributed to understatement (see table 4). In many states, workforce shortages resulted in a greater reliance on inexperienced surveyors. According to state agency directors and surveyors, this collateral effect-- inexperienced surveyors--also may have contributed to understatement. States also expressed concern about completing their workload, which appeared to be, in part, an outgrowth of workforce shortages and use of inexperienced surveyors. Table 4: State Agency Directors' Responses to Questions about Surveyor Workforce Issues: How frequently do the following issues contribute to understatement in this state survey agency? Percentage of state agency directors' responses: Inadequate number of staff to complete thorough surveys: Always/frequently: 6; Sometimes: 36; Infrequently/never: 58. Inadequate time to complete thorough surveys: Always/frequently: 8; Sometimes: 38; Infrequently/never: 54. Reluctance to cite serious deficiencies because of workload burden: Always/frequently: 8; Sometimes: 10; Infrequently/never: 86. Inexperienced surveyors not yet comfortable with job responsibilities: Always/frequently: 16; Sometimes: 48; Infrequently/never: 34. Source: GAO. [End of table] Workforce Shortages. Since 2003, we have reported that states have experienced pervasive workforce shortages, and responses to our questionnaires indicate that shortages continue to affect states. [Footnote 48] Seventy-two percent of state agency directors reported that they always or frequently had a surveyor workforce shortage, and another 16 said it occurred sometimes. The average vacancy rate for surveyors was 14 percent, and one-fourth of states had a vacancy rate of higher than 19 percent (see table 5).[Footnote 49] Among the 49 reporting states, the vacancy rate ranged from a maximum of 72 percent in Alabama to 0 percent in Nevada, Rhode Island, Vermont, and Utah. The workforce shortages have stemmed mostly from the preference to employ RNs as surveyors in state survey agencies, with half of reporting states employing RNs as more than 75 percent of their surveyor workforce.[Footnote 50] In the past, states have claimed that they had difficulty matching RN salaries offered by the private sector, and this hampered the hiring and retention of RNs. The Virginia state agency director commented during an interview that the nursing home industry values individuals who have passed CMS's SMQT and hires its surveyors after they are trained and certified by CMS. Virginia and others also identified the stress of the job--regular travel, time pressures to complete the workload, and the regulatory environment--as a challenge to retaining staff. Previously, we reported that workforce instability arising from noncompetitive RN surveyor salaries and hiring freezes affected states' abilities to complete their survey workload or resulted in the hiring of less-qualified staff.[Footnote 51] Most recently, the poor economy has further constrained state budgets for surveyors. For example, to address its budget shortfall in 2009, California will furlough its state employees including surveyors for 2 days every month from February 2009 through June 2010.[Footnote 52] An additional 11 states also reported furloughs for 2009, and 13 are considering furloughs, salary reductions, or layoffs or will employ such measures in the future. Table 5: State Survey Agency Vacancy Rates and Percentage of State Surveyors with Less Than 2 Years' Experience: Percentage of state agency directors' responses: Vacancy rate[A]: All states: 14; Low-understatement states: 12; High-understatement states: 24. Surveyors with less than 2 years' experience[B]: All states: 30; Low-understatement states: 25; High-understatement states: 38. Source: GAO. [A] Virginia did not provide the information needed to compute a vacancy rate and it was not a high-or low-understatement state. [B] Seven states did not report the number of surveyors with less than 2 years of experience. Among the high-and low-understatement states, only West Virginia, a low-understatement state, did not report this information. [End of table] [Sidebar: Surveyor Quotation about Inexperienced Staff: “I have been in this department for just over 3 years, and I still do not feel comfortable with the process. I could personally use a mentor to ensure a thorough understanding of the process. I don’t feel as if I can accurately identify deficiencies with the short amount of time given the survey teams to conduct surveys. I feel as if I overlook things due to trying to meet survey length time frames.” End of sidebar] Inexperienced Surveyors. Many states are attempting to accomplish their workload with a larger share of inexperienced surveyors, and state agency directors sometimes linked this reliance on inexperienced staff to the understatement of nursing home deficiencies. On average, 30 percent of surveyors had less than 2 years' experience (see table 5); however the percentage of inexperienced surveyors ranged from 10 to 82 percent across states who reported this information.[Footnote 53] Among state agency directors, 16 percent indicated that inexperienced surveyors always or frequently contributed to understatement, while another 48 percent indicated that surveyor inexperience sometimes contributed to understatement in their states. In response to our questionnaires, 26 percent of surveyors indicated that survey teams always or frequently had too many inexperienced surveyors and another 33 percent indicated that sometimes survey teams had too many inexperienced surveyors (see table 6). Half or more of all surveyors in six states--Alabama, Alaska, Arizona, Idaho, New Mexico, and Utah-- reported that there were always or frequently too many new surveyors who were not yet comfortable with their job responsibilities. For example, 79 percent of surveyors in Arizona reported that too many new surveyors were not comfortable with their job responsibilities, and the state agency director was among the 34 percent who reported that survey teams sometimes had an insufficient number of experienced surveyors. Overall, 26 percent of state agency directors indicated that the skill level of surveyors has decreased in the last 5 years. Table 6: Surveyors' and State Agency Directors' Responses to Questions on Workforce Issues: Percentage of responses: Respondents: Surveyors; Question: How frequently have you observed the following problems on the nursing home surveys that you have worked on? Too many new surveyors not yet comfortable with job responsibilities; Always/frequently: 26; Sometimes: 33; Infrequently/never: 35. Question: Survey team too small to conduct a thorough survey; Always/frequently: 21; Sometimes: 33; Infrequently/never: 42. Respondents: State Agency Directors; Question: In this state survey agency, how frequently do the following occur? Survey team not given sufficient time to conduct a thorough survey; Always/frequently: 25; Sometimes: 29; Infrequently/never: 42. Survey teams have a sufficient number of experienced surveyors; Always/frequently: 62; Sometimes: 34; Infrequently/never: 4. Survey teams are sufficient size to conduct thorough surveys; Always/frequently: 74; Sometimes: 18; Infrequently/never: 8. Survey teams are given sufficient time to conduct thorough surveys; Always/frequently: 78; Sometimes: 16; Infrequently/never: 6. Source: GAO. [End of table] In interviews, six state agency directors commented that inexperienced surveyors possessed different skills or needed more time than experienced surveyors to complete surveys and that workforce shortages resulted in constant recruiting, over-burdened experienced surveyors, or the need for additional supervision and training resources. Four states--Kentucky, Nevada, New Mexico, and Virginia--reported not having enough dedicated training staff to handle the initial training for new surveyors. [Sidebar: Surveyor Quotation about Insufficient Time to Complete Surveys: “Frequently G level or I/J level are not cited [due to a] lack [of] staff time.” End of sidebar] Workload. State inability to complete workload was, in part, an outgrowth of the workforce shortages and reliance on inexperienced surveyors. More than two-thirds of state agency directors reported on our questionnaire that staffing posed a problem for completing complaint surveys, and more than half reported that staffing posed a problem for completing standard or revisit surveys.[Footnote 54] In addition, 46 percent of state agency directors reported that time pressures always, frequently, or sometimes contributed to understatement in their states. In response to our questionnaire, 16 percent of surveyors nationwide reported that workload burden influenced the citation of deficiencies--including 14 states with 20 percent or more surveyors reporting the same. More than 50 percent of surveyors identified insufficient team size or time pressures as having an effect on the thoroughness of surveys. Surveyors' comments reiterated these concerns--over 15 percent of surveyors who wrote comments complained about the amount of time allotted to complete surveys or survey paperwork, and 11 percent indicated that staffing was insufficient to complete surveys.[Footnote 55] One state agency director suggested to us that CMS establish a national team of surveyors to augment states' when they fell behind on their workload or had staffing shortages. He thought the availability of national surveyors could assist states experiencing workforce shortages and help ensure state workloads were completed. This state had experience with a similar arrangement when it hired a national contractor to complete its surveys of Intermediate Care Facilities for the Mentally Retarded. Training Inadequacies May Compromise Survey Accuracy: Surveyors, state agency directors, and state performance on federal observational surveys indicated that inadequacies in initial or ongoing training may compromise the accuracy of nursing home surveys and lead to the understatement of deficiencies. In addition, workload affected surveyors' ability to attend training. Initial Surveyor Training. As noted earlier, even though CMS has established specific training requirements, including coursework and the SMQT certification test, states are responsible for preparing their new surveyors for the SMQT. According to CMS, 94 percent of new surveyors nationally passed the SMQT test in 2008 and, on average, surveyors answered about 77 percent of the questions correctly. These results seem to support the state agency directors' assertions that initial training was insufficient and suggest that the bar for passing the test may be set too low. Even though we cannot be certain whether the inadequacies are with the federal or state components of the training, reported differences among states in satisfaction with the initial surveyor training also could reflect gaps in state training programs. About 29 percent of surveyors in high-understatement states reported that initial training was not sufficient to cite appropriate scope and severity levels, compared with 16 percent of surveyors in low- understatement states (see table 7). Similarly, 28 percent of surveyors in high-understatement states, compared with 20 percent of those in low- understatement states, indicated that initial training was not sufficient to identify deficiencies for nursing homes. Further, 18 percent of state agency directors linked the occurrence of understatement always, frequently, or sometimes with insufficient initial training. From 16 to 20 percent of state agency directors indicated that initial training was insufficient to (1) enable surveyors to identify deficiencies and (2) assign the appropriate level of scope and severity. Table 7: Responses from Surveyors and State Agency Directors to Key Questions on Training: Initial training is not sufficient: To ensure surveyors are able to cite appropriate scope and severity levels; Percentage of surveyors' responses: All states: 24; Percentage of surveyors' responses: Low-understatement states: 16; Percentage of surveyors' responses: High-understatement states: 29; Percentage of directors' responses: 20. Initial training is not sufficient: To enable surveyors to identify deficiencies; Percentage of surveyors' responses: All states: 26; Percentage of surveyors' responses: Low-understatement states: 20; Percentage of surveyors' responses: High-understatement states: 28; Percentage of directors' responses: 16. Additional training is needed to: Interview nursing home residents; Percentage of surveyors' responses: All states: 13; Percentage of surveyors' responses: Low-understatement states: 9; Percentage of surveyors' responses: High-understatement states: 16; Percentage of directors' responses: 36[A]. Additional training is needed to: Identify scope and severity levels; Percentage of surveyors' responses: All states: 26; Percentage of surveyors' responses: Low-understatement states: 16; Percentage of surveyors' responses: High-understatement states: 34; Percentage of directors' responses: 56[B]. Additional training is needed to: Document deficiencies; Percentage of surveyors' responses: All states: 32; Percentage of surveyors' responses: Low-understatement states: 27; Percentage of surveyors' responses: High-understatement states: 35; Percentage of directors' responses: 62. Source: GAO. [A] Two state agency directors did not respond to this question. [B] One state agency director did not respond to this question. [End of table] Ongoing Training. Ongoing training programs are the purview of state agencies; therefore, differences between states about the sufficiency of this training also may point to gaps in the state training programs. On our questionnaire, about 34 percent of surveyors in high- understatement states indicated a need for additional training on (1) identifying appropriate scope and severity levels and (2) documenting deficiencies. This was significantly more than those from low- understatement states, which indicated less of a need for additional training in these areas--16 and 27 percent, respectively. Among state agency directors, 10 percent attributed understatement always or frequently to insufficient ongoing training, while 14 percent indicated that insufficient ongoing training sometimes gave rise to understatement. Although 74 percent of state agency directors indicated that the state had ongoing annual training requirements, the required number of hours and the type of training varied widely by state in 2007. Among the 33 states that provided the required amount of annual state training, these hours ranged from 0 to 120 hours per year. Meanwhile, 37 states reported one or more type of required training: 32 states required surveyors to attend periodic training, 22 required on- the-job training, 10 required online computerized training, and 13 states required some other type of training. State agency directors indicated that they relied on CMS materials for ongoing training of experienced surveyors, yet many reported additional training needs and suggested that use of electronic media could make continuing education and new guidance more accessible. While 98 percent of states indicated that the CMS written guidance materials and resources were useful, over 50 percent of all state agency directors identified additional training needs in documenting deficiencies, citing deficiencies at the appropriate scope and severity level, and applying CMS guidance. On federal observational surveys, an average of 17 to 12 percent of survey teams in high-understatement states received below-satisfactory ratings for Deficiency Determination and General Investigation, respectively--two skills critical for preventing understatement. In contrast, an average of 4 percent of survey teams in low-understatement states received the same below-satisfactory scores for both deficiency determination and investigative skills. Furthermore, of the 476 surveyors who commented about training