Medicare Current Beneficiary Survey (MCBS) Files

The MCBS is a continuous, multi-purpose survey of a representative sample of the Medicare population and is conducted by the Office of Strategic Planning. The central goals of the MCBS are provided below:

determine volume of utilization by type of service and the associated payments by source (i.e., by payer) for all health care used by Medicare beneficiaries, including covered and non-covered services.
To ascertain all types of health insurance coverage and relate coverage to sources of payment.
To trace processes, such as changes in health status, the "spending down" of resources to meet Medicaid eligibility requirements, and the impacts of program changes.

File Creation

The MCBS is a rotating panel survey in which a sample of Medicare beneficiaries (or appropriate proxy) is interviewed three times a year over the course of several years to form a continuous profile of their health care experience. The first round of interviews began in late 1991, and the survey is projected to continue indefinitely.

The sample for the MCBS is drawn from CMS' Medicare enrollment database and was designed to be representative of the entire beneficiary population. The sample is stratified by age. Certain age strata (two strata for the disabled or under age 65 population and one stratum for the oldest-old or population age 85 and over) were over-sampled to generate sufficient information to support detailed analyses of the special health care needs of these beneficiaries.

The sample is replenished annually to account for attrition (deaths, refusals, movement out of the sampling area, termination of coverage, etc.); to draw newly-enrolled beneficiaries into the sample; and to allow sample persons to be "rotated out" of the survey after a number of years. Supplemental samples are introduced to the survey in the fall round of each year (September - December).

All sample persons are included in the study regardless of living arrangement, i.e., in their own household, in that of a relative or non-relative, or in an institutional setting. Interviews are conducted for beneficiaries in long-term care facilities and for beneficiaries who reside in the community; a survey instrument appropriate to the setting is used.

The first time an interview is conducted for a new sample person, the respondent is introduced to the design and purpose of the study. Demographic, insurance, and baseline data on health status and access to care are collected. These introductory interviews do not collect detailed information about use of medical services and associated expenditures. In subsequent rounds of the survey, respondents are asked about health care events "since the last interview."

Community interviews are conducted using Computer-Assisted Personal Interviewing (CAPI) survey instruments. The CAPI program guides the interviewer through the survey questions, records the respondent's answers, and edits the answers for consistency and reasonableness. CAPI guides the interviewer through complex skip patterns and inserts follow-up questions when certain data are missing or incorrect. These automated features help ensure the accuracy of the data.

Interviews for beneficiaries in long-term care facilities are conducted with facility staff. Similar to the community questionnaires, the facility survey instrument is modified to be appropriate to a facility setting. The introductory interview collects demographic, insurance, and baseline health-related information and does not include the attitudinal or subjective items asked of community residents.

Initially, the facility interview was conducted from a printed questionnaire during which entries are handwritten on the form by the interviewer for later data entry. Since 1997, data collected was entered directly into a second CAPI program designed for facility interviews.

The MCBS is intended to create an ongoing time series of data for each respondent based on the information collected during each round of the survey. These data address utilization of health care services, medical care expenditures, health insurance coverage, sources of payment, health status and functioning, and a variety of demographic characteristics and behavioral patterns. Additional questions are asked periodically to supplement the time series data with information related to topics of special interest such as income, assets, access to medical care, satisfaction with care received, and sources of information about the program.

Data collected during the fall survey round are edited and combined with Medicare claims and other administrative data to create a cross-sectional MCBS file series, called the Access to Care File. To date, MCBS Access to Care files have been created for calendar years 1991 through 1997.

There is a second series of MCBS files known as the MCBS Cost and Use files. For the Cost and Use files utilization and payment data from each survey round are edited and linked, where possible, to Medicare claims, and input as necessary. This full year use and payment data is combined with other administrative and survey data to create the annual Cost and Use file. To date, a MCBS Cost and Use files have been created for calendar years 1992 through 1996.

File Maintenance

The MCBS interview process is repeated three times each year for each beneficiary in the sample. The data from the individual interviews are subjected to a series of consistency edits and are combined to develop the MCBS annual file. The claims data are subjected to the National Claims History Quality Assurance System as described in the National Claims History (NCH) Files group in the Claims and Utilization Data section of this document. For Calendar years 1991 through 1997, the Office of Strategic Planning has combined the demographic, insurance, and either baseline or updated health status and attitudinal data collected in the September - December round with Medicare claims for the applicable calendar year and released the data under a Data Use Agreement (DUA) with CMS(previously HCFA). Because of the emphasis on attitudinal questions, these files are collectively referred to as the Access to Care Files.

For calendar years 1992 through 1996, detailed event-level use and expenditure data For survey-reported-only events, linked survey-claim events (i.e., Medicare-covered service reported in the survey) and claims-only medical events (i.e., Medicare-covered service not reported during the survey) are released as a separate series of files under a DUA with CMS. These files are known as the Cost and Use Files.

File Structure and Usage

The MCBS Access to Care files For calendar years 1991 through 1997 are comprised of a number of data files that can be categorized into two types of dataset. The first category is based on survey and administrative summary data and includes the following files:

Key record file
Administrative Identification record file
Survey Identification record file
Survey Health Status and Functioning record file
Survey Access to Care record file
Survey Health Insurance record file
Survey Enumeration record file
Survey Facility Residence History record file
Survey HMO supplement record file (1996 and 1997)
Survey Facility Identification record file
Survey Interview record file
Survey Cross-Sectional Weights file
Survey Longitudinal Weights file

Some of the files above have one record for each respondent in the survey; others will have a record for each respondent if appropriate for the individual.

