������������������yyyyyzzzz{{{{{||||

Design and Methods of the Medical Expenditure Panel Survey Household Component

Report 1Methodology Methodology

U.S. Department of Health and Human ServicesPublic Health ServiceAgency for Health Care Policy and Research A

genc

y fo

r H

ealt

h C

are

Polic

y an

d R

esea

rch

������� ������� �������� �������� yyyyyyy zzzzzzz {{{{{{{{ ||||||||

Health Care Information and Electronic Ordering Through the AHCPR Web Site

The Agency for Health Care Policy and Research’s Website—http://www.ahcpr.gov/—makes practical, science-basedhealth care information available in one convenient place.

Six buttons correspond to major categories of availableholdings on the Web site: Offices/Centers, News &Resources, Research Portfolio, Data & Methods, Guidelines& Medical Outcomes, and Consumer Health.

The Web site features an Electronic Catalog to the more than450 information products generated by AHCPR, withinformation on how to obtain these resources. Manyinformation products have an electronic ordering form andare mailed free of charge from the AHCPR Clearinghousewithin 5 working days.

http://www.ahcpr.gov/

i

BackgroundThe Medical Expenditure Panel Survey (MEPS) is

conducted to provide nationally representative estimatesof health care use, expenditures, sources of payment,and insurance coverage for the U.S. civiliannoninstitutionalized population. MEPS also includes anationally representative survey of nursing homes andtheir residents. MEPS is cosponsored by the Agency forHealth Care Policy and Research (AHCPR) and theNational Center for Health Statistics (NCHS).

MEPS comprises four component surveys: theHousehold Component (HC), the Medical ProviderComponent (MPC), the Insurance Component (IC), andthe Nursing Home Component (NHC). The HC is thecore survey, and it forms the basis for the MPC sampleand part of the IC sample. The separate NHC samplesupplements the other MEPS components. Togetherthese surveys yield comprehensive data that providenational estimates of the level and distribution of healthcare use and expenditures, support health servicesresearch, and can be used to assess health care policyimplications.

MEPS is the third in a series of national probabilitysurveys conducted by AHCPR on the financing and useof medical care in the United States. The NationalMedical Care Expenditure Survey (NMCES) wasconducted in 1977, the National Medical ExpenditureSurvey (NMES) in 1987. Beginning in 1996, MEPScontinues this series with design enhancements andefficiencies that provide a more current data resource tocapture the changing dynamics of the health caredelivery and insurance system.

The design efficiencies incorporated into MEPS arein accordance with the Department of Health andHuman Services (DHHS) Survey Integration Plan ofJune 1995, which focused on consolidating DHHSsurveys, achieving cost efficiencies, reducingrespondent burden, and enhancing analytical capacities.To accommodate these goals, new MEPS designfeatures include linkage with the National HealthInterview Survey (NHIS), from which the sample forthe MEPS HC is drawn, and enhanced longitudinal datacollection for core survey components. The MEPS HCaugments NHIS by selecting a sample of NHISrespondents, collecting additional data on their healthcare expenditures, and linking these data with additionalinformation collected from the respondents’ medicalproviders, employers, and insurance providers.

Household ComponentThe MEPS HC, a nationally representative survey

of the U.S. civilian noninstitutionalized population,collects medical expenditure data at both the person andhousehold levels. The HC collects detailed data ondemographic characteristics, health conditions, healthstatus, use of medical care services, charges andpayments, access to care, satisfaction with care, healthinsurance coverage, income, and employment.

The HC uses an overlapping panel design in whichdata are collected through a preliminary contactfollowed by a series of six rounds of interviews over a 2 1/2-year period. Using computer-assisted personalinterviewing (CAPI) technology, data on medicalexpenditures and use for 2 calendar years are collectedfrom each household. This series of data collectionrounds is launched each subsequent year on a newsample of households to provide overlapping panels ofsurvey data and, when combined with other ongoingpanels, will provide continuous and current estimates ofhealth care expenditures.