needs, one-quarter indicated a need for training support from either CMS or state agencies; and between 12 to 7 percent of those who commented on training needs identified topics such as: documenting deficiencies, identifying scope and severity, CMS guidance, and medical knowledge. [Footnote 56] Inability to Attend Training. States' workload requirements and workforce shortages affected the surveyors' ability to attend initial and ongoing training. Seven of the eight state agency directors we interviewed linked workforce shortages and resource constraints to their state's ability to complete the survey workload or allow staff to participate in training courses. One director stated that workload demands compromised comprehensive training for new staff, and another reported difficulty placing new staff in CMS's initial training programs. Due to workload demands, a third state agency director stated that she could not allow experienced staff time away from surveying to attend training courses even when staff paid their own way. Five of the seven state agency directors suggested that it would be more efficient for training activities to be conducted more locally such as in their states or to be available through online, video, or other electronic media, and several emphasized the need to reduce or eliminate travel for training. Although four states also expressed a preference for interactive training opportunities, one state believed that technological solutions could allow for more accessible training that was also interactive. State Supervisory Reviews Often Are Not Designed to Identify Understatement: State supervisory reviews, which generally occurred more frequently on higher-level deficiencies, often are not designed to identify understated deficiencies. State agencies generally conducted more supervisory reviews on surveys with higher-level deficiencies, compared to surveys with deficiencies at the potential for more than minimal harm level (D through F)--the deficiencies most likely to be understated. While focus on higher-level deficiencies enables states to be certain that such deficiencies are well documented, not reviewing surveys with deficiencies at lower levels represents a missed opportunity to ensure that all serious deficiencies are cited. State surveyors who reported having frequent changes made to their survey reports during supervisory reviews also more often reported they were burdened by other factors contributing to understatement, such as workforce shortages and survey methodology weaknesses. Supervisory Reviews Often Focused on Higher-Level Deficiencies: According to state agency directors' responses to our questionnaire, states generally focused supervisory review on surveys with higher- level deficiencies, rather than on the surveys with deficiencies at the potential for more than minimal-harm level (D through F)--the deficiencies most likely to be understated. During supervisory reviews, either direct-line supervisors or central state agency staff may review draft survey records.[Footnote 57] On average, surveys at the D through F level underwent about two steps of review, while surveys with deficiencies at the immediate jeopardy level (J through L) went through three steps.[Footnote 58] For example, Washington reviews its surveys using either a two-step review that includes survey team and field manager reviews or a three-step process that includes both these reviews and an additional review by central state agency staff for serious deficiencies. As a result, central state agency staff in Washington do not review deficiencies below the level of actual harm. In addition we found that five states--Alaska, Hawaii, Illinois, Nebraska, and Nevada--did not review all surveys with deficiencies at the D through F levels. In fact, Hawaii did not report supervisory review of deficiencies at any level (see figure 3).[Footnote 59] It is difficult to know if additional supervisory reviews--the second, third, or fourth review--help make survey records more accurate and less likely to be understated, or if these reviews result in more frequent changes to deficiency citations. However, if deficiency citations with the potential for more than minimal-harm level (D through F) are not reviewed, states miss the opportunity to assess whether these deficiencies warrant a higher-level citation, for example, the level of actual harm or immediate jeopardy. Figure 3: Number of State Supervisory Reviews at the Potential for More than Minimal Harm (D-F) and Immediate Jeopardy Levels (J-L): [Refer to PDF for image: two map of the U.S. with associated data] States with D-F supervisory reviews: Alabama: 4; Alaska: 0; Arizona: 3; Arkansas: 2; California: 2; Colorado: 4; Connecticut: 4; Delaware: 2; Florida: 3; Georgia: 1; Hawaii: 0; Idaho: 1; Illinois: 0; Indiana: 1; Iowa: 4; Kansas: 2; Kentucky: 3; Louisiana: 1; Maine: 4; Maryland: 2; Massachusetts: 1; Michigan: 1; Minnesota: 2; Mississippi: 2; Missouri: 3; Montana: 2; Nebraska: 0; Nevada: 0; New Hampshire: 3; New Jersey: 2; New Mexico: 3; New York: 3; North Carolina: 1; North Dakota: 3; Ohio: 3; Oklahoma: 2; Oregon: 2; Pennsylvania: 4; Rhode Island: 4; South Carolina: 2; South Dakota: 2; Tennessee: 3; Texas: 4; Utah: 1; Vermont: 2; Virginia: 1; Washington: 2; West Virginia: 1; Wisconsin: 3; Wyoming: 1. States with J-L supervisory reviews: Alabama: 4; Alaska: 1; Arizona: 3; Arkansas: 3; California: 3; Colorado: 4; Connecticut: 4; Delaware: 2; Florida: 5; Georgia: 1; Hawaii: 0; Idaho: 1; Illinois: 4; Indiana: 5; Iowa: 4; Kansas: 3; Kentucky: 4; Louisiana: 3; Maine: 3; Maryland: 3; Massachusetts: 2; Michigan: 3; Minnesota: 5; Mississippi: 2; Missouri: 3; Montana: 2; Nebraska: 1; Nevada: 0; New Hampshire: 3; New Jersey: 3; New Mexico: 2; New York: 5; North Carolina: 4; North Dakota: 3; Ohio: 4; Oklahoma: 2; Oregon: 3; Pennsylvania: 5; Rhode Island: 4; South Carolina: 3; South Dakota: 2; Tennessee: 6; Texas: 5; Utah: 2; Vermont: 2; Virginia: 1; Washington: 2; West Virginia: 1; Wisconsin: 6; Wyoming: 1. Source: GAO. Map: Copyright © Corel Corp. All rights reserved. Note: Hawaii did not report conducting supervisory reviews. Forty states review a sample of all draft surveys. Such reviews may include additional examination of surveys with deficiencies at either the D through F or J through L levels. [End of figure] Because a majority of states are organized into geographically-based district or regional offices, review by central state agency staff, particularly quality assurance staff, is critical to help ensure consistency and detect understatement. However, 26 states reported that no central state agency staff reviews were conducted for surveys with deficiencies at the potential for more than minimal harm (D through F). These results are consistent with a finding from our 2003 report--that half of the 16 states we contacted for that report did not have a quality assurance process to help ensure that the scope and severity of less serious deficiencies were not understated.[Footnote 60] According to most of the eight state officials we interviewed, supervisory reviews commonly focused on documentation principles or evidentiary support, not on reducing understatement. For example, all eight states used supervisory reviews to assess the accuracy and strength of the evidence surveyors used to support deficiency citations, and three of these states reported that they emphasized reviewing survey records for documentation principles. Furthermore, seven out of eight states indicated that surveys with serious deficiencies--those that may be subject to enforcement proceedings-- went through additional steps of review compared with surveys citing deficiencies with the potential for more than minimal harm (D through F). Reports of Changes to Deficiencies during Supervisory Reviews May Be Related to Other Factors That Contribute to Understatement: Surveyor reports of changes to deficiency citations during supervisory reviews may be related to other factors the state is experiencing that also contribute to understatement, such as workforce shortages and survey methodology weaknesses. [Sidebar: Surveyor Quotation about Supervisory Review: “We have problems at times with nonclinical supervisors and district managers, [and with] a past branch chief not understanding clinical issues and thus not supporting surveyor findings. We’ve had deficiencies tossed out for surveys and IDR deficiencies deleted, not for lack of documentation, but for lack of understanding of the issues involved.” End of sidebar] Changes to Deficiencies. Fifty-four percent of surveyors nationwide reported on our questionnaire that supervisors at least sometimes removed the deficiency that was cited, and 53 percent of surveyors noted that supervisors at least sometimes changed the scope and severity level of cited deficiencies. Of the surveyors, who reported that supervisors sometimes removed deficiencies, 13 percent reported that supervisors always or frequently removed deficiencies--including 12 states with 20 percent or more of their surveyors reporting that deficiencies were removed. Surveyor reports of changes in deficiency citations alone make it difficult to know whether the original deficiency citation or the supervisor's revised citation was a more accurate reflection of a nursing home's quality of care. Additionally, there are many reasons that survey records might be changed during supervisory review. When a surveyor fails to provide sufficient evidence for deficient practices, it may be difficult to tell whether the deficiency was not appropriately cited or if the surveyor did not collect all the available evidence. Kentucky's state agency director offered one possible explanation--that changes to surveys often reflected a need for more support for the deficiencies cited, such as additional evidence from observations. Nevada's state agency director stated that changes to survey records occurred when it was often too late to gather more evidence in support of deficiencies. Surveyors who reported that supervisors frequently changed deficiencies also more often reported experiencing other factors that contribute to understatement. We found associations between surveyor reports of changes to deficiencies and workforce shortages and survey methodology weaknesses. * Workforce shortages. Surveyors reporting workforce shortages, including survey teams with too many new surveyors and survey teams that were either too small or given insufficient time to conduct thorough surveys, more often also reported that supervisors frequently removed deficiencies or changed the scope and severity of deficiency citations during supervisory reviews. * Survey methodology weaknesses. Surveyors reporting weaknesses in the current survey methodology more often also reported that supervisors frequently removed deficiencies or changed the scope and severity of deficiency citations during supervisory reviews. Supervisory Reviews and Understatement. In certain cases, survey agency directors and state performance on federal comparative surveys linked supervisory reviews to understatement. Twenty-two percent of state agency directors attributed inadequate supervisory review processes to understatement in their states at least sometimes.[Footnote 61] In addition, significant differences existed between zero-understatement states and all other states, including high-understatement states, in the percentage of surveyors reporting frequent changes to citations during supervisory reviews. Only about 4 percent of surveyors in zero- understatement states reported that citations were always or frequently removed or changed and that the scope and severity cited were changed, while about 12 percent of surveyors in all other states indicated the same (see table 8). To address concerns with supervisory reviews, Nevada recently reduced its process from two steps to a single step review by survey team supervisors to address surveyor complaints about changes made during supervisory reviews. Table 8: Percentage of Surveyors Reporting Changes in Deficiency Citations during Supervisory Review: Surveyors reporting that: Supervisors always or frequently remove or change deficiency cited; Percentage of surveyors' responses: Zero-understatement states: 5; Percentage of surveyors' responses: All other states: 13. Surveyors reporting that: Supervisors always or frequently changed the scope and severity cited; Percentage of surveyors' responses: Zero-understatement states: 4; Percentage of surveyors' responses: All other states: 12. Source: GAO. [End of table] In addition, we observed a relationship between state practices to notify surveyors of changes made during supervisory reviews and surveyor reports of deficiency removal and explanation of changes. Specifically, compared to surveyors in states that require supervisors to notify surveyors of changes made during supervisory review, surveyors from states where no notification is required reported more often that supervisors removed deficiencies and less often that explanations for these changes, when given, were reasonable. Similarly, we found an association between the frequency of explained and reasonable changes and zero-understatement states, possibly demonstrating the positive effect of practices to notify surveyors of changes made during supervisory reviews. Nursing home surveyors from zero-understatement states more often reported that supervisors explained changes and that their explanations seemed reasonable compared to surveyors in all other states. State agency directors in Massachusetts and New Mexico stated that explanations of changes to the survey record provided opportunities for one-on-one feedback to surveyors and discussions about deficiencies being removed. State Agency Practices and External Pressure May Compromise Survey Accuracy and Lead to Understatement in a Few States: Nursing home surveyors and state agency directors in a minority of states told us that in isolated cases issues such as a state agency practice of noncitation, external pressure from the nursing home industry, and an unbalanced IDR process may have led to the understatement of deficiencies. In a few states, surveyors more often identified problems with noncitation practices and IDR processes compared to state agency directors. Yet, a few state agency directors acknowledged either noncitation practices, external pressure, or an IDR process that favored nursing home operators over resident welfare. Although not all the issues raised by surveyors were corroborated by the state agency directors in their states, surveyor reports clustered in a few states gives credence to the notion that such conditions may lead to understatement. Surveyors Reported Noncitation Practices in a Small Number of States: [Sidebar: Surveyor Quotation about State Noncitation Practices: “I have been criticized by my supervisor on more than one occasion for citing too many deficiencies at facilities [that] have an ongoing history of repeat tags from survey to survey and many complaints surveys between annual surveys. My supervisor states that citing too many deficiencies ‘confuses’ the facility and creates a ‘hostile’ environment.” End of sidebar] Approximately 20 percent of surveyors nationwide and over 40 percent of surveyors in five states reported that their state agency had at least one of the following noncitation practices: (1) not citing certain deficiencies, (2) not citing deficiencies above a certain scope and severity level, and (3) allowing nursing homes to correct deficiencies without receiving a citation (see figure 4). Only four state agency directors acknowledged the existence of such practices in their states on our questionnaire and only one of these directors was from the five states most often identified by surveyors. One of these directors commented on our questionnaire that one of these practices occurs only in "rare individual cases." Another director commented that a particular federal quality standard is not related to patient outcome and therefore should not be cited above a Level F. According to CMS protocols, when noncompliance with a federal requirement has been identified, the state agency should cite all deficiencies associated with the noncompliance. CMS regional officials we interviewed were not aware of any current statewide noncitation practices.