Rersponse to Public Comments

Comments Summary for Data Collection for the Nursing Home Value-Based Purchasing (NHVBP) Demonstration

CMS Collection # CMS-10240

The 60-day public comment period for the NHVBP demonstration ended September 18, 2007. We received nine comments from two commenters. These fell into several categories: Comments on the data collection forms, comments on the electronic payroll submission guidelines, comments on the quality measures, and comments on the demonstration design and evaluation. These comments and our responses are as follows:

Comments on Data Collection Forms

Comment: One commenter recommended including the following question on the nursing home application:

“Does the applicant use an electronic health record (EHR) that captures, maintains, and manages patient health information?

Response: We have included this question on the application. The question will help us identify nursing homes that are implementing health information technology (HIT). This information may be helpful when we stratify the applicant nursing homes prior to randomly assigning them to the demonstration and control groups in each State. There is no impact on the burden due to this change.

Comments on Electronic Payroll Submission Guidelines

Comment: One commenter suggested that Medication Aides and Restorative Aides should not be included in the Certified Nurse Aides (CNA) category. State training requirements may vary and in many cases these aides may not have the training that meets the requirements for CNA training. This category should include only individuals who are fully certified, not those who are still in training. Also, if an individual works two positions, such as CNA and housekeeping, then the hours worked in housekeeping should not be included in the measure.

Response: We believe that Medication Aides and Restorative Aides should be included in the nurse aides category since they provide direct care to residents. However, we agree that these aides should be trained and certified. Thus we have changed the guidelines to clarify that Medication Aides and Restorative Aides must be State certified in order to be included in the CNA category. Note that the each nursing home’s performance will be assessed relative to other homes in the State, so training requirements that are standard within State but vary across States will not cause any homes to be disadvantaged in terms of performance. We have also changed the guidelines to clarify that hours worked by CNAs in training should not be included. We have also excluded nurse aides in training from the Nurse Temporary Agency Staff section of the data collection form. Regarding the situation where CNAs work two positions, note that it is not feasible for nursing homes to report only the hours worked as CNAs because their payroll systems are not usually designed to track hours by task. Nursing homes that use CNAs for more than one function have a better workflow and greater flexibility to provide quality care. Therefore, CNAs that perform housekeeping duties will be classified as CNAs and all of their hours will be included.

Comments on Quality Measures

Comment: One commenter indicated support for the inclusion of the appropriate hospitalization measures for both short-stay and long-stay residents.

Response: We appreciate the support for these measures and believe that ‘appropriate hospitalizations’ is an important quality domain.

Comment: One commenter requested more detail regarding the survey deficiency domain. Information was requested regarding the specific level of deficiency that would screen out a nursing home from participation in the demonstration, as well as how a deficiency score would be determined.

Response: Survey deficiencies will be used in two ways. First, nursing homes that are participating in the demonstration will be ineligible to receive a performance payment if they were cited for a serious deficiency or enforcement action in the demonstration year. Second, the surveys will be used to determine a score for this domain. More detailed information about this can be found in the preliminary design report at the CMS website:

www.cms.hhs.gov/DemoProjectsEvalRpts/md/itemdetail.asp?itemID=CMS1198946

Note that CMS is refining the preliminary design and will post additional information on the website when it is available.

Comment: One commenter noted that CMS was collecting information on staff immunizations to assess whether this measure should be included in the demonstration in the future, depending on whether there was strong evidence of a link between this measure and resident outcomes. The commenter asked how CMS would determine that outcomes were linked to staff immunization rates and not other factors.

Response: We will examine staff immunization rates to see if they correlate with outcome measures such as hospitalization rates. For example, one approach would be to compare hospitalization rates for residents in the demonstration group to those in the control group who were hospitalized for influenza and pneumonia. As we analyze the immunization data in relation to outcomes, we intend to control for other factors. Additional detail on the analysis approach will be available at a later time.

Comment: One commenter suggested that the use of resident care experience surveys as a quality measure may be premature. It would be beneficial to first gather information such as the issues these surveys address and what particular survey was used. The commenter asked how CMS would determine possible links between the use of resident care surveys and resident outcomes.

Response: The purpose of this measure is to determine how nursing homes use information on the care experience of their residents. We anticipate that most homes collect or have access to information on the quality of life and care of residents, but that some homes use this information more effectively than others. Our goal is to provide incentives to use the information more effectively. We did not include questions about the kind of survey used because we do not intend to incentivize homes to use any particular kind of survey. Details on our approach for analyzing the data we collect will be available at a later time.

Comment: One commenter suggested that CMS consider measures of culture change, such as private rooms with baths, consistent assignment of staff to residents, and consumer direction. Also, the commenter was not clear as to why measures such as pain, medication errors and the use of electronic medical records were not included in the demonstration.

Response: CMS considered a wide range of quality measures for inclusion in the demonstration. We sought measures which were verifiable and which could be collected without undue burden. And we wanted measures that would encourage improvement in overall quality of care of nursing home residents. We did not include private rooms with baths because we did not believe that most nursing homes would be able to improve on this measure (i.e., change the structure of their facility) during the demonstration. Consistent assignment and consumer direction were not included because we did not believe that we could verify the accuracy of the data reported by nursing homes. Note that while we did not include culture change measures, we anticipate that nursing homes that have implemented culture change will perform well on the measures that are included in the demonstration.

Although pain is an important measure, we did not include it because of concerns about differences across nursing homes in how they assess pain. Differences in prevalence rates across nursing homes may reflect differences in assessment practices rather than differences in quality of care. We did not include medication errors because the only feasible approach for collecting accurate information would be a medical record review, which would be too burdensome. The use of electronic medical records was not included in the demonstration because we have no evidence of a quantitative relationship between electronic records and quality of care in nursing homes. As stated above, we have modified the application to include a question about the use of electronic health records. This measure has potential, but further research on the relationship between health information technology and nursing home quality of care is necessary before we can include such a measure.

Comments on Demonstration Design and Evaluation

Comment: One commenter encouraged CMS to publish and make publicly available the nursing homes’ performance scores, and urged that the information be presented in a format easily understandable to consumers.

Response: We would like to further clarify our intentions regarding publication of scores. As a non-financial incentive, CMS may publish the aggregate scores of nursing homes that qualify for performance awards under the demonstration. However, since participation is voluntary and we want to encourage poor performing homes to participate, we do not plan to publish the scores of nursing homes that do not qualify for awards.

Comment: One commenter indicated that a strong evaluation of the demonstration would be critical. An independent evaluation to examine what works and what does not work will be an important part of assessing the effectiveness of this demonstration and the impact of pay-for-performance on the quality of care in nursing homes.

Response: We agree with this comment. CMS plans to procure a contractor to conduct the evaluation.