The sampling frame for the MEPS HC is drawnfrom respondents to NHIS, conducted by NCHS. NHISprovides a nationally representative sample of the U.S.civilian noninstitutionalized population, withoversampling of Hispanics and blacks. A subsample of10,500 households was drawn from the NHIS samplingframe for the initial 1996 MEPS HC panel. Every 5years the HC sample size is increased. Beginning with����

��������������������yyyyyyyy

zzzzzzzz{{{{||||

Suggested citationCohen J. Design and methods of the Medical ExpenditurePanel Survey Household Component. Rockville (MD):Agency for Health Care Policy and Research; 1997. MEPSMethodology Report No.1. AHCPR Pub. No. 97-0026.

The Medical Expenditure Panel

Survey (MEPS)

ii

the 1997 panel, policy-relevant population subgroupsare oversampled. The subgroups initially targetedinclude adults with functional impairments, childrenwith functional limitations in their activities, individualsaged 18-64 who are predicted to have high levels ofmedical expenditures, and individuals with familyincome less than 200 percent of the poverty level.

Medical Provider ComponentThe MEPS MPC supplements and validates

information on medical care events reported in theMEPS HC by contacting medical providers identifiedby household respondents. The MPC sample includesall hospitals, hospital physicians, home health agencies,and pharmacies reported in the HC. Also included in theMPC are all office-based physicians:

• Providing care for HC respondents receivingMedicaid.

• Associated with a 75-percent sample of householdsreceiving care through an HMO (health maintenanceorganization) or managed care plan.

• Associated with a 25-percent sample of theremaining households.

The 1996 sample is projected to provide data fromapproximately 2,700 hospitals, 12,400 office-basedphysicians, 7,000 separately billing hospital physicians,and 500 home health providers.

Data are collected on medical and financialcharacteristics of medical events reported by HCrespondents, including:

• Diagnoses coded according to ICD-9 (9th Revision,International Classification of Diseases) and DSM-IV (Fourth Edition, Diagnostic and StatisticalManual of Mental Disorders).

• Physician procedure codes classified by CPT-4(Current Procedural Terminology, Version 4).

• Inpatient stay codes classified by DRGs (diagnosis-related groups).

• Charges, payments, and the reasons for anydifference between charges and payments.

The MPC is conducted through telephoneinterviews and mailed survey materials.

Insurance ComponentThe MEPS IC collects data on health insurance

plans obtained through employers, unions, and othersources of private health insurance. Data obtained in theIC include the number and types of private insuranceplans offered, benefits associated with these plans,premiums, contributions by employers and employees,eligibility requirements, and employer characteristics.

Establishments participating in the MEPS IC areselected through four sampling frames:

• A list of employers or other insurance providersidentified by MEPS HC respondents who reporthaving private health insurance at the Round 1interview.

• A Bureau of the Census list frame of private sectorbusiness establishments.

• The Census of Governments from the Bureau of theCensus.

• An Internal Revenue Service list of the self-employed.

To provide an integrated picture of health insurance,data collected from the first sampling frame (employersand other insurance providers) are linked back to dataprovided by the MEPS HC respondents. Data from theother three sampling frames are collected to provideannual national and State estimates of the supply ofprivate health insurance available to American workersand to evaluate policy issues pertaining to healthinsurance.

The MEPS IC is an annual panel survey. For thesurvey conducted in 1997, the sample includesapproximately 7,000 establishments identified throughthe MEPS HC, 27,000 identified through the businessestablishments list frame, 1,900 from the Census ofGovernments, and 1,000 identified through the list ofthe self-employed. Data are collected from the selectedorganizations through a prescreening telephoneinterview, a mailed questionnaire, and a telephonefollowup for nonrespondents.

��������������������

��������������������

��������������������

��������������������

yyyyyyyyyyyyyyyyyyyy

zzzzzzzzzzzzzzzzzzzz

{{{{{{{{{{{{{{{{{{{{

||||||||||||||||||||

Nursing Home ComponentThe 1996 MEPS NHC is a survey of nursing

homes and persons residing in or admitted to nursinghomes at any time during calendar year 1996. TheNHC gathers information on the demographiccharacteristics, residence history, health and functionalstatus, use of services, use of prescription medications,and health care expenditures of nursing homeresidents. Nursing home administrators and designatedstaff also provide information on facility size,ownership, certification status, services provided,revenues and expenses, and other facilitycharacteristics. Data on the income, assets, familyrelationships, and care-giving services for samplednursing home residents are obtained from next-of-kinor other knowledgeable persons in the community. Inkeeping with the DHHS Survey Integration Plan, theNHC is designed to be conducted every 5 years.