[Footnote 62] Figure 4: Percentage of Surveyors in Each State Reporting at Least One Noncitation Practice: [Refer to PDF for image: U.S. map and associated data] No data: Pennsylvania[A]. 0-10 Noncitation Practices: Alabama: Alaska: Georgia: Hawaii: Idaho: Montana: Oklahoma: Tennessee: Vermont: West Virginia. 11-20 Noncitation Practices: Colorado: Connecticut: District of Columbia: Florida: Illinois: Iowa: Kentucky: Maine: Massachusetts: Michigan: Minnesota: Mississippi: Missouri: New York: North Dakota: Rhode Island: South Carolina: Virginia: Washington: Wisconsin. 21-30 Noncitation Practices: Arkansas: California: Indiana: Louisiana: Maryland: New Jersey: North Carolina: Ohio: Oregon: South Dakota: Texas: Utah: Wyoming. 31-40 Noncitation Practices: Arizona: Kansas. 41 on more Noncitation Practices: Delaware: Nebraska: Nevada: New Hampshire: New Mexico. Source: GAO. Map: Copyright © Corel Corp. All rights reserved. [A] Responses from Pennsylvania surveyors could not be included because the state agency directed nursing home surveyors not to respond to our questionnaire. [End of figure] Not citing certain deficiencies. Nationally, 9 percent of surveyors reported a state agency practice that surveyors not cite certain deficiencies. However, in four states over 30 percent of surveyors reported their state agency had this noncitation practice, including over 60 percent of New Mexico surveyors. In some cases, surveyors reported receiving direct instructions from supervisors not to cite certain deficiencies. In other cases, surveyors' reports of noncitation practices may have been based on their interpretation of certain management practices. For instance, surveyors commented that some state agency practices--such as providing inadequate time to observe and document deficiencies or frequently deleting deficiency citations during supervisory review--seemed like implicit or indirect leadership from the agency to avoid citing deficiencies. One state agency director we interviewed agreed that surveyors may report the existence of noncitation practices when their citations are changed during supervisory review. This official told us that when surveyors' deficiencies are deleted or downgraded, the surveyors may choose not to cite similar deficiencies in the future because they perceive being overruled as an implicit state directive not to cite those deficiencies. Not citing deficiencies above a certain scope and severity level. Although nationwide less than 8 percent of surveyors reported a state agency practice that surveyors not cite deficiencies above a certain scope and severity level, in two states over 25 percent of surveyors reported that their state agency used this type of noncitation practice. One reason state agencies might use this noncitation practice could be to help manage the agency's workload. In particular, citing deficiencies at a lower scope and severity might help the agency avoid additional work associated with citing higher-level deficiencies, such as survey revisits or IDR.[Footnote 63] In one of the two states mentioned above, 54 percent of surveyors indicated that the workload burden influenced their citations. Additionally, as we described earlier, 16 percent of surveyors nationwide indicated that workload burden influenced the citation of deficiencies and more than half of state agency directors (including those from the two states mentioned above) responded that staffing was not sufficient to complete revisit surveys. While our questionnaire focused on not citing deficiencies above a certain scope and severity level, a few surveyors commented on being discouraged from citing lower-level deficiencies due to time pressures to complete surveys. Agency officials in two states told us that surveyors may miss some deficiencies due to limited survey time and resources. Allowing nursing homes to correct deficiencies without citing them on the survey record. Nationwide, approximately 12 percent of surveyors reported this type of noncitation practice. However, in five states, at least 30 percent of surveyors reported their state agency allowed nursing homes to correct deficiencies without citing those deficiencies on the official survey record. Comments from surveyors suggest that state agencies may use this type of practice to avoid actions that nursing homes or the industry would dispute or interpret as excessive. Similarly, several surveyors commented that they were instructed to cite only one deficiency for a single type of negative outcome, even when more than one problem existed. However, CMS guidance requires state agencies to cite all problems that lead to a negative outcome. The decrease in G-level citations that occurred after CMS implemented the double G immediate sanctions policy in January 2000 also suggests that some states may have avoided citing deficiencies that would result in enforcement actions for the nursing home.[Footnote 64] The total number of G-level deficiency citations nationwide dropped from approximately 10,000 in 1999 to 7,700 in 2000.[Footnote 65] State Agency Directors in a Few States Reported That External Pressure Contributed to Understatement: [Sidebar: Surveyor Quotation about External Pressure: “The larger corporations often pressure our [state agency] Central Office to change and delete citations. Our Central Office changes not only wording but content and intent of the citation, when they were not on site. There is a great deal of political push and pull—the interference from State Senators and Representatives protecting their re-electability and not the rights of the residents (who don’t vote).” End of sidebar] State agency directors from 12 states reported experiencing external pressure from at least one of the following stakeholder groups: (1) the nursing home surveyed, (2) the nursing home industry, or (3) state or federal legislators. Examples of such external pressure include pressure to reduce federal or state nursing home regulation or to delete specific deficiencies cited by the state agency. Of the 12 state agency directors, 7 reported that external pressure at least sometimes contributed to the understatement of deficiencies in their states, while the other 5 indicated that it infrequently or never contributed to understatement. Adversarial attitude toward nursing home surveys. Two states we interviewed--State A and State B--commented on the adversarial attitude that industry and legislative representatives had toward nursing home surveys at times.[Footnote 66] For instance, state agency officials from State A told us that the state nursing home association organized several forums to garner public and legislative support for curtailing state regulation of facilities. According to officials in this state, the influential industry groups threatened to request legislation to move the state agency to a different department and to deny the confirmation of the director's gubernatorial appointment if the citations of G level or higher deficiencies increased. CMS regional office officials responsible for State A told us that the state may be experiencing more intense external pressure this year given the current economy, because providers have greater concerns about the possible financial implications of deficiency citations--fines or increased insurance rates. Similarly, officials from State B told us that when facilities are close to termination, the state agency receives phone calls from state delegates questioning the agency's survey results. Officials from State B also told us that the Governor's office instructed the state agency not to recommend facilities for enforcement actions. Officials from the CMS regional office responsible for State B told us that this situation was not problematic because CMS was ultimately responsible for determining enforcement actions based on deficiency citations. However, this regional office's statement is inconsistent with (1) language in the SOM that calls for states to recommend enforcement actions to the regional office, and (2) assertions from the regional office responsible for State A that it infrequently disagrees with state recommendations for sanctions. A third state agency director commented that the agency had been called before state legislative committees in 2007-2008 to defend deficiency citations that led to the termination of facilities. A fourth state agency director also commented on our questionnaire that legislators had pressured the state agency on behalf of nursing homes to get citations reduced or eliminated and prevent enforcement actions for the facilities. In addition, a few surveyors commented that at times when nursing homes were unhappy with their survey results the homes or their state legislators would ask state agency management to remove the citations from the survey record, resulting in the deletion or downgrading of deficiencies. Further, comments from a few surveyors indicated that they may steer clear of citing deficiencies when they perceive the citation might cause a home to complain or exert pressure for changes in the survey record. Interference in the survey process. In a few cases, external pressure appeared to directly interfere with the nursing home survey process. State agency officials from two states--State A and an additional fifth state--reported that state legislators or industry representatives had appeared on-site during nursing home surveys. Although in some cases the legislators just observed the survey process, officials from these two states explained that third parties also have interfered with the process by questioning or intimidating surveyors. The state agency director from the fifth state commented on our questionnaire that the nursing home industry sent legal staff on-site during surveys to interfere with the survey process. Similarly, officials from State A told us that during one survey, a home's lawyer was on-site reviewing nursing home documentation before surveyors were given access to these documents. Officials from State A also told us that state legislators have attended surveys to question surveyors about their work and whether state agency executives were coercing them to find deficiencies. We discussed this issue with the CMS regional officials responsible for State A, who acknowledged that this type of interference had occurred. States' need for support from CMS. In the face of significant external pressure, officials from States A and B suggested that they need support from CMS; however, CMS regional office officials did not always acknowledge external pressure reported by the states. This year, State A terminated a survey due to significant external pressure from a nursing home and requested that the CMS regional office complete the revisit survey for them. Six weeks later, the federal team completed the survey and found many of the same problems that this state team had previously identified before it stopped the survey. Officials from State A suggested the need for other support as well, such as creating a federal law that would require state agencies to report external pressure and ensure whistleblower protections for state officials who report pressure and allowing sanctions for inappropriate conduct. CMS officials from the regional office responsible for State A stated that external pressure might indirectly contribute to understatement by increasing surveyor mistakes from the additional stress, workload, focus on documentation, and supervisory reviews. Conversely, CMS regional officials did not acknowledge that State B experienced external pressure and officials from State B thought that CMS should be more consistent in its requirements and enforcement actions. Unbalanced IDR Processes Might Have Contributed to Understatement in a Few States: States with unbalanced IDR processes may experience more understatement. IDR processes vary across states in structure, volume of proceedings, and resulting changes. According to state agency directors' responses to our questionnaire, 16 IDRs were requested per 100 homes in fiscal year 2007, with this number ranging among states from 0 to 57 per 100 homes.[Footnote 67] For IDRs occurring in fiscal year 2007, 20 percent of disputed deficiencies were deleted and 7 percent were downgraded in scope or severity, but in four states, at least 40 percent of disputed deficiencies were deleted through this process.[Footnote 68] CMS does not provide protocols on how states should operate their IDR processes, leaving IDR operations to state survey agencies' discretion. For example, states may choose to conduct IDR meetings in writing, by telephone, or through face-to-face conferences. State agencies also have the option to involve outside entities, including legal representation, in their IDR operations. [Sidebar: Surveyor Quotation about the IDR Process: “The IDR process is inconsistent. Over the years we have been on all ends of the spectrum—between having involved panels who have an understanding of the survey process versus people who know nothing about the process and have no idea how to apply the federal regulations. (The latter is the current make-up.) When we have a panel made up of the latter, the word spreads throughout the state and there is a very large increase in requests for IDR. The reason is that this type of panel tends to delete most everything for ‘insufficient evidence’ but cannot coherently explain how they came [to] that decision.” End of sidebar] On the basis of responses from surveyors and state agency directors clustered in a few states, problems with the IDR processes--such as frequent hearings, deficiencies that are frequently deleted or downgraded through the IDR process, or outcomes that favor nursing home operators over resident welfare--may have contributed to the understatement of deficiencies in those states. Although reports of such problems were not common--only 16 percent of surveyors nationwide reported on our questionnaire that their state's IDR process favored nursing home operators--in four states over 40 percent of surveyors reported that their IDR process favored nursing home operators (see figure 5), including one state where a substantial percentage of surveyors identified at least one noncitation practice. While only one state agency director reported that the IDR process favored nursing home operators, three other directors acknowledged that frequent IDR hearings at least sometimes contributed to the understatement of deficiencies. For example, in some states surveyors may hesitate to cite deficiencies that they believe will be disputed by the nursing home. Figure 5: Percentage of Surveyors in Each State Reporting the IDR Process Favored Concerns of Nursing Home Operators over Resident Welfare: [Refer to PDF for image: U.S. map and associated data] No data: Pennsylvania[A]. 0-10% of Surveyors: Alaska: Arkansas: California: Colorado: Connecticut: District of Columbia: Georgia: Hawaii: Idaho: Illinois: Indiana: Kansas: Maine: Michigan: Mississippi: Montana: New Hampshire: New Mexico: North Dakota: Oklahoma: Oregon: Rhode Island: South Dakota: Utah: Vermont: Wisconsin. 11-20% of Surveyors: Alabama: Arizona: Florida: Iowa: Minnesota: Missouri: Nevada: New York: North Carolina: South Carolina: Tennessee: West Virginia. 21-30% of Surveyors: Kentucky: Maryland: Nebraska: New Jersey: Ohio: Texas: Washington: Wyoming. 