Changes to Nursing Home Compare: Some Good, Some Bad

The Centers for Medicare & Medicaid Services (CMS) announced plans to expand its focused surveys on resident assessments and nurse staffing for nursing facilities nationwide (but not in all facilities), beginning in early fiscal year 2015.  Expanded surveys should lead to more accurate reporting of quality measures and staffing data on the federal website Nursing Home Compare.  At the same time, however, CMS is currently providing misleading information regarding health survey deficiencies, the third domain on the Nursing Home Compare website. 

Expanded Focused Surveys

The New York Times reported on August 25, 2014 that nursing facilities "game" Nursing Home Compare by self-reporting resident assessment and staffing information that inflates their star ratings on the quality measure and staffing domains of the federal website.[1]  The report led to White House and CMS announcements on October 6, 2014 that CMS would make six changes to improve Nursing Home Compare, including conducting focused surveys nationwide, beginning in January 2015, and initiating the process of implementing the statutory requirement to publish nurse staffing levels based on payroll-based staffing information.[2]

As discussed below, CMS described the new focused surveys in an October 31, 2014 Survey & Certification Letter sent to State Survey Agency Directors.[3]

            Expanded Focused Surveys: Resident Assessments

CMS reports that in mid-2014, it pilot-tested a short-term focused survey to evaluate resident assessments, called minimum data set (MDS) 3.0, "coding practices and its relationship to resident care in nursing homes in five states."  The focused surveys, completed in August 2014, identified deficiencies in 24 of 25 pilot surveys, including "inaccurate staging and documentation of pressure ulcers, lack of knowledge regarding the classification of antipsychotic drugs, and poor coding regarding the use of restraints."

CMS's Survey & Certification Letter cites reports by the HHS Office of Inspector General (OIG) that found problems in three assessment-related areas: care planning and discharge planning,[4] assessment errors (in 99.5% of a sample of assessments and care plans for residents given atypical antipsychotic drugs),[5] and inaccuracies on at least one MDS item in 47% of claims reviewed.[6]  CMS writes that these reports "reinforced CMS' conclusion that additional steps should be put in place to support the resident assessment process and accurate codification of information in the MDS." 

CMS has been concerned about the accuracy of MDS data for many years, particularly since the 1998 implementation of the Medicare prospective payment system that uses resident assessment information to set Medicare reimbursement rates.  In 1998, CMS contracted with Abt Associates "to develop and test various on-site and off-site approaches for verifying and improving the accuracy of MDS data."[7]  CMS's "contractor found widespread MDS errors that resulted in a change in Medicare payment categories for 67 percent of the resident assessments sampled."[8]  In 2001, CMS implemented "a nationwide MDS review program" called the Data Assessment and Verification Project (DAVE), which continued until 2007.[9]

            Expanded Focused Surveys: Nurse Staffing

In addition to improving the accuracy of resident assessments, CMS reports in the Survey & Certification Letter that it identified a similar need to assess "the accuracy of information on the staffing of nursing homes."  Some or all of the focused surveys will assess staffing.  CMS explains, "This assessment will aim to verify the data self-reported by the nursing home, and identify changes in staffing levels throughout the year."

At present, during annual health surveys, facilities complete a standard survey form, CMS-671,[10] which lists "the type of staff working in the facility and the number of hours they worked" in "the most recent complete pay period" or the last 14 days.  These forms are not audited for accuracy, but their information is reported on Nursing Home Compare. 

Improvements in reporting staffing information are mandated by §6106 of the Affordable Care Act, which required facilities, by March 2012, to submit, electronically to HHS, direct care staffing information (including agency and contract staff), "based on payroll and other verifiable and auditable data in a uniform format."  Congress provided funding for §6106 in the Improving Medicare Post-Acute Care Transformation Act of 2014.