The 1996 MEPS NHC sample was selected usinga two-stage stratified probability design. In the firststage, facilities were selected; in the second stage,facility residents were sampled, selecting both personsin residence on January 1, 1996, and those admittedduring the period January 1 through December 31.

The sample frame for facilities was derived fromthe National Health Provider Inventory, which isupdated periodically by NCHS. The MEPS NHC dataare collected in person in three rounds of datacollection over a 1 1/2-year period using the CAPIsystem. Community data are collected by telephoneusing computer-assisted telephone interviewing (CATI)technology. At the end of data collection, the samplewill consist of approximately 800 responding facilities,3,100 residents in the facility on January 1, andapproximately 2,200 eligible residents admitted during 1996.

Survey ManagementMEPS data are collected under the authority of the

Public Health Service Act. They are edited and

published in accordance with the confidentialityprovisions of this act and the Privacy Act. NCHSprovides consultation and technical assistance.

Data collection is conducted under contract byWestat, Inc., Rockville, MD, and the National OpinionResearch Center at the University of Chicago, as wellas through an interagency agreement with Bureau ofthe Census. Technical consultation is provided byMedstat, Inc., Boston, MA. Data processing support isprovided under contract by Social & ScientificSystems, Inc., Bethesda, MD.

As soon as data collection and editing arecompleted, the MEPS survey data are released to thepublic in staged releases of summary reports andmicrodata files. Summary reports are released asprinted documents and electronic files. Microdata filesare released on CD-ROM and/or as electronic files.

Printed documents and CD-ROMs are availablethrough the AHCPR Publications Clearinghouse. Writeor call:

AHCPR Publications ClearinghouseAttn: (publication number)P.O. Box 8547Silver Spring, MD 20907800/358-9295410/381-3150 (callers outside the United States only)888/586-6340 (toll-free TDD service; hearing impaired only)Be sure to specify the AHCPR number of the

document or CD-ROM you are requesting. Selectedelectronic files are available on the Internet in theMEPS section of the AHCPR home page:

http://www.ahcpr.gov/

Additional information on MEPS is available fromthe MEPS project manager or the MEPS public usedata manager at the Center for Cost and FinancingStudies, Agency for Health Care Policy and Research,2101 East Jefferson Street, Suite 500, Rockville, MD20852 (301/594-1406).

iii

Table of Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Household Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Preliminary Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2

Core Rounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Uses of Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Tables

1. Panel design for the MEPS Household Component, 1996 and 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2. Design features of the MEPS Household Component, 1996 panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

iv

1

Introduction

The Medical Expenditure Panel Survey (MEPS) isthe third in a series of nationally representative surveysof medical care use and expenditures sponsored by theAgency for Health Care Policy and Research (formerlythe National Center for Health Services Research). Thefirst of these surveys, called the National Medical CareExpenditure Survey (NMCES), was conducted in 1977,and the second, called the National Medical ExpenditureSurvey (NMES), in 1987. The 1996 MEPS, which iscosponsored by the National Center for Health Statistics(NCHS), will update the 1987 data to reflect thedramatic changes that have occurred in the U.S. healthcare system over the last decade.

Background

Major changes have taken place in the Nation’shealth care delivery system since NMES was conductedalmost 10 years ago. The most notable is the rapidexpansion of managed care arrangements such as HMOs(health maintenance organizations), PPOs (preferredprovider organizations), and other provider networks thatseek to minimize increases in health care costs. Newhybrid forms of health insurance coverage also haveappeared. Changes such as these have affected both theprivate and public sectors. The new MEPS providesinformation about the current state of the health caresystem in the United States and the changes that havetaken place since the last national survey of medicalexpenditures was conducted in 1987. The informationcollected by MEPS also provides valuable baseline datafor use in evaluating future changes in the system.