The second category is comprised of the Medicare claims by type of claim and includes the following files:

Inpatient hospital claims record file
Skilled Nursing Facility (SNF) (inpatient) claims record file
Hospice claims record file
Home Health Agency claims record file
Outpatient claims record file
Physician/Supplier claims record file
Durable Medical Equipment (DME) claims record file (beginning 1993)

For the files in the Medicare claims dataset, each record represents one claim. The claims records represent services provided during the calendar year. To facilitate analysis, the administrative identification record file contains a summary of the utilization presented in detail in the claims record files.

The 1992 Cost and Use MCBS file is comprised of data files that can be categorized into three types of datasets. The first category is based on survey and administrative summary data and includes the following files:

Key record file
Administrative Identification record file
Survey Identification record file
Survey Health Status and Functioning record file
Survey Health Insurance record file
Survey Enumeration (Household Characteristics) record file
Survey Facility Identification (Facility Characteristics) record file
Survey Interview record file
Residence Time Line record file
Survey Cross-Sectional Weights file

Some of the files above have one record for each respondent in the survey; others will have a record for each respondent if appropriate for the individual.

The second category is comprised of detailed and summary information from both the survey and from claims about medical goods and services the beneficiary used in the calendar year, the costs associated with those services, and the share of those costs borne by all payers.

For some services, records are provided in two levels of aggregation -- detail at the event level and summary information. For others only summary information is provided. The files included are:

Inpatient use and costs (detail and summary)
Outpatient use and costs (detail and summary)
Drug use and costs (detail and summary)
Facility use and costs (detail and summary)
Dental use and costs (detail and summary)
Medical services and goods (detail and summary)
Home health use and costs (summary only)
Hospice use and costs (summary only)
Person summary of all use and costs

The detail event records are present only for people who had services of a given type. More than one record may be present for a given individual. The summary records are present for every respondent.

The third category is comprised of the Medicare claims by type of claim and includes the following files:

Inpatient hospital claims record file
SNF (inpatient) claims record file
Hospice claims record file
Home Health Agency claims record file
Outpatient claims record file
Physician/Supplier claims record file
DME Claims record file

These records are available only for people who had service of a given type. There may be more than one record for a given individual.

The MCBS Cost and Use File provides a current and accurate picture of health care service use, expenditures, and sources of payment such as Medicare, Medicaid, VA, other public programs, private insurance, and out-of-pocket. These data enable CMS to perform the following tasks:

Monitor financial effects of changes in the Medicare program on beneficiaries.
Develop reliable and current information on the use and cost of services not covered by Medicare such as prescription drugs and long-term care.
Develop reliable and current information on the sources of payment for costs of covered services not reimbursed by Medicare.

Data Structure and Usage

The MCBS Files were originally designed to yield approximately 12,000 annual records for both the Access to Care and the Cost and Use series. With the change from a longitudinal to a rotating sample design (that is, rotating new sample persons into the survey in the fall, rotating retiring sample persons out in the subsequent winter or spring), approximately 16,000 cases are included in the Access to Care Files and 12,000 - 13,000 in the Cost and Use series.

The volume of survey, administrative, and claim-related data is extensive and requires storage on magnetic tape reels or cartridges for data dissemination. Survey-reported data and data summarized from administrative records are stored in sequential flat files consisting of fixed formatted records sorted by an assigned beneficiary identification number. Claim records are in variable length, packed-decimal EBCDIC format. Working with claims in this format may prove difficult for individuals in a non-IBM environment. Some summary information from the claims is therefore provided in the Administrative file (RIC A).

All MCBS File records begin with the assigned record identification code and year field to identify the record type and year of interest. These fields are followed by a unique beneficiary identification number and allow linkage between files. To obtain complete information for an individual, analysts must link records for an individual beneficiary from the various data files. The MCBS data have and are being used for a variety of purposes including the following:

provide the basic information needed to estimate the cost of program changes and expansions.
To monitor the effects of recent changes to the Medicare program.
To assess the effects of Medicare Physician Payment Reform on access to and costs of care.
To assist in the development of the National Health Accounts.
To conduct epidemiological studies associated with aging and the process of institutionalization.
To examine health care use and expenditures by the near-poor elderly - those who are not eligible for Medicaid and are relatively unprotected against increasing health care costs.
To understand the distribution of supplementary insurance coverage among different types of beneficiaries and its effect on Medicare utilization.
To understand the health care needs of the disabled, including the use of services and payments as well as access issues for this relatively unexplored subpopulation.
To trace processes, such as changes in health status or the exhaustion of assets and income through the process known as "spend-down," and the effects on individuals and government spending.

Methods of Access

CMS releases MCBS data only under a DUA. CMS will release some billing and administrative data with the MCBS data, commensurate with demonstrated need. Researchers who have specific needs for more detailed geographic information or for Medicare claims data may request Research Identifiable files from CMS. Requests for these files must include a study protocol with specific justification for the additional data required, along with an Identifiable DUA.

The MCBS Access to Care files (1991 through 1997) and the MCBS Cost and Use files (1992 through 1996) and accompanying documentation are available. Further information regarding these MCBS releases may be obtained either by contacting the OIS, EDG, Division of Data Liaison and Distribution via the Data Liaison Help Line at (410) 786-3690 or by visiting the MCBS web site at www.cms.hhs.gov/mcbs/default.asp.

Migration of Data

Currently, there are no changes planned in the file creation, data storage, or data retrieval processes for the MCBS File.

List of Data Elements

The list of data elements will be included in the future on the CMS web site at www.cms.hhs.gov.