31-40% of Surveyors: None. 41% of more of Surveyors: Delaware: Louisiana: Massachusetts: Virginia. Source: GAO. Map: Copyright © Corel Corp. All rights reserved. [A] Responses from Pennsylvania surveyors could not be included because the state agency directed nursing home surveyors not to respond to our questionnaire. [End of figure] In isolated cases, a lack of balance with the IDR process appeared to be a result of external pressure. In one state, the state agency director reported that the nursing home industry sent association representatives to the IDR, which increased the contentiousness of the process. In another state, officials told us that a large nursing home chain worked with the state legislature to set up an alternative to the state IDR process, which has been used only by facilities in this chain. Through this alternative appeals process, both the state agency and the nursing home have legal representation, and compliance decisions are made by an adjudicator. According to agency officials in this state, the adjudicators for this alternative appeals process do not always have health care backgrounds. While CMS gives states the option to allow outside entities to conduct the IDR, the states should maintain ultimate responsibility for IDR decisions.[Footnote 69] CMS regional officials stated it would not consider the outcome of this alternative appeals process when assessing deficiencies or determining enforcement actions. Regardless, these actions may have affected surveyors' perceptions of the balance of the states' IDRs, because over twice the national average of surveyors in this state reported that their IDR process favored nursing home operators. Conclusions: Reducing understatement is critical to protecting the health and safety of vulnerable nursing home residents and ensuring the credibility of the survey process. Federal and state efforts will require a sustained, long-term commitment because understatement arises from weaknesses in several interrelated areas--including CMS's survey process, surveyor workforce and training, supervisory review processes, and state agency practices and external pressure. * Concerns about CMS's Survey Process. Survey methodology and guidance are integral to reliable and consistent state nursing home surveys, and we found that weaknesses in these areas were linked to understatement by both surveyors and state agency directors. Both groups reported struggling to interpret existing guidance, and differences in interpretation were linked to understatement, especially in determining what constitutes actual harm. Surveyors noted that the current survey guidance was too lengthy, complex, and subjective. Additionally, they had fewer concerns about care areas for which CMS has issued revised interpretive protocols. In its development of the QIS, CMS has taken steps to revise the nursing home survey methodology. However, development and implementation of the QIS in a small group of states has taken approximately 10 years, and full implementation of the new methodology is not expected to be completed until 2014. The experience of the QIS was mixed regarding improvement in the quality of surveys, and the independent evaluation generated a number of recommendations for improving the QIS. CMS concluded that it needed to focus future QIS development efforts on improving survey consistency and giving supervisors more tools to assess performance of surveyor teams. * Ongoing Workforce and Surveyor Training Challenges. Workforce shortages in state survey agencies increase the need for high-quality initial and ongoing training for surveyors. Currently, high vacancy rates can place pressure on state surveyors to complete surveys under difficult circumstances, including compressed time frames, inadequately staffed survey teams, and too many inexperienced surveyors. States are responsible for hiring and retaining surveyors and have grappled with pervasive and intractable workforce shortages. State agency directors struggling with these workforce issues reported the need for more readily accessible training for both their new and experienced surveyors that did not involve travel to a central location. Nearly 30 percent of surveyors in high-understatement states stated that initial surveyor training, which is primarily a state activity that incorporates two CMS on-line computer courses and a 1-week federal basic training course culminating in the SMQT, was not adequate to identify deficiencies and cite them at the appropriate scope and severity level. State agency directors reported that workforce shortages also impede states' ability to provide ongoing training opportunities for experienced staff and that additional CMS online training and electronic training media would help states maintain an experienced, well-informed workforce. They noted that any such support should be cognizant of states' current resource constraints, including limited funding of travel for training. * Supervisory Review Limitations. Currently, CMS provides little guidance on how states should structure supervisory review processes, leaving the scope of this important quality-assurance tool exclusively to the states and resulting in considerable variation throughout the nation in how these processes are structured. We believe that state quality assurance processes are a more effective preventive measure against understatement because they have the potential to be more immediate and cover more surveys than the limited number of federal comparative surveys conducted in each state. However, compared to reviews of serious deficiencies, states conducted relatively fewer reviews of deficiencies at the D through F level, those that were most frequently understated throughout the nation, to assess whether or not such deficiencies were cited at too low a scope and severity level. In addition, we found that frequent changes to survey results made during supervisory review were symptomatic of workforce shortages and survey methodology weaknesses. For example, surveyors who reported that survey teams had too many new surveyors, more often also reported either frequent changes to or removals of deficiencies during supervisory reviews--indicating that states with inexperienced workforces may rely more heavily on supervisory reviews. In addition, variation existed in the type of feedback surveyors receive when deficiencies are changed or removed during supervisory reviews, providing surveyors with inconsistent access to valuable feedback and training. CMS did not implement our previous recommendation to require states to have a quality assurance process that includes, at a minimum, a review of a sample of survey reports below the actual harm level to assess the appropriateness of the scope and severity cited and help reduce understatement. * State Agency Practices and External Pressure. In a few states, noncitation practices, challenging relationships with the industry or legislators, or unbalanced IDR processes--those that surveyors regard as favoring nursing home operators over resident welfare--may have had a negative effect on survey quality and resulted in the citation of fewer nursing home deficiencies than was warranted. In one state, both the state agency director and over 40 percent of surveyors acknowledged the existence of a noncitation practice such as allowing a home to correct a deficiency without receiving a citation. Forty percent of surveyors in four other states also responded on our questionnaire that noncitation practices existed. Currently, CMS does not explicitly address such practices in its guidance to states, and its oversight is limited to reviews of citation patterns, feedback from state surveyors, state performance reviews, and federal monitoring surveys to determine if such practices exist. Twelve state agency directors reported on our questionnaire experiencing some kind of external pressure. For example, in one state a legislator attended a survey and questioned surveyors as to whether state agency executives were coercing them to find deficiencies. Under such circumstances, it is difficult to know if the affected surveyors are consistently enforcing federal standards and reporting all deficiencies at the appropriate scope and severity levels. States' differing experiences regarding the enforcement of federal standards and collaboration with their CMS regional offices in the face of significant external pressure also may confuse or undermine a thorough and independent survey process. If surveyors believe that CMS does not fully or consistently support the enforcement of federal standards, these surveyors may choose to avoid citing deficiencies that they perceive may trigger a reaction from external stakeholders. In addition, deficiency determinations may be influenced when IDR processes are perceived to favor nursing home operators over resident welfare. Because many aspects of federal and state operations contribute to the understatement of deficiencies on nursing home surveys, mitigating this problem will require the concerted effort of both entities. The interrelated nature of these challenges suggests a need for increased CMS attention on the areas noted above and additional federal support for states' efforts to enforce federal nursing home quality standards. Recommendations for Executive Action: To address concerns about weaknesses in CMS survey methodology and guidance, we recommend that the Administrator of CMS take the following two actions: * make sure that action is taken to address concerns identified with the new QIS methodology, such as ensuring that it accurately identifies potential quality problems; and: * clarify and revise existing CMS written guidance to make it more concise, simplify its application in the field, and reduce confusion, particularly on the definition of actual harm. To address surveyor workforce shortages and insufficient training, we recommend that the Administrator of CMS take the following two actions: * consider establishing a pool of additional national surveyors that could augment state survey teams or identify other approaches to help states experiencing workforce shortages; * evaluate the current training programs and division of responsibility between federal and state components to determine the most cost- effective approach to: (1) providing initial surveyor training to new surveyors, and (2) supporting the continuing education of experienced surveyors. To address inconsistencies in state supervisory reviews, we recommend that the Administrator of CMS take the following action: * set an expectation through guidance that states have a supervisory review program as a part of their quality-assurance processes that includes routine reviews of deficiencies at the level of potential for more than minimal harm (D-F) and that provides feedback to surveyors regarding changes made to citations. To address state agency practices and external pressure that may compromise survey accuracy, we recommend that the Administrator of CMS take the following two actions: * reestablish expectations through guidance to state survey agencies that noncitation practices--official or unofficial--are inappropriate, and systematically monitor trends in states' citations; and: * establish expectations through guidance to state survey agencies to communicate and collaborate with their CMS regional offices when they experience significant pressure from legislators or the nursing home industry that may affect the survey process or surveyors' perceptions. Agency and AHFSA Comments and Our Evaluation: We provided a draft of this report to HHS and AHFSA for comment. In response, the Acting Administrator of CMS provided written comments. CMS noted that the report adds value to important public policy discussions regarding the survey process and contributes ideas for solutions on the underlying potential causes of understatement. CMS fully endorsed five of our seven recommendations and indicated it would explore alternate solutions to our remaining two recommendations, one of which the agency did not plan to implement on a national scale. (CMS's comments are reprinted in appendix II.) AHFSA's comments noted that several states agreed with one of our recommendations, but did not directly express agreement or disagreement with the other recommendations. AHFSA made several other comments on our findings and recommendations as summarized below. CMS: CMS agreed with five of our recommendations that called for: (1) addressing issues identified with the new QIS methodology, (2) evaluating current training programs, (3) setting expectations that states have a supervisory review program, (4) reestablishing expectations that noncitation practices are inappropriate, and (5) establishing expectations that states communicate with their CMS regional office when they experience significant pressure from legislators or the nursing home industry. In its comments, the agency cited several ongoing efforts as mechanisms for addressing some of our recommendations. While we acknowledge the importance of these ongoing efforts, in some areas we believe more progress and investigation are likely needed to fully address our findings and recommendations. For example, we recommended that CMS ensure that measures are taken to address issues identified with the new QIS methodology, such as ensuring that it accurately identifies potential quality problems; CMS's response cited Desk Audit Reports that enable supervisors to provide improved feedback to surveyors and quarterly meetings of a user group as evidence of efforts under way to continuously improve the QIS and to increase survey consistency. However, we noted that a 2007 evaluation of the QIS did not find improved survey accuracy compared to the traditional survey process and recommended that CMS evaluate how well the QIS accurately identifies areas in which there were potential quality problems. While improving the consistency of the survey process is important, CMS must also focus on addressing the accuracy of QIS surveys. For the remaining two recommendations, CMS described alternative solutions that it indicated the agency would explore: * Guidance. The agency agreed in principle with our recommendation to clarify and revise existing written guidance to make it more concise, simplify its application in the field, and reduce confusion. However, CMS disagreed with shortening the guidance as the preferred method for achieving such clarification. Instead, the agency suggested an alternative--the creation of some short reference documents for use in the field that contain cross-links back to the full guidance--that we believe would fulfill the intent of our recommendation. * National surveyor pool. CMS indicated it did not plan to implement our recommendation to consider establishing a pool of additional national surveyors that could augment state survey teams experiencing workforce shortages, at least not on a national scale. The agency stated that the establishment of national survey teams was problematic for several reasons, including that it (1) began to blur the line between state accountability for meeting performance expectations and compensating states for problematic performance due to state management decisions, and (2) was improper for CMS to tell states how to make personnel decisions While the agency noted that it used national contractors to perform surveys for other types of facilities such as organ transplant centers, it expressed concern about their use to compensate for state performance issues because of the more frequent nursing home surveys. We believe that state workforce shortages are a separate issue from state performance on surveys. Since 2003, we have reported pervasive state workforce shortages and this report confirms that such shortages continue.