The MEPS study design was developed to enhancethe capabilities to study changes in health care deliveryand the effects of new health policies. These are

important objectives in view of the various health reforminitiatives being implemented by States and the FederalGovernment. The MEPS design allows for theproduction of annual estimates for 2 calendar years. Italso permits the tracking of changes in employment,income, health status, and medical care use andexpenditures over the 2 consecutive years during whichhouseholds will be interviewed. In addition, NationalHealth Interview Survey (NHIS) baseline data areavailable for persons in the MEPS panels, thereby addinganother data point for comparisons of change over time.

MEPS extends the NMES series of studies onmedical expenditures and health insurance and provides,for the first time, data suitable for detailed analysis oftrends and changes in these areas. The survey is aunique resource for a number of reasons, including:

• Scope. MEPS provides information on a broadspectrum of the population. The survey sample baserepresents the civilian noninstitutionalized populationand, in a separate component, the populationinstitutionalized in nursing homes. MEPS alsoprovides information on many types of health careservices, expenditures, and sources of payment forboth individuals and families.

• Population basis. Because MEPS is a survey ofpersons, population groups that are or may become ofspecial policy concern can be identified andanalyzed. This is especially important for analyzingthe effect of particular eligibility requirements on theenrollment and budgets of public programs and onthose who are not eligible for such programs.

• Cost-effectiveness. MEPS will collect data neededby groups that might otherwise either sponsorseparate or overlapping surveys, or do without crucialinformation needed for important decisions.Experience has demonstrated that broad-based dataon use, expenses, and financing of health care

Design and Methods of the Medical Expenditure Panel SurveyHousehold Component

by Joel Cohen, Ph.D.,Agency for Health Care Policy and Research

collected from a nationally representative sample canmeet the data needs of a wide variety of users in acost-effective manner.

Household Component

The sample design of the NMES household surveyshas been revised for MEPS. The MEPS sample is notdefined through an initial screening round. Instead, it isselected as a nationally representative subsample ofhouseholds that participated in NHIS. The 1996 MEPSsample (based on the 1995 NHIS) is being carriedforward into 1997 and combined with a new subsampleof households responding to the 1996 NHIS. These twopanel samples (the 1996 MEPS sample and the newMEPS selections from the 1996 NHIS) will jointlydefine the sample base for the 1997 MEPS HouseholdComponent. Table 1 shows the study design of the1996 and 1997 MEPS Household Components. Table 2summarizes various features of the study design of theHousehold Component.

In 1996, the MEPS sample linked to the 1995 NHISwas selected. It was drawn from a nationallyrepresentative NHIS subsample that included 195primary sampling units and approximately 1,700segments, yielding approximately 10,500 respondingNHIS households. These households were recontactedin MEPS. This NHIS subsample reflects an oversampleof Hispanics and blacks. Other groups with high publicpolicy relevance in the areas of health care use andfinancing are oversampled as part of the MEPS 1997panel to improve the precision of the estimates for thosegroups.

Households selected for participation in the 1996 or1997 MEPS Household Components are interviewed inperson five times (Rounds 1-5) and a last time during abrief telephone interview (Round 6). The rounds of datacollection are spaced approximately 4 months apart.The interviews take place with a family respondent whoreports for himself or herself and for other familymembers.

Preliminary Contact

Mail and telephone contacts take place prior to thefirst MEPS interview with the NHIS participatinghouseholds selected for each MEPS panel (Round 1).The purpose of the preliminary contact is to enlist thehousehold respondent into MEPS and plan for thedelivery of record-keeping materials before the studyobservation period begins on January 1st of the surveyyear. In December, an advance letter announcing MEPSis mailed to the family respondent at the address wherethe NHIS interview was conducted. An interviewerfollows up on the letter with a telephone call to confirmits arrival, verify the identity of the household, identifythe MEPS family respondent (if different from the NHISrespondent), and announce the future mailing of a studycalendar and record file. These materials are sent, alongwith $5 to compensate respondents for the time andeffort devoted to keeping records in preparation for theRound 1 interview. An interviewer telephones a secondtime to confirm the arrival of these materials andarrange for a convenient time to conduct the Round 1interview.

Households that do not have a telephone or cannotbe reached using the telephone number from NHIS arecontacted by mail and asked to return a postcardidentifying a telephone number where they can becontacted (e.g., the number at work or a neighbor'shouse).

