MEPS-IC
Sample Design
and Data Collection Process
The
Insurance Component of the Medical Expenditure Panel Survey (MEPS-IC)
is an annual survey of employers that collects information on
the employer’s health insurance offerings. There are two
distinct samples fielded in the MEPS IC survey: the List sample
and the Household (link) sample. While these samples are designed
to address different survey goals, the two have been combined
to make data collection more efficient. The U.S. Census Bureau
serves as the data collection agent for the MEPS-IC survey.
MEPS-IC List sample. The MEPS-IC List
sample is an independently drawn, nationwide sample
of establishments and state/local governments. This sample is specifically
designed to make national and state estimates each year. Tables
of estimates, derived from the List sample, are posted on the MEPS
Web site. The List sample is drawn from the Business Register,
a confidential list of establishments in the United States maintained
by the Census Bureau and from the Census of Governments, a frame
maintained by the Census Bureau and updated once every five years.
Due to the confidentiality requirements surrounding any samples
drawn from the Business Register, there can be no Public Use Files
released with micro data. For researchers, the MEPS-IC List sample
microdata can only be accessed through the Census
Bureau's Research Data Centers. The data are not available
at the AHRQ Data Center and the List sample is not linked to the
MEPS Household
Component (MEPS-HC) survey.
MEPS-IC Household sample. The MEPS-IC
Household sample (also known as the
Link sample or the Household Link sample) is a sample of employers
that are identified
by respondents in the MEPS-HC as their main employer or secondary
employer that is the source of their health insurance. This sample
is directly linkable to the MEPS-HC survey and is specifically
designed for that purpose. The Household sample does not support
the making of national- or state-level estimates. While the Household
sample is not derived from a confidential frame, the promise of
confidentiality given to respondents makes it impossible to issue
a public use file with the microdata. Go to Household Component/Insurance Component Link Files for
more information on how researchers can access these files.
Data collection methods. The MEPS-IC uses
multiple data collection methods. Establishments
(and governmental units) are initially screened by telephone to
confirm their mailing addresses and to establish a point-of-contact
with a knowledgeable respondent. If an establishment does not offer
health insurance to any of its employees, the noninsurance questions
about establishment and firm characteristics are asked at that
time, thus completing the survey by telephone.
Establishments that offer health insurance and those that do not
respond to the telephone screening are mailed survey questionnaires.
The mailing consists of an establishment-level questionnaire and
separate questionnaires for the health insurance plans offered.
A second mailing is sent if a response to the first mailing is
not received within a three-week period. If the establishment does
not return either mailing, it is called and interviewed over the
telephone. A computer-assisted telephone interview questionnaire
is used at this stage to conduct the interview and enter the establishment’s
responses directly into a database. Large companies and governments
are occasionally interviewed in person due to the large amount
of data being requested of them.
The health insurance plan questions are asked for each offered
plan, up to a maximum of four plans. Most companies do not offer
more than four plans. However, if an establishment indicates during
the telephone prescreening interview that it offers more than four
plans, the names of all the plans are collected and then plan-level
questionnaires are sent for those plans with the largest enrollments—but
making sure that we get at least one plan of each type offered.
The MEPS-IC data tables are published annually. In order to collect
data to support both the List sample and the Household sample of
the MEPS-IC, the questions asked of employers are for the previous
calendar year. This retrospective data collection, along with the
time it takes to process survey data of this magnitude, creates
a significant delay. Private sector tables are posted on the MEPS
Web site in July and the public sector tables a month later in
August of the year after the data are collected. For example, data
used to create the 2003 MEPS-IC tables were collected in 2004,
and the tables were posted in July/August 2005.
Go to Survey Questionnaires — Insurance
Component to access the main questionnaires used in the collection of the MEPS-IC survey.
Suggested
Citation:
MEPS-IC Sample Design and Data
Collection Process. Agency
for Healthcare Research and Quality, Rockville, Md. http://www.meps.ahrq.gov/survey_comp/ic_data_collection.jsp
|
|
|