Asking Questions about Diseases: How Questionnaire Design Changes Impact the Chronic Disease Data of the Medicare Current Beneficiary Survey

Background The Medicare Current Beneficiary Survey (MCBS) is a continuous study of a nationally representative sample of aged and disabled Medicare beneficiaries sponsored by the Centers for Medicare & Medicaid Services (CMS). It provides comprehensive data on health and functional status, health care expenditures, and health insurance for Medicare beneficiaries. Data on sample persons may be collected in the Community Interview, the Facility Interview, or both, depending on whether they reside at home or in a long-term care setting. The MCBS Facility Interview collects information about institutionalized patients from proxy respondents, most of whom are nurses or other primary care givers. To enhance data quality, proxy respondents are encouraged by the interviewers to refer to the sample person’s medical records when answering questions about health status, including chronic diseases. One of the key components of every resident’s medical record in a nursing home is the Minimum Data Set (MDS), a set of federally-mandated clinical assessments administered to nursing home residents. Information on the sample person’s chronic diseases is recorded in the Disease Diagnosis section in the MDS, in the subsection of Diseases. Forty-three categories of chronic diseases, and the options of “other” and “none of above,” were listed in this subsection. The Disease Diagnosis section includes an instruction for nursing home staff to “check only those diseases that have a relationship to current ADL status, cognitive status, mood and behavior status, medical treatments, nursing monitoring, or risk of death.” Inactive diagnoses are not recorded. The Facility Questionnaires underwent significant questionnaire design changes in 1997. In particular, the Health Status section of the Facility Instruments was redesigned to build around the use of the MDS in data collection. To mirror MDS format and wording, the reference period, question format, and wording of the chronic disease questions, among others in the Health Status section, were revised. At the same time, the mode of administration was changed from paper-and-pencil interview (PAPI) to computer-assisted personal interview (CAPI). As a result, the distributions of some of the items in the Health Status section after the changes were very different from those before the changes. The chronic disease items were especially affected by these changes. This paper discusses the background and nature of the questionnaire design changes to the chronic disease questions, the impact of these changes on the data, and the implications of these changes.