Core Rounds

Data collection for the MEPS HouseholdComponent takes place using a computer-assistedpersonal interview (CAPI) system. A core instrument isadministered in each of the first five rounds of datacollection, with periodic supplements added in selectedrounds to deal with specific topics in greater depth.Dependent interviewing methods, in which respondentsare asked to confirm or revise data provided in earlierinterviews, is used to update information in several ofthe core questionnaires, such as employment and healthinsurance, after the initial interview.

2

3

Core InstrumentThe core instrument is used to collect data about all

persons in sampled households. It includesquestionnaires on demographics, health status andconditions, use, charges and payments, prescribed andover-the-counter medicines purchased, employment, andhealth insurance.

Periodic SupplementsSupplements scheduled for inclusion in the survey

include questionnaires on access to and satisfaction withcare, income and assets, long-term care, and alternativecare (which includes approaches to health care that aredifferent from those typically practiced by medicaldoctors in the United States, such as acupuncture andhomeopathic treatments).

Self-Administered QuestionnaireAll adults in sample households are asked to

complete a self-administered questionnaire in Round 2.This questionnaire collects information about healthbehaviors and opinions that would be difficult, if notimpossible, to collect on a proxy basis from the familyrespondent. Similar information is collected for childrenas part of the regular interview with the householdsurvey respondent, usually the mother.

Medical Provider Permission FormsSigned permission forms are requested in Round 1,

much earlier than in past NMES studies, in order toexpedite the timetable for the later Medical ProviderComponent of the survey, which collects data aboutspecific medical events directly from providers. Resultsfrom a previous methodological study suggested thatearly requests for signed permission forms involvingoffice-based physicians have a modest negative effect onsurvey cooperation rates in later rounds. Therefore, therequests for signed permission forms in Round 1 arelimited to events taking place in hospitals. In Round 2and subsequent rounds, requests for signed permissionforms apply to all types of providers included in theMedical Provider Component of MEPS (that is,

hospitals, physicians, and home health agencies),including those associated with use reported in Round 1.

Health Insurance Permission FormsSigned permission forms are needed to contact

sources of employment and private health insurancecoverage in the Insurance Component of the survey,which collects data directly from individuals’ sources ofhealth insurance (typically their employers). Theserequests are initiated in Round 2 and apply to theinsurance sources associated with plans held at the timeof the Round 1 interview.

Health Insurance Policy BookletRequests

Following procedures tested successfully in aprevious methodological study, MEPS interviewers askrespondents to provide health insurance booklets or othersummary materials that describe the characteristics ofprivate plans held by family members at the time of theRound 1 interview. The requests for policy informationinclude all sources of private insurance coverage, not justemployment-related coverage. Respondents arereimbursed $15 for the time and effort involved inprocuring policy booklets.

Provider DirectoriesTo assist in the identification of medical providers

and the preparation of an unduplicated list of medicalproviders, interviewers use a computerized database(directory) of health providers that has been loaded intothe CAPI laptop computer. With search software loadedinto the laptops, interviewers can query the database ofproviders in the course of the MEPS interview. If amatch is found in the database for the provider specifiedby the household respondent, the matched directoryrecord is associated with the household member.Directory records include the following information foreach provider: a unique provider identification number;the provider’s name, address, and telephone number; andthe provider’s specialty (for individual office-basedphysicians).

4

Uses of Data

At the most basic level, the objective of the MEPSHousehold Component is to collect data that can be usedto produce annual estimates for a variety of measuresrelated to the characteristics of individuals; their healthinsurance coverage; and their health care use,expenditures, and sources of payment. The data can alsobe used to support behavioral analyses that informresearchers and policymakers about how thecharacteristics of individuals and families, including theirhealth insurance, affect medical care use and spending.

Data obtained in this study will be used to producethe following national estimates for calendar years 1996and 1997:

• Annual estimates of health care use and expendituresfor persons and families.

• Annual estimates of sources of payment for healthcare expenses, including amounts paid by publicprograms, such as Medicare and Medicaid, and byprivate insurance, as well as out-of-pocket payments.