[Footnote 70] For example, we reported that one-fourth of states had vacancy rates higher than 19 percent and that one state reported a 72 percent vacancy rate. We also believe that addressing workforce shortages is critical to creating an effective system of oversight for nursing homes and reducing understatement throughout the nation. However, CMS noted that it would explore this issue with a state- federal work group in order to identify any circumstances in which a national pool may be advisable and to identify any additional solutions. Reflecting this comment from CMS, we have revised our original recommendation to include other potential solutions as well as a national pool of surveyors. One suggestion in AHFSA comments may be worth exploring in this regard--providing funds to state survey agencies for recruitment and retention activities. AHFSA: AHFSA commented that vigorous oversight and enforcement are essential to improving the quality of life and quality of care for health care consumers and are critical if improvements already achieved are to be maintained. The association noted that several states agreed with our recommendation on the need for CMS to revise existing written guidance to make it more concise. While the association did not directly express agreement or disagreement with our other recommendations, it did note that most states would need additional funding to meet any new staffing requirements associated with our recommendation that CMS set an expectation for states to have a supervisory review program. However, AHFSA noted what it considered to be conflicting assertions within the report. For example, it noted that we cited inexperienced staff as a factor that contributes to understatement but also appeared to take issue with the practice of supervisors changing reports prepared by inexperienced staff. While our report identifies a wide variety of factors that may contribute to understatement, we did not and could not meaningfully prioritize among these factors based on the responses of nursing home surveyors and state agency directors. We did find that many states were attempting to accomplish their survey workload with a large share of inexperienced surveyors and that state agency directors sometimes linked this reliance on inexperienced staff to the understatement of nursing home deficiencies. In addition, we found that frequent changes made during supervisory review were symptomatic of workforce shortages and survey methodology weaknesses. For example, surveyors who reported that survey teams had too many new surveyors, more often also reported either frequent changes to or removals of deficiencies during supervisory reviews. We believe that state quality assurance processes have the potential to play an important role in preventing understatement, which may result in states with inexperienced workforces relying more heavily on supervisory reviews. AHFSA also stated that our report did not address limitations of federal monitoring surveys, specifically the potential inconsistency among CMS regional offices in how these surveys are conducted. Assessing CMS's performance on federal monitoring surveys was beyond the scope of this report. However, our May 2008 report noted several improvements CMS had made since fiscal years 2002 and 2003 in federal comparative surveys intended to make them more comparable to the state surveys they are assessing; these improvement include; (1) reducing the time between the state and federal surveys to ensure that they more accurately capture the conditions at the time of the state survey, (2) including at least half of the residents from state survey investigative samples to allow for a more clear-cut determination of whether the state survey should have cited a deficiency, and (3) using the same number of federal surveyors as the corresponding state survey, again to more closely mirror the conditions under which the state survey was conducted.[Footnote 71] Finally, AHFSA questioned whether the information that we received from surveyors about the IDR process was universally valid because their input about quality assurance reviews might be biased. Our methodology did not rely solely on surveyor responses to our questionnaire but used a separate questionnaire sent to state survey agency directors to help corroborate their responses. Thus we reported both that (1) over 40 percent of surveyors in four states indicated that their IDR process favored nursing home operators and (2) one state survey agency director agreed and three others acknowledged that frequent IDR hearings sometimes contributed to the understatement of deficiencies. We also collected and reported data on the number of deficiencies modified or overturned, which AHFSA said was a more accurate measure of the effect of IDRs. We also incorporated technical comments from AHFSA as appropriate. As arranged 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 Administrator of the Centers for Medicare & Medicaid Services and appropriate congressional committees. In addition, the report will be available at no charge on GAO's Web site at [hyperlink, http://www.gao.gov]. If you or your staffs have any questions about this report, please contact me at (202) 512-7114 or dickenj@gao.gov. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix II. Signed by: John E. Dicken: Director, Health Care: [End of section] Appendix I: Scope and Methodology: This appendix describes the data and methods we used to identify the factors that contribute to the understatement of serious deficiencies on nursing home surveys.[Footnote 72] This report relies largely on the data collected through (1) two GAO-administered Web-based questionnaires to nursing home surveyors and state agency directors and (2) analysis of federal and state nursing home survey results as reported in the federal monitoring survey database and the On-Line Survey, Certification, and Reporting (OSCAR) system. Summary results from the GAO questionnaires are available as an e-supplement to this report. See Nursing Homes: Responses from Two Web-Based Questionnaires to Nursing Home Surveyors and State Agency Directors (GAO-10-74SP), an E-supplement to GAO-10-70. To augment our quantitative analysis, we also interviewed officials at the Centers for Medicare & Medicaid (CMS) Survey and Certification Group and select regional offices;[Footnote 73] reviewed federal regulations, guidance, and our prior work; and conducted follow-up interviews with eight state agency directors and a select group of surveyors. Except where otherwise noted, we used data from fiscal year 2007 because they were the most recently available data at the time of our analysis. Development of Questionnaires and Analysis of Responses: We developed two Web-based questionnaires--one for the nursing home surveyors and one for the state agency directors. Development of the Questionnaires: The questionnaires were developed and the data collection and analysis conducted to (1) minimize errors arising from differences in how a particular question might be interpreted and in the sources of information available to respondents and (2) reduce variability in responses that should be qualitatively the same. GAO social science survey specialists aided in the design and development of both questionnaires. We pretested the two questionnaires with six surveyors from a local state and five former or current state agency directors, respectively. Based on feedback from these pretests, the questionnaires were revised to improve clarity and the precision of responses, and ensure that all questions were fair and unbiased. Most questions were closed-ended, which limited the respondent to answers such as yes or no, or to identifying the frequency that an event occurred using a scale--always, frequently, sometimes, infrequently, or never. For reporting purposes, we grouped the scaled responses into three categories--always/frequently, sometimes, and infrequently/never. Both questionnaires included some open-ended questions to allow respondents to identify specific training needs or other concerns. With few exceptions, respondents entered their responses directly into the Web-based questionnaire databases.[Footnote 74] These questionnaires were sent to the eligible population of nursing home surveyors and all state agency directors. We performed computer analyses to identify illogical or inconsistent responses and other indications of possible error. We also conducted follow-up interviews with select respondents to clarify and gain a contextual understanding of their responses.[Footnote 75] Questionnaire for Nursing Home Surveyors: This questionnaire was designed to gather information from nursing home surveyors nationwide about the process for identifying and citing nursing home deficiencies. It included questions about various aspects of the survey process identified by our prior work that may contribute to survey inconsistency and the understatement of deficiencies. Such aspects included survey methodology and guidance, deficiency determination, surveyor training, supervisory review of draft surveys, and state agency policies and procedures.[Footnote 76] We fielded the questionnaire from May through July 2008 to 3,819 eligible nursing home surveyors. To identify the eligible population, we downloaded a list of identification numbers for surveyors who had conducted at least one health survey of a nursing home in fiscal years 2006 or 2007 from CMS's OSCAR database and we obtained surveyors' e- mail addresses from state survey agencies. We received complete responses from 2,340 state surveyors, for a 61 percent response rate.[Footnote 77] The state-level response rates were above 40 percent for all but three states--Connecticut, Illinois, and Pennsylvania.[Footnote 78] We excluded Pennsylvania from our analysis because Pennsylvania's Deputy Secretary for Quality Assurance instructed the state's surveyors not to respond to our survey and few responded. (For response rates by state, see table 9.) Table 9: Response Rates to GAO's Questionnaire of Nursing Home Surveyors, 2008: State: Alabama; Number of respondents: 36; Number of eligible surveyors: 52; Response rate: 69%. State: Alaska; Number of respondents: 4; Number of eligible surveyors: 6; Response rate: 67%. State: Arizona; Number of respondents: 19; Number of eligible surveyors: 28; Response rate: 68%. State: Arkansas; Number of respondents: 28; Number of eligible surveyors: 54; Response rate: 52%. State: California; Number of respondents: 306; Number of eligible surveyors: 544; Response rate: 56%. State: Colorado; Number of respondents: 16; Number of eligible surveyors: 38; Response rate: 42%. State: Connecticut; Number of respondents: 17; Number of eligible surveyors: 61; Response rate: 28%. State: Delaware; Number of respondents: 13; Number of eligible surveyors: 16; Response rate: 81%. State: District of Columbia; Number of respondents: 6; Number of eligible surveyors: 10; Response rate: 60%. State: Florida; Number of respondents: 128; Number of eligible surveyors: 226; Response rate: 57%. State: Georgia; Number of respondents: 47; Number of eligible surveyors: 54; Response rate: 87%. State: Hawaii; Number of respondents: 4; Number of eligible surveyors: 7; Response rate: 57%. State: Idaho; Number of respondents: 6; Number of eligible surveyors: 11; Response rate: 55%. State: Illinois; Number of respondents: 34; Number of eligible surveyors: 171; Response rate: 20%. State: Indiana; Number of respondents: 92; Number of eligible surveyors: 101; Response rate: 91%. State: Iowa; Number of respondents: 37; Number of eligible surveyors: 59; Response rate: 63%. State: Kansas; Number of respondents: 34; Number of eligible surveyors: 59; Response rate: 58%. State: Kentucky; Number of respondents: 44; Number of eligible surveyors: 86; Response rate: 51%. State: Louisiana; Number of respondents: 79; Number of eligible surveyors: 134; Response rate: 59%. State: Maine; Number of respondents: 24; Number of eligible surveyors: 28; Response rate: 86%. State: Maryland; Number of respondents: 29; Number of eligible surveyors: 47; Response rate: 62%. State: Massachusetts; Number of respondents: 39; Number of eligible surveyors: 88; Response rate: 44%. State: Michigan; Number of respondents: 50; Number of eligible surveyors: 80; Response rate: 63%. State: Minnesota; Number of respondents: 58; Number of eligible surveyors: 85; Response rate: 68%. State: Mississippi; Number of respondents: 21; Number of eligible surveyors: 35; Response rate: 60%. State: Missouri; Number of respondents: 175; Number of eligible surveyors: 192; Response rate: 91%. State: Montana; Number of respondents: 20; Number of eligible surveyors: 21; Response rate: 95%. State: Nebraska; Number of respondents: 26; Number of eligible surveyors: 33; Response rate: 79%. State: Nevada; Number of respondents: 21; Number of eligible surveyors: 25; Response rate: 84%. State: New Hampshire; Number of respondents: 9; Number of eligible surveyors: 15; Response rate: 60%. State: New Jersey; Number of respondents: 35; Number of eligible surveyors: 77; Response rate: 45%. State: New Mexico; Number of respondents: 18; Number of eligible surveyors: 25; Response rate: 72%. State: New York; Number of respondents: 108; Number of eligible surveyors: 250; Response rate: 43%. State: North Carolina; Number of respondents: 53; Number of eligible surveyors: 84; Response rate: 63%. State: North Dakota; Number of respondents: 12; Number of eligible surveyors: 16; Response rate: 75%. State: Ohio; Number of respondents: 77; Number of eligible surveyors: 145; Response rate: 53%. State: Oklahoma; Number of respondents: 63; Number of eligible surveyors: 92; Response rate: 68%. State: Oregon; Number of respondents: 31; Number of eligible surveyors: 45; Response rate: 69%. State: Rhode Island; Number of respondents: 20; Number of eligible surveyors: 27; Response rate: 74%. State: South Carolina; Number of respondents: 15; Number of eligible surveyors: 27; Response rate: 56%. State: South Dakota; Number of respondents: 20; Number of eligible surveyors: 22; Response rate: 91%. State: Tennessee; Number of respondents: 52; Number of eligible surveyors: 79; Response rate: 66%. State: Texas; Number of respondents: 201; Number of eligible surveyors: 281; Response rate: 72%. State: Utah; Number of respondents: 16; Number of eligible surveyors: 25; Response rate: 64%. State: Vermont; Number of respondents: 11; Number of eligible surveyors: 16; Response rate: 69%. State: Virginia; Number of respondents: 36; Number of eligible surveyors: 42; Response rate: 86%. State: Washington; Number of respondents: 60; Number of eligible surveyors: 85; Response rate: 71%. State: West Virginia; Number of respondents: 14; Number of eligible surveyors: 22; Response rate: 64%. State: Wisconsin; Number of respondents: 66; Number of eligible surveyors: 83; Response rate: 80%. State: Wyoming; Number of respondents: 10; Number of eligible surveyors: 10; Response rate: 100.0%. State: Total; Number of respondents: 2,340; Number of eligible surveyors: 3,819; Response rate: 61%. Source: GAO. [End of table] Questionnaire for State Agency Directors: The questionnaire for state agency directors was designed to gather information on the nursing home survey process in each state. Directors were asked many of the same questions as the surveyors, but the survey agency directors' questionnaire contained additional questions on the overall organization of the survey agency, resource and staffing issues, CMS's Quality Indicator Survey (QIS), and experience with CMS's federal monitoring surveys.[Footnote 79] In addition, the questionnaire for state agency directors asked them to rank the degree to which several factors, derived from our previous work, contributed to understatement.