The inaccuracy of nurse staffing levels reported on Nursing Home Compare is highlighted by a new analysis by the Center for Public Integrity, which reported "widespread discrepancies in staffing levels reported by nursing homes."[11]  On November 12, 2014, the Center for Public Integrity reported, "More than 80 percent of [more than 10,000] facilities reported higher registered nurse staffing levels" on Nursing Home Compare than the investigative report calculated from Medicare cost reports.[12]  In addition, the investigative report found, "In more than 25 percent of nursing homes nationwide, the listed amount on Compare was at least double the level in the cost reports."[13]  Overstated staffing levels were part of The New York Times' August analysis of problems in Nursing Home Compare. 

            CMS Plans for MDS/Staffing Focused Surveys

CMS will provide states with supplemental Medicare funds to conduct the focused surveys, which, like the pilot surveys, will be expected to last two days, with two surveyors.  Although CMS plans mandatory training for state agency staff and on-going phone and email support during surveys, many details remain undecided, including how many surveys will be conducted and when, which facilities will be surveyed, and the survey protocol and tool to be used.

Survey Deficiencies on Nursing Home Compare

The health survey is the only part of Nursing Home Compare that has been based on objective third-party information (i.e., surveys conducted by state survey agencies, using a federal protocol), rather than on self-reported, unaudited information provided by facilities. 

However, in November, in the health survey portion of the website, CMS began to include a statement when no deficiencies were cited in each of nine regulatory areas (mistreatment, quality of care, resident assessment, resident rights, nutrition and dietary, pharmacy services, environmental, and administration) as a result of the annual standard survey and complaint surveys conducted in the prior year.  For example, the website reports for a facility, "No Resident Rights Deficiencies were found during this inspection period."  By comparison, no statement is made on Nursing Home Compare that deficiencies were not cited in the five other regulatory Requirements of Participation at 42 C.F.R. Part 483: Admission, Transfer, and Discharge Rights; Nursing Services; Physician Services; Specialized Rehabilitation Services; and Infection Control.

While CMS may have intended to provide additional information and clarity for consumers, the problem with CMS's new language is that it implies that the facility is in full compliance with the regulatory standards for which deficiencies were not cited.  However, since surveys regularly understate or undercite deficiencies, this statement, while accurately reflecting the absence of cited deficiencies, is often seriously misleading. 

CMS's own pilot surveys in 2014 found problems in pressure ulcers and antipsychotic drugs, two of the 14 care areas included in the Quality of Care Requirement. Moreover, the Government Accountability Office (GAO) has repeatedly reported that states understate deficiencies during surveys.  For example, in a May 2008 report entitled Federal Monitoring Surveys Demonstrate Continued Understatement of Serious Care Problems and CMS Oversight Weakness,[14] the GAO reported that many state surveys failed to cite deficiencies that federal comparative surveys identified.  The GAO found that about 15% of the federal comparative surveys identified deficiencies at the most serious levels of noncompliance – actual harm and immediate jeopardy deficiencies – that the state survey agency had failed to cite in the facility's annual standard survey.[15]  At lower levels of harm to residents, states' failures to identify problems were even higher.  More than 40% of the federal comparative surveys identified deficiencies that the state survey agency had not cited.[16]  The GAO reported, "At both levels of noncompliance, the most frequently missed deficiencies involved Quality of Care standards,"[17] which include "prevention of pressure sores, nutrition and hydration, accident prevention, and assistance with bathing and grooming."[18]  The GAO found some improvement, but still high levels of underciting of deficiencies, in a follow-up report issued in April 2010 .[19]

An analysis by the Center for Medicare Advocacy similarly found under-reporting of nurse staffing deficiencies.  The Center found that only .022% of the country's 16,100 nursing homes were cited with a staffing deficiency in 2013, despite the fact that most facilities do not have sufficient staffing.[20]

Conclusion

While CMS is making plans to improve the accuracy of sections of Nursing Home Compare that reflect resident assessment and staffing information, it is providing misleading information about the survey portion of the website.  CMS should not report that facilities are deficiency-free in certain regulatory requirements when state surveys fail to cite so many significant deficiencies.  The Center for Medicare Advocacy has called on CMS to withdraw the misleading information from the federal website.

T. Edelman, 11/20/2014