Re: CMS-4062-N - Medicare and Medicaid Program: Solicitation for Information on the Hospital CAHPS (68 Federal Register 27154) June 27, 2003
Liberty Place, Suite 700
325 Seventh Street, NW
Washington, DC 20004-2802
(202) 638-1100 Phone
Monday, July 28th 2003
Thomas A. Scully
The Centers for Medicare & Medicaid Services
Room 443-G, Hubert H. Humphrey Building
200 Independence Avenue, SW
Washington, DC 20201
Dear Mr. Scully:
On behalf of our nearly 5,000 member hospitals, health care systems, networks and other providers of care, the American Hospital Association (AHA) appreciates the opportunity to provide comments on the Centers for Medicare & Medicaid Services’ (CMS) Hospital CAHPS pilot tool and survey process. We pledge to work with you and others to ensure that this survey can be combined with data from clinical care measures as part of the hospital-led Quality Initiative to meet the needs of the public for information on hospital quality.
To compile a useful and well-informed comment letter, the AHA solicited input from its members and consulted with experts from the survey vendor companies. Based in part on that information and in part on the vision for how the survey might be incorporated into The Quality Initiative, we urge CMS and Agency for Healthcare Research and Quality (AHRQ) to:
- Develop appropriate instructionson administering the survey to ensure the comparability of data collected through different modes (i.e. mail, telephone, etc.);
- Encourage hospitals to collect data as soon as possible but no later than 45 days after discharge;
- Conduct the necessary research to determine how factors – such as the health status of the patient, racial/ethnic background, payer source, and generational differences – may affect patient responses, and develop methods to adjust for these differences.
Mode of Administration
To the extent feasible, hospitals should be able to incorporate the HCAHPS survey into existing patient surveys and have their current vendors administer it. Hospitals have long-standing and valued relationships with these vendors, and many hospitals have tracked their performance data over several years using the same questions to ensure they know if they are improving their care in the eyes of those they are serving. Hospitals and their patients are best served if these relationships are maintained, and that will mean allowing a variety of modes of administration of the HCAHPS survey.
However, flexibility in administration should not be placed above comparability of the information. More research mustbe done to ensure that different modes of administration do not result in different ratings of care. Before the survey protocol is finalized and the survey is conducted nationally, additional pilot studies must be conducted using the different modes of administration for the questions likely to be contained in the final questionnaire. Without this analysis, hospitals will rightfully question whether the data displayed for the public fairly and accurately portrays differences in performance or simply differences in survey administration.
Ideally, data collection would take place on an ongoing basis. Continuous data collection will decrease the likelihood of seasonal fluctuations or skewed results caused by extraordinary events. It also will provide constant feedback to the hospital, thus maintaining internal momentum for improving quality in the patient experience. However, because not every hospital has an ongoing survey, continuous collection might disrupt current processes and increase costs to the hospitals. CMS and AHRQ should experiment with periodic survey administration throughout the year as a way to reduce the effects of seasonal variations and other factors while containing the costs of conducting the survey.
Further, conducting the survey within 45 days of patient discharge is critical to ensuring reliable survey results. In the pilot project, the gap between discharge and survey administration is six to nine months. This large amount of time is likely to reduce the accuracy of survey responses because patients will be less able to recall the details of their hospital experience. It also may reduce patients’ willingness to respond to the survey.
In consensus with the patient survey vendors, we believe the survey should center around eight aspects of hospital care that are critical to patients’ experiences:
- The care rendered by nurses,
- The care rendered by physicians,
- An overall impression of care,
- The patient’s assessment of whether they would recommend the hospital,
- Communication with the patients about their care needs, especially at discharge,
- The coordination of care and communication among their caregivers,
- Whether there was adequate pain management, and
- The patients’ insights on issues of patient safety and the reliability of the system of care.
We believe these eight topic areas are appropriate because the public will find the information understandable and meaningful when evaluating a hospital. Although there are questions on the draft survey that are suitable for gathering broad assessments of these aspects of care – question #10, for example, asks for an overall rating of nursing care –the survey should not be limited solely to eight broad questions. More specific questions assessing the key aspects of care within each of these broad areas will be needed to give the public meaningful information on hospital performance and hospitals the opportunity to target areas for quality improvement. Because many vendors currently administer the surveys via telephone , and experts have found that telephone surveys with more than 25 questions have a lower response rate, the HCAHPS survey should be limited to a maximum of 25 questions.
When the survey is finally developed, it is imperative that it can be conducted in multiple languages besides English and Spanish to allow hospitals to obtain input from as broad an array of patients as possible. The survey must be administered at minimum in English, Spanish, Chinese, Vietnamese, Russian, and Arabic. Further, we urge CMS and AHRQ to translate the survey into the top 20 languages used in the U.S. as the program flourishes.
Adjusting for Differences of Patients
Factors such as racial/ethnic backgrounds, generational differences, payer source, health status, and age may affect a patient’s hospital experience. . In our conversations with vendors, we learned that adjustments (beyond simply separating obstetrics from medicine, surgery, and other types of patient care) must be developed for the HCAHPS survey to be used to report comparable data to the public. Unfortunately, we lack evidence that will allow us to identify the most critical factors for which adjustments must be made, and we lack data to help calculate what adjustments should be supplied. We strongly urge CMS and AHRQ to carry out further studies that will enable all to determine how patients’ ratings of hospitals may be affected by such factors as:
- Racial or ethnic backgrounds,
- Age or generational differences,
- Hospital location, and
- Hospital size.
The AHA is eager to see the program implemented as soon as possible, but it is more important to ensure that survey information is accurate and comparable. Therefore, the AHA urges CMS and AHRQ to conduct further studies to determine how different factors can affect the patient’s hospital experience.. Once the differences are discerned, a tool must be developed to adjust for these differences. It is imperative that this is done before the final version of the survey is released and implemented.
Thank you for the opportunity to provide comments. If you have any questions about these comments, please feel free to contact Nancy Foster at 202-626-2337.
Executive Vice President