[Footnote 80] This questionnaire was fielded from September to November 2008 to all 50 state survey agency directors and the survey agency director for the District of Columbia. We received completed responses from 50 of 51 survey agency directors, for a 98 percent response rate. The District of Columbia survey agency director did not respond. Analysis of Responses: To analyze results from the survey questions among groups, we used standard descriptive statistics. In addition, we looked for associations between questions through correlations and tests of the differences in means for groups. For certain open-ended questions, we used a standard content review method to identify topics that respondents mentioned such as "applying CMS guidance," "on-the-job training," "time to complete survey onsite," or "time to complete the survey paperwork." Our coding process involved one independent coder and an independent analyst who verified a random sample of the coded comments. For open-ended questions that enabled respondents to provide additional general information, we used similar standard content review methods, including independent coding by two raters who resolved all disagreements through discussion. Validity and Reliability of Data: In addition to the precautions taken during the development of the questionnaires, we performed automated checks on these data to identify inappropriate answers. We also reviewed the data for missing or ambiguous responses.[Footnote 81] Where comments on open-ended questions provided more detail or contradicted responses to categorical questions, the latter were corrected. On the basis of the strength of our systematic survey processes and follow-up procedures, we determined that the questionnaire responses were representative of the experience and perceptions of nursing home surveyors and state agency directors nationally and at the state level, with the exception of Pennsylvania surveyors and the survey agency director of the District of Columbia. On the basis of the response rates and these activities, we determined that the data were sufficiently reliable for our purposes. We also interviewed directors and other state agency officials in eight states to better understand unusual or interesting circumstances related to surveyor workforce and training, supervisory review, or state policies and practices. We selected these eight states based on our analysis of questionnaire responses from state agency directors and nursing home surveyors. Analysis of Federal Comparative and Observational Surveys: We used information from our May 2008 report on federal comparative surveys nationwide for fiscal years 2002 through 2007 to categorize states into groups.[Footnote 82] We used these results to identify states with high and low percentages of serious missed deficiencies. [Footnote 83] We classified nine states as high-understatement states-- those that had 25 percent or more federal comparative surveys identifying at least one missed deficiency at the actual harm or immediate jeopardy levels across all years. These states were Alabama, Arizona, Missouri, New Mexico, Oklahoma, South Carolina, South Dakota, Tennessee, and Wyoming. Zero-understatement states were those that had no federal comparative surveys identifying missed deficiencies at the actual harm or immediate jeopardy levels. These seven states were Alaska, Idaho, Maine, North Dakota, Oregon, Vermont, and West Virginia. Low-understatement states were the 10 with the lowest percentage of missed serious deficiencies (less than 6 percent)--Arkansas, Nebraska, Ohio, and all seven zero-understatement states. Response rates among the high-, low-, and zero-understatement states-- approximately 77, 62, and 71 percent, respectively--supported statistical testing of associations and differences among these state groupings. Therefore, in addition to descriptive statistics, we used correlations and tests of the differences in means for groups to identify questionnaire responses that were associated with differences in understatement.[Footnote 84] We reported the statistically significant results for tests of association and differences between group averages at the 5 percent level, unless otherwise noted. In a previous report, we found a possible relationship between the understatement of nursing home deficiencies on the federal comparative surveys and surveyor performance in General Investigation and Deficiency Determination on federal observational surveys--that is, high-understatement states more often had below-satisfactory ratings in General Investigation and Deficiency Determination than low- understatement states.[Footnote 85] For this report, we applied the same statistical analysis to identify when responses to our questionnaires were associated with satisfactory performance on General Investigative and Deficiency Determination skills on the federal observational surveys. We interpreted such relationships as an indication of additional training needs. Analysis of OSCAR: We used information from OSCAR and the federal monitoring survey databases to (1) compare the deficiencies cited by state and federal surveyors, (2) analyze the timing of nursing home surveys, and (3) assess trends in deficiency citations. OSCAR is a comprehensive database that contains information on the results of state nursing home surveys. CMS reviews these data and uses them to compute nursing home facility and state performance measures. When we analyzed these data, we included automated checks of data fields to ensure that they contain complete information. For these reasons, we determined that the OSCAR data were sufficiently reliable for our purposes. * We used OSCAR and the federal monitoring survey database to compare average facility citations on state survey records with the average citations on federal observational survey records for the same facilities during fiscal years 2002 through 2007.[Footnote 86] We computed the average number of serious deficiencies cited on federal observational surveys between fiscal years 2002 through 2007, and for the same facilities and time period, calculated the average number of serious deficiencies cited on state surveys. Next, we determined which facilities had greater average serious deficiency citations on federal observational surveys compared to state standard surveys between fiscal years 2002 through 2007. For these facilities, we computed the percentage difference between the average number of serious deficiencies cited on federal observational surveys and those cited on state surveys. * We used OSCAR to determine the percentage of the most recent state surveys that were predictable because of their timing. Our analysis of survey predictability compared the time between state agencies' current and prior standard nursing home surveys as of June 2008. According to CMS, states consider 9 months to 15 months from the last standard survey as the window for completing standard surveys because it yields a 12-month average. We considered surveys to be predictable if (1) homes were surveyed within 15 days of the 1-year anniversary of their prior survey or (2) homes were surveyed within 1 month of the maximum 15-month interval between standard surveys. * We calculated the number of serious deficiencies on state surveys in OSCAR from calendar year 1999 through 2007. We examined the trend in G- level and higher deficiencies to assess whether CMS's expanded enforcement policy appeared to affect citation rates. Effective January 2000, CMS completed the implementation of its immediate-sanctions policy, requiring the referral of homes that caused actual harm or immediate jeopardy on successive standard surveys or intervening complaint investigations. [End of section] Appendix II: Comments from the Department of Health & Human Services: Department Of Health & Human Services: Office Of The Secretary: Assistant Secretary for Legislation: Washington, DC 20201: October 30, 2009: John Dicken: Director, Health Care: U.S. Government Accountability Office: 441 G Street N.W. Washington, DC 20548: Dear Mr. Dicken: Enclosed are comments on the U.S. Government Accountability Office's (GAO) report entitled: "Nursing Homes: Addressing the Factors Underlying Understatement of Serious Care Problems Requires Sustained CMS and State Commitment" (GAO-10-70). The Department appreciates the opportunity to review this report before its publication. Sincerely, Signed by: Andrea Palm: Acting Assistant Secretary for Legislation: Enclosure: [End of letter] Department Of Health & Human Services: Centers for Medicare & Medicaid Services: Administrator: Washington, DC 20201: Date: October 30, 2009: To: Andrea Palm: Acting Assistant Secretary for Legislation: Office of the Secretary: From: [Signed by] Charlene Frizzera: Acting Administrator: Subject: Government Accountability Office (GAO) Draft Report: "Nursing Homes: Addressing the Factors Underlying Understatement of Serious Care Problems Requires Sustained CMS and State Commitment" (GA0-10-70): The Centers for Medicare & Medicaid Services (CMS) appreciates the opportunity to review and comment on the subject GAO Draft Report. The GAO was asked to examine four factors relevant to potential understatement of nursing home deficiencies including: 1. CMS' survey process; 2. Workforce shortages and training; 3. Supervisory reviews of surveys; and; 4. State agency practices. We believe that the GAO study adds value to the important public policy discussions regarding the survey process. The report derives its information primarily from opinion surveys of State surveyors and directors of State survey agencies. We particularly appreciate the GAO's effort to identify underlying causes of issues in the survey process and contribute ideas for solutions. We fully endorse five of the seven GAO recommendations. For the remaining two, we will convene a workgroup of State and Federal officials to explore alternate solutions that may be responsive to the intent of the GAO recommendations. We offer the following, more detailed responses to the GAO recommendations. GAO Recommendation 1: Quality Indicator Survey (OIS): To improve the survey process, CMS must make sure measures are taken to address issues identified with the new QIS methodology. CMS Response: We agree with this recommendation. We have designed the QIS specifically to improve consistency of survey processes and to provide both States and CMS with tools for continuous improvement. Since the tools for continuous improvement were not part of the original QIS design, their development has slowed QIS implementation somewhat, but the investment has been very worthwhile. A set of QIS "Desk Audit Reports" (DARs) represents one such tool. The DARs contain over 30 key investigative and decision-making factors for each QIS survey. The DARs also enable State survey supervisors to provide improved feedback to their surveyors, identify patterns in deficiency citation, and strengthen the consistency of the survey process between State surveyor teams. These tools provide information for the CMS regional offices and become part of a quarterly review teleconference in which each State can review the patterns of survey results together with the CMS regional office. Finally, we established a user group that convenes quarterly to continue to address QIS issues as they arise. This user group may also represent a useful forum to share ideas for methods to facilitate implementation and to generate additional ideas for refinements to the system or to the implementation process. For example, Florida has just completed its statewide conversion from the traditional survey to the QIS and made a total revision to their surveyor orientation program materials. We plan to share these materials with the other States, so they can upgrade their orientation efforts to more effectively address the investigation and deficiency determination skills of newly hired surveyors. GAO Recommendation 2: CMS Guidance for Surveyors: To improve guidance to surveyors, CMS should clarify and revise existing CMS written guidance to make it more concise, simples its application in the field and reduce confusion, particularly on the definition of actual harm. CMS Response: We will seek alternate methods to address these issues, and will work with a workgroup of State and Federal officials to do so. We agree that it is desirable to clarify any areas of Federal guidance that may be ambiguous, but do not agree that shortening the guidance is necessarily the preferable method of doing so, or that greater conciseness would have the desired effect. One method of striking a balance between full guidance and conciseness may be to create some short reference documents for use in the field that contain cross-links back to the full guidance. Based on previous GAO recommendations to ensure consistency, CMS embarked on a multi-year project to upgrade and clarify our interpretive guidelines for key regulations (Tags) in order to provide surveyors with accurate and up-to-date information in each topic area (such as nutrition, infection control, and incontinence, etc.). The guidance was developed with assistance from expert panels and is much more informative than previous guidance. We believe it is imperative for all surveyors to master the guidance and that the GAO's recommendation to bolster training efforts (recommendation #4) is the superior method, rather than seeking to shorten the guidance. Many survey agencies that the revised guidance has enhanced the ability of their surveyors to correctly investigate concerns and select the correct levels of severity for non-compliance, including making distinctions as to when "actual harm" (Level 3) is reached for common types of deficient practices that fall under the various Tag topics. An example of a positive impact from revised guidance (in addressing deficiency understatement issues) can be seen in CMS' issuance of improved pharmacy guidelines. The new guidance clearly improved surveyors' ability to identify the use of unnecessary medications. Widely publicized concerns about the use of unnecessary medications had previously failed to alter the extent to which such problems were identified by surveyors. CMS issued the expanded guidance in late 2006 and combined the issuance with national training of surveyors. The percent of standard surveys in which unnecessary' medications was identified by surveyors subsequently increased from the previously consistent rate of 13-14 percent in fiscal year (FY) 2000 through 2006 to 18 percent in FY 2007. These results are portrayed in Figure 1. Figure 1: Percent of Surveys Citing Unnecessary Drug Use: [Refer to PDF for image: vertical bar graph] Calendar years 2000 through 2007 plotted versus Percentage of Surveys 9% through 19%. [End of figure] GAO Recommendation 3: Establish a National Pool of Surveyors for States: To address surveyor workforce shortages, CMS should consider establishing a pool of additional national surveyors that could augment State survey teams experiencing workforce shortages. CMS Response: We do not plan to implement this recommendation, at least on any scale that would make a national difference. Section 1864 of the Social Security Act (the Act) directs the Secretary of the Department of Health and Human Services to enter into agreement with any State that is willing and capable of carrying out certain survey and certification responsibilities identified in the Act. According to the 1864 Agreement, the responsibility for conducting inspections and making recommendations for enforcement rests with the State survey agency. Establishing national survey teams to augment State surveys is problematic for a variety of reasons. First, it begins to blur the lines between (a) holding States accountable for meeting performance expectations versus (b) compensating for problematic performance due to State management decisions. While it is proper for CMS to set survey performance expectations and to establish qualifications for surveyor knowledge and training, we believe it is improper to tell States how to make personnel decisions, establish pay scales, recruit staff or, hire staff, or for CMS to try to alleviate State performance problems that arise because of State personnel management decisions. It is incumbent upon each State to determine alternate methods