• Annual estimates of health care use, expenditures,and sources of payment for persons and families bytype of service, including inpatient hospital stays,ambulatory care, home health care, dental care, andpurchases of prescribed and over-the-countermedicines.

• The number and characteristics of the populationeligible for each of the public programs, including theuse of services and expenditures of the populationeligible for benefits under Medicare, Medicaid,CHAMPUS and CHAMPVA (Civilian Health and

Medical Program for the Uniformed Services andCivilian Health and Medical Program, Veterans’Affairs), and the Department of Veterans’Affairs.

• The number, characteristics, use of services,expenditures, and benefits of persons and familieswith individual or group coverage; commercial ornonprofit coverage; and coverage through HMOs orother managed care arrangements.

In addition to national estimates, data collected inthis longitudinal study will be used to study determinantsof the use of services and expenditures, and the effects ofindividual characteristics and policy changes on medicalcare use and expenses. These behavioral analyses willinclude studies of:

• Social and demographic factors such as employmentand income.

• Methods of financing health care and healthinsurance.

• Health habits, lifestyles, and behavioral patterns ofindividuals and families.

• Health needs of specific subpopulation groups ofcurrent or potential policy interest, such as the elderlyand members of racial or ethnic minorities.

Finally, data collected in this survey–in conjunctionwith data from the 1977 NMCES and the 1987NMES–will be used to study trends in the nature anddistribution of national health expenditures, sources ofcare, and amounts and types of services used by the U.S.noninstitutionalized population.

���

���

���

���

���

���

���

���

���

������������������

���

���

���

���

���

���

���

���

���

������������������

yyy

yyy

yyy

yyy

yyy

yyy

yyy

yyy

yyy

zzzzzzzzzzzzzzzzzz

{{{

{{{

{{{

{{{

{{{

{{{

{{{

{{{

{{{

||||||||||||||||||

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

���

yyy

yyy

yyy

yyy

yyy

yyy

yyy

yyy

yyy

zzz

zzz

zzz

zzz

zzz

zzz

zzz

zzz

zzz

{{{

{{{

{{{

{{{

{{{

{{{

{{{

{{{

{{{

|||

|||

|||

|||

|||

|||

|||

|||

|||

5

Table 1. Panel design for the MEPS Household Component, 1996 and 1997

Item Calendar year Calendar year Calendar year1996 1997 1998

1996 panel (from 1995 NHIS) Round 1 Round 2 Round 3 Round 4 Round 5 Round 6

Field period 3/96-7/96 8/96-11/96 2/97-5/97 8/97-11/97 2/98-5/98 6/98-7/98

Responding households 9,500 9,000 8,800 8,500 8,300 8,100

1997 panel (from 1996 NHIS) — — Round 1 Round 2 Round 3 Round 4

Field period — — 3/97-7/97 8/97-11/97 2/98-5/98 8/98-11/98

Responding households — — 5,800 5,500 5,400 5,200

Total responding households 9,500 9,000 14,600 14,000 13,700 13,300

Note: MEPS is Medical Expenditure Panel Survey. NHIS is National Health Interview Survey.

Table 2. Design features of the MEPS Household Component, 1996 panel

Feature 1995 1996 1997 1998

Data collection Preliminary Round 1 Round 2 Round 3 Round 4 Round 5 Round 6

contact

Reference period

—

1/1/96 to date Date of Round 1 Date of Round 2 Date of Round 3 Date of Round 4

—of Round 1 interview to date interview to date interview to date interview to

interview of Round 2 of Round 3 of Round 4 12/31/97

interview interview interview

Field period 12/95-1/96 3/96-7/96 8/96-11/96 2/97-5/97 8/97-11/97 2/98-5/98 6/98-7/98

Interview mode Mail and In-person, CAPI In-person, CAPI In-person, CAPI In-person, CAPI In-person, CAPI Telephone

telephone

Note: CAPI is computer-assisted personal interview. MEPS is Medical Expenditure Panel Survey.

Source: Agency for Health Care Policy and Research, Center for Cost and Financing Studies

U.S. Department of Health and Human ServicesPublic Health ServiceAgency for Health Care Policy and Research2101 East Jefferson Street, Suite 501Rockville, MD 20852

AHCPR Pub. No. 97-0026July 1997