to fulfill the terms of the 1864 Agreement; it is not up to CMS to compensate for those State decisions. While we have national contractors for some provider types (such as organ transplant centers and psychiatric hospitals), the use of national contractors has generally been limited to areas of specialty knowledge. The number of providers in many States may be so small that it is more cost-effective to have a national contractor than to contract with States and seek to maintain specialty surveyors for infrequent surveys. In those instances, we have offered the contracting to States and utilized a national contractor where States declined the offer. However, we regard as questionable the significant expansion of the use of national contractors to perform the more frequent nursing home surveys, in order to compensate for State performance issues. We will explore this issue with our State-Federal workgroup in more detail in order to identify any circumstances in which a national pool may be advisable, and to identify any additional solutions. In the past, CMS has examined several promising practices with States and published policy briefs to assist States on issues such as surveyor recruitment strategies, retaining surveyors, and strategies to promote consistent surveyor performance. GAO Recommendation 4: Training of Surveyors: To address insufficient training, CMS should (a) evaluate the current training programs and division of responsibility between Federal and State components to determine the most cost-effective approach; and (b) support the continuing education of experienced surveyors. CMS Response: We agree with this recommendation. Training is critically important in fulfilling the mission of the survey and certification program. We have made significant investments to increase: (a) the number and types of CMS surveyor training courses, (b) the use of distance learning (satellites and Web-based training), and (c) the accessibility of training in geographical areas with large numbers of surveyors ("magnet areas"). To increase the accessibility of training, we successfully inaugurated "Magnet Area Training (MAT)" in Florida and California. Evaluations of the training showed that MAT instruction produced results comparable to our other training events. We also initiated Web-based training (WBT) for one course in FY 2008, and we continue to develop new training courses that employ Web technology. These efforts have improved the survey and certification training profile. Figure 2, for example, shows the results of our annual survey of State training coordinators. The percent of respondents who indicated they were not satisfied with the number of CMS training courses declined from 30 percent in FY 2005 to 6 percent in FY 2008. The percent of respondents indicating they were satisfied increased from 30 percent in FY2005 to 41 percent in FY 2008. Figure 2: Percent State Training Coordinators Not Satisfied with the Number of CMS Training Courses: [Refer to PDF for image: vertical bar graph] FY 2005: 30%; FY 2006: 10%; FY 2008: 6%. [End of figure] While the trendline is very positive, the FY 2008 results still left 53 percent of participants indicating that they were only "somewhat satisfied." Gaps remain in both the number and types of courses. As part of our ongoing dialogue with States, we will (a) poll States to complete a needs assessment for the continuing education of experienced surveyors; (h) offer selected continuing education opportunities based on the needs assessment; and (c) continue to expand the above initiatives. GAO Recommendation 5: State Supervisory Reviews: To address inconsistencies in State supervisory reviews, CMS should set an expectation through guidance that States have a supervisory review program as a part of their Quality Assurance processes that includes routine reviews of deficiencies at the level of potential for more than minimal harm (1)-F) and that provides feedback to surveyors regarding changes made to citations. CMS Response: We agree with the recommendation. We will start by working with a State- Federal workgroup to identify promising practices, and initiate the process of setting more defined expectations for quality review. Historically, CMS has set forth expectations for documentation and for quality of the surveys, however, ensuring that deficiency determination and severity selection are consistent and accurate has been left to State management. We believe many States have developed effective programs or methods of supervisory review. Such methods include offsite review of citations and/or at least occasional participation onsite of a supervisor during deficiency decision-making meetings of surveyors, in order to coach and correct mistaken processes used by survey teams. In addition, CMS regional offices conduct validation reviews of State surveys and, particularly through the follow-along surveys, provide feedback to the State survey teams. We will build on these existing processes and identify additional steps that will be taken. GAO Recommendation 6: Guidance on Non-Citation Practices: To address State agency practices and external pressure, CMS should reestablish expectations through guidance to State survey agencies that non-citation practices—official or unofficial—are inappropriate and systematically monitor trends in States' citations. CMS Response: We agree with this recommendation. We will issue a Survey and Certification policy letter outlining our expectations regarding the survey and certification program as well as avenues to pursue if inappropriate practices are applied. GAO Recommendation 7: Guidance on Communicating with CMS when External Pressures Exist: To address State agency practices and external pressure, CMS should establish expectations through guidance to State survey agencies to communicate and collaborate with their CMS Regional offices when they experience significant pressure from legislators or the nursing home industry that may have an impact on the survey process or surveyors' perceptions. CMS Response: We agree with this recommendation. We will provide guidance for communicating surveyor concerns and provide feedback to State surveyors outlining our expectations regarding the survey and certification program, including avenues to pursue if inappropriate pressures are applied. We appreciate the effort that went into this report and look forward to working with the GAO on this and other issues. [End of section] Appendix III: GAO Contact and Staff Acknowledgments: GAO Contact: John E. Dicken, (202) 512-7114 or dickenj@gao.gov: Staff Acknowledgments: In addition to the contact named above, Walter Ochinko, Assistant Director; Stefanie Bzdusek; Leslie V. Gordon; Martha R. W. Kelly; Katherine Nicole Laubacher; Dan Lee; Elizabeth T. Morrison; Dan Ries; Steve Robblee; Karin Wallestad; Rachael Wojnowicz; and Suzanne Worth made key contributions to this report. [End of section] Related GAO Products: Nursing Homes: Opportunities Exist to Facilitate the Use of the Temporary Management Sanction. [hyperlink, http://www.gao.gov/products/GAO-10-37R]. Washington, D.C.: November 20, 2009. Nursing Homes: CMS's Special Focus Facility Methodology Should Better Target the Most Poorly Performing Homes, Which Tended to Be Chain Affiliated and For-Profit. [hyperlink, http://www.gao.gov/products/GAO-09-689]. Washington, D.C.: August 28, 2009. Medicare and Medicaid Participating Facilities: CMS Needs to Reexamine Its Approach for Funding State Oversight of Health Care Facilities. [hyperlink, http://www.gao.gov/products/GAO-09-64]. Washington, D.C.: February 13, 2009. Nursing Homes: Federal Monitoring Surveys Demonstrate Continued Understatement of Serious Care Problems and CMS Oversight Weaknesses. [hyperlink, http://www.gao.gov/products/GAO-08-517]. Washington, D.C.: May 9, 2008. Nursing Home Reform: Continued Attention Is Needed to Improve Quality of Care in Small but Significant Share of Homes. [hyperlink, http://www.gao.gov/products/GAO-07-794T]. Washington, D.C.: May 2, 2007. Nursing Homes: Efforts to Strengthen Federal Enforcement Have Not Deterred Some Homes from Repeatedly Harming Residents. [hyperlink, http://www.gao.gov/products/GAO-07-241]. Washington, D.C.: March 26, 2007. Nursing Homes: Despite Increased Oversight, Challenges Remain in Ensuring High-Quality Care and Resident Safety.[hyperlink, http://www.gao.gov/products/GAO-06-117]. Washington, D.C.: December 28, 2005. Nursing Home Quality: Prevalence of Serious Problems, While Declining, Reinforces Importance of Enhanced Oversight. [hyperlink, http://www.gao.gov/products/GAO-03-561]. Washington, D.C.: July 15, 2003. Nursing Homes: Quality of Care More Related to Staffing than Spending. [hyperlink, http://www.gao.gov/products/GAO-02-431R]. Washington, D.C.: June 13, 2002. Nursing Homes: Sustained Efforts Are Essential to Realize Potential of the Quality Initiatives. [hyperlink, http://www.gao.gov/products/GAO/HEHS-00-197]. Washington, D.C.: September 28, 2000. Nursing Home Care: Enhanced HCFA Oversight of State Programs Would Better Ensure Quality. [hyperlink, http://www.gao.gov/products/GAO/HEHS-00-6]. Washington, D.C.: November 4, 1999. Nursing Homes: Proposal to Enhance Oversight of Poorly Performing Homes Has Merit. [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-157]. Washington, D.C.: June 30, 1999. Nursing Homes: Additional Steps Needed to Strengthen Enforcement of Federal Quality Standards. [hyperlink, http://www.gao.gov/products/GAO/HEHS-99-46]. Washington, D.C.: March 18, 1999. California Nursing Homes: Care Problems Persist Despite Federal and State Oversight. [hyperlink, http://www.gao.gov/products/GAO/HEHS-98-202]. Washington, D.C.: July 27, 1998. [End of section] Footnotes: [1] Medicare is the federal health care program for elderly and certain disabled individuals. Medicaid is a joint federal-state health care financing program for certain categories of low-income individuals. [2] In addition to the oversight of nursing homes, CMS and state survey agencies are responsible for oversight of other Medicare and Medicaid providers, such as home health agencies, intermediate care facilities for the mentally retarded, and hospitals. [3] See a list of related GAO products at the end of this report. [4] See GAO, Nursing Homes: Federal Monitoring Surveys Demonstrate Continued Understatement of Serious Care Problems and CMS Oversight Weaknesses, [hyperlink, http://www.gao.gov/products/GAO-08-517] (Washington, D.C.: May 9, 2008). [5] CMS's Survey and Certification Group is responsible for ensuring the effectiveness of state survey activities and managing the federal monitoring survey program. [6] See GAO, Nursing Home Quality: Prevalence of Serious Problems, While Declining, Reinforces Importance of Enhanced Oversight, [hyperlink, http://www.gao.gov/products/GAO-03-561] (Washington, D.C.: July 15, 2003) and GAO, Nursing Home Reform: Continued Attention Is Needed to Improve Quality of Care in Small but Significant Share of Homes, [hyperlink, http://www.gao.gov/products/GAO-07-794T] (Washington, D.C.: May 2, 2007). In response to our recommendation to finalize the development, testing, and implementation of a more rigorous survey methodology, CMS evaluated and is currently implementing a revised survey methodology. [7] See GAO, Nursing Homes: Despite Increased Oversight, Challenges Remain in Ensuring High-Quality Care and Resident Safety, [hyperlink, http://www.gao.gov/products/GAO-06-117] (Washington, D.C.: Dec. 28, 2005). [8] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. Our analysis of survey predictability considered surveys to be predictable if (1) homes were surveyed within 15 days of the 1-year anniversary of the prior survey or (2) homes were surveyed within 1 month of the maximum 15-month interval between standard surveys. We used this rationale because homes know the maximum allowable interval between surveys, and those whose prior surveys were conducted 14 or 15 months earlier are aware that they are likely to be surveyed soon. [9] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [10] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [11] Eligible surveyors are those who had conducted at least one health survey of a nursing home in fiscal year 2006 or 2007 and for whom we could obtain an e-mail or other address from their state survey agency. [12] We excluded Pennsylvania from our analysis because Pennsylvania's Deputy Secretary for Quality Assurance instructed the state's surveyors not to respond to our survey. Two other states had response rates below 40 percent--Connecticut (28 percent), and Illinois (20 percent). Illinois' response rate probably reflected that surveyors' access to their e-mail accounts and our Web-based survey was limited to only 1 day per month. [13] The District of Columbia agency director did not respond to our questionnaire. [14] We did not ask nursing home surveyors a similar question because survey agency directors, as a result of their positions, were a more consistent source of knowledge about the influence of these factors on understatement. [15] See [hyperlink, http://www.gao.gov/products/GAO-08-517]. This database captures the results of two types of federal monitoring surveys. Federal comparative surveys are conducted independently by federal surveyors to evaluate state surveys. Federal surveyors resurvey a home that was recently inspected by state surveyors and compare the deficiencies identified during the two surveys. When federal surveyors accompany state surveyors to directly observe them during a nursing home survey it is considered a federal observational survey. [16] We use the term survey record to refer to CMS's Form 2567, which is the official statement of deficiencies with respect to federal quality standards. [17] Other areas include Admission, Transfer and Discharge Rights, Resident Rights, Resident Behavior and Facility Practices, Nursing Services, Pharmacy Services, Dietary Services, Physician Services, Specialized Rehabilitative Services, Dental Services, Infection Control, and Physical Environment. Surveys also examine compliance with federal fire safety requirements. [18] Revisits are not required for most deficiencies cited below the actual-harm level--that is, A through F. [19] Nursing homes can also appeal deficiency citations, which result in hearings before an administrative law judge; nursing homes may also request HHS's Departmental Appeals Board to review. [20] On-site sources include observations, interviews, and records review. [21] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [22] The federal government funds state surveys through the Medicare and Medicaid programs. States contribute a share of Medicaid and non- Medicaid funds to support survey activities. State non-Medicaid contributions are to reflect the benefit states derive from health care facilities that meet federal quality standards as well as the cost of assessing compliance with state licensing requirements. See GAO, Medicare and Medicaid Participating Facilities: CMS Needs to Reexamine Its Approach for Funding State Oversight of Health Care Facilities, [hyperlink, http://www.gao.gov/products/GAO-09-64] (Washington, D.C.: Feb. 13, 2009). [23] See [hyperlink, http://www.gao.gov/products/GAO-09-64]. [24] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [25] CMS commented on the importance of quality-assurance processes and noted it had already incorporated such reviews into CMS regional offices' reviews of the state performance standards. However, the agency did not require states to initiate an ongoing process that would evaluate the appropriateness of the scope and severity of documented deficiencies, as we recommended. See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [26] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [27] In addition to the federal monitoring surveys, CMS established annual state performance reviews in fiscal year 2001 to measure a state's compliance with specific standards. These standards generally focus on the timeliness and quality of surveys, complaint investigations, and enforcement actions. CMS's state performance reviews include (1) an examination of the quality of state survey agency investigations and decision making and (2) the timeliness and quality of complaint investigations. [28] In 1998, the Health Care Financing Administration, the HHS agency now known as CMS, acknowledged the need to perform a greater number of comparative surveys and have done so. Between October 1998 and July 1999 only about 9 percent (64) of federal monitoring surveys were comparative. However, in our May 2008 report, we found that for the period of fiscal years 2002 through 2007 about 20 percent (976) of federal monitoring surveys were comparative surveys and the remaining 80 percent were observational surveys. By statute, comparative surveys must be conducted within 2 months of the completion of the state survey. [29] CMS began requiring regional offices to make this determination in fiscal year, 2002 and it is captured by a yes/no validation question. [30] See [hyperlink, http://www.gao.gov/products/GAO-08-517]. [31] In May 2008, we found that understatement also occurred when state survey teams cited deficiencies at too low a level of scope and severity. At that time, CMS did not require federal surveyors to evaluate scope and severity differences between state and federal comparative surveys. However, as of October 2008, CMS began requiring such assessments. [32] See [hyperlink, http://www.gao.gov/products/GAO-08-517]. [33] For purposes of this report, we defined the federal nursing home survey process as both the traditional methodology used to evaluate compliance of nursing homes with federal requirements and the written guidance provided by CMS to help state agencies carry out survey activities. [34] Survey methodology is defined as the traditional approach used to evaluate nursing home compliance with federal regulations as outlined by CMS in Appendix P of the SOM. [35] For purposes of this report, we defined CMS written guidance as the information in the SOM on the long-term care survey process, including the survey protocol for long-term care facilities in Appendix P and the guidance on federal quality standards in Appendix PP as well as any additional materials provided by CMS to assist surveyors, such as Survey and Certification letters. [36] Guidance for determining actual-harm level deficiencies is provided in Section IV, Appendix P of the SOM. [37] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [38] Federal nursing home quality standards detail requirements for the delivery of care, resident outcomes, and facility conditions. State survey teams use these federal quality standards to assess compliance during state nursing home surveys. [39] In October 2000, CMS began revising investigative protocols for assessing specific deficiencies. The intent of this initiative is to enable surveyors to better (1) identify specific deficiencies, (2) investigate whether a deficiency is the result of poor care, and (3) document the level of harm resulting from a home's identified deficient care practices. See [hyperlink, http://www.gao.gov/products/GAO-03- 561]. [40] Our questionnaire included 13 topics of the approximately 200 federal quality standards. Seven of these were taken from the Quality of Care category of federal quality standards, the others originated from different categories such as Resident Assessment and Dietary Services. See [hyperlink, http://www.gao.gov/products/GAO-08-517]. [41] Beginning January 1, 1999, CMS directed states to avoid scheduling surveys for the same month of the year as a nursing home's previous survey. However, surveys can also be considered predictable if occurring at a time other than near the 1-year anniversary or 15-month maximum date. For example, nursing home operators could be alerted when the state agency is surveying a facility in a nearby area if all the facilities in that area were surveyed at about the same time. [42] In 2003, we found that 34 percent of nursing home surveys were predictable. See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [43] According to CMS, states consider 9 months to 15 months from the last standard survey as the window for completing standard surveys because it yields a 12-month average. Thus, to maintain an average survey interval of 12 months, given that some facilities are not surveyed until near or after 15 months, more surveys would need to occur within 9 months of the last standard survey. See GAO-06-117. [44] At the time of our survey, the QIS methodology was being implemented in eight state agencies: Connecticut, Florida, Kansas, Louisiana, Minnesota, New Mexico, North Carolina, and Ohio. According to a CMS official, Connecticut is the only one of these states that has implemented the QIS statewide. As of May 2008, CMS projected that the QIS would be fully implemented in all states in 2014. [45] The QIS evaluation was conducted to answer questions about accuracy, documentation, changes in the number and types of deficiencies, and whether the QIS process is more efficient. HHS, CMS, Evaluation of the Quality Indicator Survey (QIS), Final Report (December 2007), [hyperlink, http//:www.cms.hhs.gov/CertificationandComplianc/Downloads/QISExecSummar y.pdf] (accessed July 17, 2009). [46] Scope refers to the number of residents potentially or actually affected and has three levels--isolated, pattern, and widespread. A pattern scope refers to deficiencies at the B, E, H, and K levels and a widespread scope refers to deficiencies at the C, F, I, and L levels. [47] We asked all 42 directors who had not participated in the QIS to provide their opinions on the new methodology; we received comments from 18 of the 42 directors. [48] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [49] Virginia did not provide the information needed to compute a vacancy rate. [50] Michigan and Illinois did not provide this information. [51] See [hyperlink, http://www.gao.gov/products/GAO-09-64]. [52] This information was last updated in June 2009 before the governor of California signed the state's budget revisions. [53] Seven states did not report the number of surveyors with less than 2 years of experience--Illinois, Michigan, Minnesota, Texas, Washington, West Virginia, and Wisconsin. [54] Revisit surveys are generally conducted in facilities when a G- level or higher deficiency is cited by a survey team, to verify that serious deficiencies have been corrected by the home. [55] In an open-ended question at the end of the questionnaire, 842 surveyors commented on a wide range of topics related to surveys. These comments represented about 36 percent of all nursing home surveyor respondents; 15 percent of the comments represented about 6 percent of all respondents. [56] The 476 surveyors who responded to this open-ended question about training needs constituted about 20 percent of all respondents. [57] For supervisory review processes, we defined direct-line supervisors as including survey team leaders, direct supervisors, and supervisors at district or regional offices. Central state agency staff was defined as including quality assurance teams, legal counsel, state training coordinators, and compliance specialists. [58] These review steps could be done at the direct-line supervisor level or by central state agency staff. The difference between supervisory review levels for surveys with J through L citations and those for surveys with D through F citations was significant at the 1 percent level. [59] Forty states review a sample of all draft surveys. Such reviews may include additional examination of surveys with deficiencies at either the D through F or J through L levels. [60] See [hyperlink, http://www.gao.gov/products/GA0-03-561]. [61] Seventy-four percent of state agency directors indicated that inadequate supervisory review processes infrequently or never contributed to understatement in their state, and 4 percent of state agency directors were unsure or had no opinion on this topic. [62] CMS previously identified the existence of a potential noncitation practice in one state that had an unusually high number of homes with no deficiencies on their standard surveys. Contrary to federal guidance, surveyors in that state were not citing all identified deficiencies but rather brought them to the homes' attention with the expectation that the deficiencies would be corrected. See GAO-03-561. [63] CMS requires on-site revisits for any noncompliance identified at level F (with a finding of substandard quality of care) or any level higher than F. [64] Effective January 2000, CMS expanded its immediate sanctions policy, requiring referral of homes found to have harmed one or a small number of residents (G-level deficiencies) on successive routine surveys or intervening complaint investigations. [65] CMS officials previously acknowledged that the double G policy may have had an unintended negative effect on the rate of deficiency citations. See GAO, Nursing Homes: Efforts to Strengthen Federal Enforcement Have Not Deterred Some Homes from Repeatedly Harming Residents, [hyperlink, http://www.gao.gov/products/GAO-07-241] (Washington, D.C.: Mar. 26, 2007). [66] In this section, we refer to two states we interviewed as State A and State B to maintain confidentiality for the officials from these state agencies. The corresponding regional offices are referred to as regional office responsible for State A or State B, respectively. [67] Alaska, Hawaii, Illinois, and West Virginia did not report information on the number of IDRs. [68] The following states did not report the number of deficiencies deleted or downgraded through the IDR process: Alaska, Hawaii, Illinois, Kentucky, Nebraska, New Jersey, New Mexico, Vermont, or West Virginia. Maine, Oklahoma, Utah, Washington, and Wyoming provided the number of deficiencies deleted but not the number that were downgraded. [69] According to CMS guidance, if an outside entity conducts the IDR, the results of the process may serve only as a recommendation to the state survey agency of noncompliance or compliance with the federal requirements for nursing homes. [70] See [hyperlink, http://www.gao.gov/products/GAO-03-561]. [71] See [hyperlink, http://www.gao.gov/products/GAO-08-517]. [72] This report follows and expands on our May 2008 report, which examined (1) the information contained in federal monitoring surveys about understatement nationwide, and (2) CMS management and oversight of the federal monitoring survey program, see GAO-08-517. [73] CMS's Survey and Certification Group is responsible for ensuring the effectiveness of state survey activities and managing the federal monitoring survey program. [74] We mailed paper copies of the questionnaire to 15 surveyors in Arkansas, who did not have a state-issued e-mail address; on request, an additional copy was faxed to a surveyor. Seven out of the 16 paper copies were completed and returned to GAO. [75] Although nursing home surveyors' responses were anonymous to preserve their confidentiality, a few surveyors voluntarily provided their contact information and agreed to be interviewed. [76] Questions about CMS's survey methodology directed surveyors to respond about the traditional survey methodology, not the new Quality Indicator Survey (QIS) methodology, which had been implemented in eight states. [77] When respondents indicated that they did not conduct health safety surveys of nursing homes and therefore should have been excluded from the population of eligible nursing home surveyors, these surveyors and their responses were excluded. [78] The Illinois response rate likely reflects that surveyors' access to their e-mail accounts, and our Web-based survey, was limited to only 1 day per month. [79] All state agency directors were asked about CMS's traditional survey methodology, which all states used in 2008. However eight state agency directors, who indicated that the QIS has been implemented in at least part of their states, were asked additional questions specifically about the QIS. [80] We did not ask nursing home surveyors a similar question because survey agency directors, as a result of their positions, were a more consistent source of knowledge about the influence of these factors on understatement. [81] Where responses to particular questions were fewer than the overall number of responses for the questionnaire, this limitation is indicated in the text. [82] During this period, fiscal year 2002 was the first year that the database contained all the information needed to assess the results of federal comparative surveys. See [hyperlink, http://www.gao.gov/products/GAO-08-517]. [83] Federal comparative surveys are done on a small group of facilities that are not randomly selected, and the understatement of deficiencies identified through comparative surveys are not representative of all nursing home surveys or survey teams within each state. [84] For our descriptive statistics, we computed means, the minimum and maximum responses, responses at the 25th, 50th, and 75th percentiles, frequencies among categories of respondents, such as those from high- and low-understatement states, as well as frequencies across two or more categories of respondents. Correlations were computed as Pearson's correlations of association. T-tests were done to identify when the mean response from two different categories of respondents, such as high-and low-understatement states, were significantly different from each other. [85] [hyperlink, http://www.gao.gov/products/GAO-08-517]. The General Investigation segment assesses the effectiveness of state survey team actions such as collection of information, discussion of survey observations, interviews with nursing home residents, and implementation of CMS investigative protocols. The Deficiency Determination segment evaluates the skill with which the state survey teams (1) integrate and analyze all information collected and (2) use the guidance for surveyors and identify deviations from regulatory requirements. [86] We use the term survey record to refer to CMS's Form 2567, which is the official statement of deficiencies with respect to federal quality standards. [End of section] GAO's Mission: The Government Accountability Office, the audit, evaluation and investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO's commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of GAO Reports and Testimony: The fastest and easiest way to obtain copies of GAO documents at no cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each weekday, GAO posts newly released reports, testimony, and correspondence on its Web site. To have GAO e-mail you a list of newly posted products every afternoon, go to [hyperlink, http://www.gao.gov] and select "E-mail Updates." Order by Phone: The price of each GAO publication reflects GAO’s actual cost of production and distribution and depends on the number of pages in the publication and whether the publication is printed in color or black and white. Pricing and ordering information is posted on GAO’s Web site, [hyperlink, http://www.gao.gov/ordering.htm]. Place orders by calling (202) 512-6000, toll free (866) 801-7077, or TDD (202) 512-2537. Orders may be paid for using American Express, Discover Card, MasterCard, Visa, check, or money order. Call for additional information. To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: E-mail: fraudnet@gao.gov: Automated answering system: (800) 424-5454 or (202) 512-7470: Congressional Relations: Ralph Dawn, Managing Director, dawnr@gao.gov: (202) 512-4400: U.S. Government Accountability Office: 441 G Street NW, Room 7125: Washington, D.C. 20548: Public Affairs: Chuck Young, Managing Director, youngc1@gao.gov: (202) 512-4800: U.S. Government Accountability Office: 441 G Street NW, Room 7149: Washington, D.C. 20548: