Statistical Aspectsof PhysicianPayment SystemsA Report of the NationalCommittee on Vital andHealth Statistics

Findings and conclusions concerningcurrent and anticipated needs of users ofdata from the patient-physician encounter inthe ambulatory care setting are summarized

in this report. The report emphasizes theincreasing plurality of data uses and of sitesof care in the ambulatory arena and

underscores the value of standard

definitions to facilitate comparison of

different data bases across systems and toachieve as much uniformity as possible at

the national level.

Documents and Committee ReportsSeries 4, No. 24

DHHS Publication No. (PHS) 87–1 461

U.S. Department of Health and HumanServices

Public Health Service

National Center for Health StatisticsHyattsville, Md.

January 1987

Copyright information

All material appearing in this report is in the public domain and may bereproduced or copied without permission; citation as to source, however, is

appreciated.

Suggeatad Citetiin

National Center for Health Statistics: Statistical aspects of physician payment

systems. Vita/ and Health Statistics. Series 4, No. 24. DHHS Pub. No. (PHS)87-1461. Public Health Service. Washington. U.S. Government Printing Office,Jan. 1987.

Library of (kngraae Cateloging4n-Publication Data

Statistical aspects of physician payment systems.(Series 4, Documents and committee reports : no. 24)

(DHHS publication ; no. (PHS) 87-1461)Prepared by the Subcommittee on Statistical Aspects of Physician Payment

Systems.Supt. of Dots. no.: HE20.6209:4/241. Ambulatory medical cere-Utilization-Reporling-United States.

2. Medical record linkeg~United States 1. United States. National Committee

on Vital and Health Statistics. Il. United States. National Committee on Vital andHealth Statistics. Subcommittee on Statistical Aspects of Physician Payment

Systems. Ill. National Center for Health Statistics (U.S.) IV. Series: Vital andhealth statistics. Series 4, Documents and committee reports ; no. 24.V. Series: DHHS publication ; no. (PHS) 87-1461. [DNLM: 1. Ambulatory

Car%conomics-UniWd States. 2. Insurance, Health, Reimbursement—United states. W2AN148vd no. 24]HA37.U1 693 no. 24 362.1 ‘0723 S 86-600375

[RA408.5] [362.1 ’72]ISBN 0-6406-0357+

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402

National Center for Health Statistics

Manning Feinleib, M. D., Dr. P.H., Director

Robert A. Israel, Deputy Director

Jacob J. Feldman, Ph. D., Associate Director for Analysis and

Epidemiology

George A. Schnack, Associate Director for Data Processing and

Services

Alvan O. Zarate, Ph. D., Assistant Director for International

Statistics

Peter L. Hurley, Acting Associate Director for Interview and

Examination Statistics

Stephen E. Nieberding, Associate Director for Management

Gail F. Fisher, Ph. D., Associate Director for Program Planning,Evaluation, and Coordination

Monroe G. Sirken, Ph. D., Associate Director for Research and

Methodology

Peter L. Hurley, Associate Director for Vital and Health Care

Statistics

Preface

This report is the culmination of 2 years of work bythe National Committee on Vital and Health Statistics(NCVHS) Subcommittee on Statistical Aspects of PhysicianPayment Systems and its predecessor work group. The NCVHSis a legislatively mandated advisory committee to the Secretaryof the Department of Health and Human Services. The reportwas approved by the full NCVHS at its meeting on June 6,1986, and was submitted to the Assistant Secretary for Healthin July 1986.

National Committee on Vital andHealth Statistics

CHAIRMANRonald G. Bkmkenbaker, M.D. (1987)Vice President for Medical AffairsSt. Vincent Hospital and Health Care Center2001 West 86th StreetIndianapolis, Indiana 46260

EXECUTIVE SECRETARYGail F. Fisher, Ph.D.Associate Director for Program Planning, Evaluation,

and CoordinationNational Center for Health Statistics3700 East-West HighwayHyattsville, Maryland 20782

Robert E. Windom, M.D.Assistant Secretary for HealthRoom 7 16G, Hubert H. Humphrey Building200 Independence Avenue, SW.Washington, D.C. 20201

Manning Feinleib, M. D., Dr.P.H.Ex OfficioDirector, National Center for Health Statistics3700 East-West HighwayHyattsville, Maryland 20782

Membership

Elizabeth L. Barrett-Connor, M.D. (1987)Department of Community and Family MedicineSchool of Medicine, University of California

at San Diego, M-007La Jolla, California 92093

William R. Felts, Jr., M.D: (1987)Professor of MedicineGeorge Washington Univ. Medical Center2150 Pennsylvania Ave., NW., Rm. 405-CWashington, D.C. 20037

Ms. Mary Anne Freedman (1989)Director, Division of Public Health StatisticsVermont Department of HealthP.O. Box 70Burlington, Vermont 05402

Lawrence W. Green, Dr.P.H. (1987)Director, Center for Health Promotion Research

and DevelopmentUniv. of Texas Health Science CenterHouston, Texas 77225

Jerome H. Grossman, M.D. (1988)Chairman and Chief Executive OftlcerNew England Medical Center, Inc.Boston, Massachusetts 02111

Mr. James K. Hutchison (1989)Chief ActuaryThe Blue Cross/Blue Shield Association676 North St. Clair StreetChicago, Illinois 60611

Carmault B. Jackson, Jr., M.D. (1987)Medical Advisory Services16902 Hidden Timber WoodSan Antonio, Texas 78248

William H. Kirby, Jr., M.D. (1989)401 Walpole CourtTimonium, Maryland 21093

George C. Myers, Ph.D. (1988)Center for Demographic StudiesDuke University2117 Campus DriveDurham, North Carolina 27706

Lloyd F. Novick, M. D., M.P.H. (1988)Director, Center for Community ServicesNew York State Health DepartmentGovernor Nelson Rockefeller Empire State PlazaComing Tower Building, Room 1492Albany, New York 12237

John D. Reid, Ph.D. (1989)Graduate Professor of SociologyHoward University

Mailing Address:Versailles Plaza2400 Queens Chapel Road, Apt. #105Hyattsville, Maryland 20782

iv

William J. ScanIon, Ph.D. (1988)Co-DirectorCenter for Health Policy StudiesDept. of Community/Family MedicineGeorgetown Univ. School of Medicine2233 Wisconsin AvenueWashington, D.C. 20007

Fernando M. Trevino, Ph.D. (1988)Associate ProfessorDept. of Preventive Medicine and Community HealthThe Univ. of Texas Medical BranchGalveston, Texas 77550

Karel M. Weigel, R.R.A. (1989)Coordinator, Records UnitDept. of Medical Statistics and EpidemiologyMdyo Clinic200 SW. First StreetRochester, Minnesota 55905

Members retired June 1986

Robert H. Barnes, M.D. (1986)Chairman1014 East Blaine StreetSeattle, Washington 98102

Mr. Theodore Allison (1986)Vice PresidentGovernment and Industry RelationsMetropolitan Life Insurance CompanyOne Madison AvenueNew York, New York 10010

Mr. Walter P. Bailey (1986)Chief, Office of Cooperative Health and DemographyState Budget and Control BoardRempert C. Dennis Bldg., Room 3371000 Assembly StreetColumbia, South Carolina 29201

Suzanne S. Harris, Ph.D. (1986)Deputy Assistant Secretary for Food andConsumer Services

USDA Administration Building, 207WWashington, D.C. 20250

Grayson B. Miller, Jr., M.D. (1986)Division of EpidemiologyVirginia Department of Health109 Governor StreetRichmond, Virginia 23219

Subcommittee onStatisticalAspects ofPhysician PaymentSystems

Members

WNiam R. Felts, Jr., M.D., ChairmanTheodore AllisonCarmault B. Jackson, Jr., M.D.

Staff

Marjorie S. GreenbergOffIce of Program Planning, Evaluation and CoordinationNational Center for Health Statistics3700 East-West HighwayHyattsville, Maryland 20782

Alan BradtBureau of Data Management and StrategyHealth Care Financing AdministrationMeadows East Building6325 Security BoulevardBaltimore, Maryland 21207

Contents

—.I-’reface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

NationalCommitteeonVitalandHealth Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Subcommittee on Statistical As~ctsof Physician Payment Systems . . . . . . . . . . . . . . . . . . . . . . . . . .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Activities of the Work group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Activities of the Subcommittee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

General tindings and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Appendixes

I. UniformAmbulatoryMedicalCareMinimumDataSet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .II. Minutes ofthe NCVHS Subcommittee on Statistical Aspects ofPhysician Payment Systems . . . . . . . . . . . .III, Placeofservicedesignations(FiguresII-VI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .IV, Data flow fortraditional fee-for-semice physicim ~imbumement (Hgu~VII) . . . . . . . . . . . . . . . . . .

...111

iv

v

1

2

3

4

5

11

13142933

vii

Statistical Aspects ofPhysician PaymentSystems

Introduction

A number of forces are currently converging to increaseinterest in ambulatory care data and stimulate reevaluationof the Uniform Ambulatory Medical Care Minimum DataSet, The changing patterns for delivering and financing patientcare are all having an impact on the ambulatory care arena.Procedures p~viously performed exclusively within hospitalsincreasingly are now takhg place on an ambulatory or outpa-tient basis. The technological advancements and changes ininsurance coverage which have made these shifts possiblehave stimulated the growth of various alternate care sites,such as free-standing surgical centers and urgicenters, as wellas the expansion of existing hospital outpatient services. Thepublic and private commitment in this decade to health promo-tion and disease prevention has been reflected in a growinginterest in primary care and self-help involving ambulatorycare services. The emphasis in care for our aging populationis alsoon independentlivingandambulatorycare.

The Medicare Prospective Payment System (PPS), enactedfor hospitalsin 1983,has put increasedpressureon institutionsto reduce inpatient stays and costs per case. Under PPS andother cost containment measures, efforts to shift care to outpa-tient, and potentially less costly, sites are likely to continue.Recent national statistics show that for the first time since1965 hospital care expenditures have declined as a share of

total health care dollars, from 41.9 percent in 1983 to 40.8percent in 1984. For Medicare, this shift from inpatient tooutpatient care has significant fiscal implications, becausethe trust fund for hospital care is financed exclusively throughSocial Security taxes, whereas physician services are financedthrough a combination of beneficiary premiums and generalrevenues and involve considerably more beneficiary cost shar-ing. Adequate and comparable statistical data will be necessaryto track and assess these possible shifts.

The emerging interest in ambulatory care services hascoincided with the growing focus on reimbursement for physi-cian services. The legislation enacting PPS required the Depart-ment of Health and Human Services (DHHS) to study andreport on the advisability and feasibility of covering inpatientphysician services within the hospital’s prospective paymentrate. Because physician services, inpatient and outpatient,now generally are reimbursed by Medicare on a fee-for-servicebasis, and because an episode of care frequently includesboth types of physician services, the broadening of concernto the full spectmm of physician services has followed natur-ally. Current approaches being considered by the Departmentfor reimbursing physician inpatient services all have ‘implica-tions for ambulatory care services as well.

1

Background

Recognizing the numerous statistical issues involved incurrent and future policy choices for delivering and reimbursingphysician services, the National Committee on Vital and HealthStatistics (NCVHS) formed a work group in May 1984 togather further information and to make recommendations tothe full Committee. Originally charged to examine statisticalaspects of a pre-payment system for physician services, atits first meeting the work group expanded its investigationto encompass the adequacy of any data systems related tothe provision of physician services, whether designed for reim-bursement, planning, or research. It was decided to give par-ticular attention to services rendered in ambulatory care settingsand to the information needed about these services. The ulti-mate goals of the work group’s inquity were to encouragecomparability and standardization; to enhance the multipleutility of data bases; to assure that data requirements by third-party payers and others were justified; and to prevent unneces-sary duplications. Underlying the effort was the convictionthat in times of decreasing resources, high quality statisticsbecome increasingly important in making optimal allocationdecisions. The work group was also aware that there wouldbe growing pressure on the analysis of ambulatory care dataand that development of standard definitions, adoption ofthose standards, and training in their application would benecessary first steps toward turning that data into usefulinformation.

The NCVHS had contributed to the standardization ofhospital data through the development of the Uniform HospitalDischarge Data Set (UHDDS) in the early 1970’s and therecent reassessment of the UHDDS in 1984. The UHDDShas been approved for use in Department of Health and Human

Services (DHHS) programs since 1974 and is widely acceptedand implemented in the public and private sectors.

Developments in standardization of ambulatory care datahave been considerably slower, even though ambulatory careis the largest component of the health care system in thiscountry, accounting for approximately 1 billion contacts withmedical doctors per year. Traditionally, ambulatory care rec-ords have been far less standardized than records of hospitalcare, and patient complaints and episodes of care have beenless clearly defined and documented. There has been no parallelin the ambulatory care field to the private sector developmentof abstracting services for inpatient care, although computerapplications to physician offices have proliferated in recentyears.

Following the lead of the UHDDS, a Uniform AmbulatoryMedical Care Minimum Data Set was developed by theNCVHS in the mid- 1970’s and revised in 1980 (see appen-dix I). Reflecting the state of the art at that time, the focuswas on defining what information should be entered inpatients’ambulatory medical care records rather than on what informa-tion should be abstracted from existing records.

Although the NCVHS submitted the revised version ofthe ambulatory data set to the Secreta~ of DHHS in 1981,no official action has been taken in response to this transmittal.Although the data set has received wide distribution, it isnot known to be in current use in its entirety in any programsof the Department. In 1983, the DHHS Health InformationPolicy Council (HIPC) gave reexamination of the data seta lesser priority than the review of the UHDDS and the Long-Term Care Minimum Data Set.

Activities of the work group

The work group formed by the full Committee met onthree occasions between June and October 1984 to obtaininformation on current and proposed Departmental data ac-tivities and to identify issues that would merit further NCVHSstudy and assessment. Because physician payment systemsare a major source of data on ambulatory services, and Medi-care requirements often influence activities in the private sec-tor, detailed information was sought and obtained on the datarequirements related to Medicare reimbursement for physicianservices. The work group received extensive background onthe Medicare data requirements, the data bases maintainedby HCFA at the national level, and studies underway to assessphysician reimbursement methods.

The review of data requirements, data bases, studies,and data sets by the work group focused on the followingfive areas: patient identification, physician identification, placeof service designation, diagnostic information, and procedurecoding systems. Many of the issues explored, although specificto Medicare, represented generic concerns that had surfacedduring the earlier minimum data set development.

The work group concluded that the evolving activitiesrelated to changes in methods for physician reimbursement,particularly in the ambulatory care setting, are likely to havesignitlcant implications for each of these data areas and forthe comparability of data through time and across health caresettings. Continuing liaison among the Health Care FinancingAdministration (HCFA), the National Center for Health Statis-tics (NCHS), and the NCVHS concerning the data systems

to support these activities was considered desirable, at leastuntil the systems have been fully defined. Therefore, thework group recommended to the NCVHS at its meeting inDecember 1984 that a subcommittee be established to providethis liaison and address the following tasks as its charge:

. Determine more clearly the specific needs of users ofdata from patient-physician encounters in the ambulatorycare setting.

● Develop a schematic overview of the flow of data fromvarious ambulatory settings into the multiple data bases.

. Define better the different sites of care in the ambulatorysetting and the types of services delivered so that under-standing of data requirements can be improved.

The results of these tasks were to be input into the possiblereview and revision of the Uniform Ambulatory Medical CareMinimum Data Set.

The National Committee accepted the recommendationsof the work group and designated a Subcommittee on StatisticalAspects of Physician Payment Systems, consisting of WilliamR. Felts, Jr., M. D., Chairman, Professor of Medicine atGeorge Washington University Medical Cente~ Theodore A1li-son, Vice President for Government and Industry Relations,Metropolitan Life Insurance Company; and Carrnault B.Jackson, Jr., M. D., Medical Advisor for Baptist MemorialHospital System in San Antonio, Texas. Staff from NCHSand HCFA were requested and assigned to work with theSubcommittee.

Activities of the Subcommittee

The Subcommittee held four meetings to hear testimonyon June 19, September 10, October 15, 1985, and January14, 1986 (see minutes in appendix II), in order to pursuefurther the work group goals and to address the specific tasksincluded in its charge. Following the extensive informationreceived from the Medicare program during the work groupmeetings, the Subcommittee continued to be updated on MedLcare activities and sought to obtain comparable informationfrom other public and private insurers. The latter providedan overview of the ambulatory care data requirements andthe flow of data for the patient-physician encounter in theirrespective programs or organizations, discussed uses of dataand data problems, and described any work currently underwayto resolve data inadequacies and improve data quality. Re-searchers and planners who use data from the patient-physicianencounters also contributed to the discussions. Dr. Felts andDr. Jackson, both of whom have had extensive involvementwith provider data systems and strong afilliations with nationalprofessional organizations, offered the additional perspectiveof practicing physicians. A representative of the AmericanMedical Association and other observers representing profes-sional associations also attended the sessions.

The purpose of these meetings was to:

. Gather information related to the Subcommittee’s charge.

. Anticipate changes in physician payment systems thatmay require collection and analysis of additional itemsof data.

. Examine the possible impact of these changes and addi-tions on availability, quality, and comparability of data.

. Analyze the applications of data for trend analysis, re-search, and program administration.

The meetings also served to bring together interested organiza-tions in the public and private sectors to share mutual concernsand seek common solutions.

The organizations, grouped by type, that have made pre-sentations to the Subcommittee and the primary focus of theirremarks follow:

Organization Primary focus

Public insurers and data users

Civilian Health and Medical ● Program requirementsProgram of the Uniformed . Ambulatory care data needsServices (CHAMPUS),Department of Defense

Centera for Disease Control

Department of the Army,Department of Defense

Medicare Program, HCFA

National Center for HealthStatistics

Medicaid Program, HCFA

Pennsylvania Department ofPublic Welfare

Peer Review OrganizationProgram, HCFA

Veterans Administration

. Ambulatory care data needs forepidemiologic surveillance

. Ambulatory Care DataBase Project

● Claims processing● Statistical system

. National Ambulato~ MedicalCare Survey

● Ambulatory Visit Groups

. National Medicaid repottingsystem

● State Medicaid program datarequirements

. Data needs and plans

● Program requirements● Ambulatory care data needs

Commercial insurers

Metropolitan Life Insurance ● Claims processingCompany . Uses of claims data

Prudential Life Insurance ● Medicare and Medicaid carrierCompany dafa requirements

. Uses of claims data

Blue Cross Blue Shield

Blue Cross and Blue Shield ● Data needsAssociation ● Uses of claims data

National Capital Area Blue ● Claims processingCrose Blue Shield . Uses of claims data

Health Maintenance Organizations

Group Health Association of . Current data requirements ofAmerica different HMO model types

● Data needs

Health planners

United Hospital Fund . Collection and use of hospitaloutpatient data

Self-insured employar

Honeywell, Inc. ● Data needs of employers andbusiness coalitions

● Uses of data

The Subcommittee also requested the views and perspectives ofthe House and Senate Subcommittees on Health but was unableto arrange a mutually convenient meeting time. Information onactivities was sent to the House subcommittee, at the latter’srequest.

General findings andconclusions

Several themes emerged from the numerous presentationsheard by the Subcommittee:

. All presenters recognized the increasing need for ambula-tory care data,

After two decades of focusing almost exclusivelyon the utilization and cost of inpatient care, public andprivate insurers are beginning to turn their attention tooutpatient services. The Medicare Program has given con-siderable priority to the development of the data systemfor Part B, which reimburses for physician services, dur-able medical equipment, and laboratory services. ThePart B Medicare Data System (BMAD) includes benefi-ciary, provider, procedure, and prevailing charge files.The Subcommittee believes that this data base representsthe most important step at the national level to standardizeand use aggregate data on physician services for policyevaluation, The Peer Review Organizations (PRO’s),while currently reviewing hospital care for Medicare andMedicaid beneficiaries, strongly recognize the need toexamine pre-care and after-care in the ambulatory caresetting. Commercial insurers and Blue Cross Blue Shieldplans are following the lead of the Medicare Programin addressing data needs from the patient-physician en-counter. A self-insured corporation reported increased in-terest in examining ambulatory care data, noting that itsexpenditures for ambulatory care have shifted from 30percent of total dollars spent to 55 percent in the past2 years. The Department of the Army has undertakena study in six sites with the purpose of establishing amajor new reporting system for ambulatory services. TheUnited Hospital Fund has recently completed a studyof hospital-based ambulatory care visits in New YorkCity, of which there are between 10 and 12 million annu-ally. One Medicare carrier has initiated a study, withHCFA support, of overutilization of physician office visitsby beneficiaries.

In the area of epidemiologic research, both the Centersfor Disease Control and the Department of the Armyemphasized the importance of ambulatory care data.

. Demands upon ambulatory care data systems for moreextensive data elements and more comparative analyseswill continue to increase. Those collecting and analyzingthe chta must exercisecautionthat the qualityand contextof the data merit the interpretationsthat are made. Thecommunity of users must assure accurate, meaningful,and responsible reporting.

Many presenters predicted that data will be soughtto perform more complex and sophisticated analyses oncethe basic systems are in place. Subcommittee membersexpressed concern that policymakers may attempt to sub-ject data to uses for which the data were not originallydesigned. Examples were cited of release of aggregatedata on physician reimbursement under public programsin the late 1970’s and more recently on hospital mortalityrates for Medicare patients, each of which resulted inerroneous and potentially damaging conclusions. Data in-accuracies and inappropriate use of raw data were bothat issue. Relevant public and private sector organizationsmust set standards not only for data definitions but alsofor data quality, adequate sample sizes, and meaningfulcomputations. Public organizations collecting potentiallysensitive information may need stricter confidentiality leg-islation to protect raw data files adequately from misuse.

. There are both many commonalities and some significantdifferences in the needs of data users. The data variationsstem primarily from administrative requirements.

First, current ambulatory care data requirements differconsiderably depending on the reimbursement system.Perceived data needs similarly differ widely among insur-ers. The fee-for-service environment, whether in the publicor private sector, requires considerably more encotmter-level data than a cavitation system for payment. Evenwithin the Health Maintenance Organization (HMO) ap-proach, the particular model of HMO impacts significantlyon the type of data collected. Second, each system hasspecific and often unique administrative requirements re-lated to determining eligibility and coverage of servicesand coordinating benefits. These are the most variableitems collected by the different data systems but are crucialfor benefit management.

. The data currently collected by public and private insurersmeet most of their current needs for reimbursement butmay be quite inadequate for alternative reimbursementmethods. Third party payers are interested in reviewingquality of care and in conducting relevant research, butthe basic data needed for processing and paying claimscontinue to govern their data requirements. and data aremore limited for other uses.

Making changes and adding items to a claim formcan be very costly to the insurers and providers. ThePennsylvania State Medicaid program reported that chang-ing invoices costs about a million dollars. The Group

5

Health Association of America reported that the type ofdata system used in fee-for-service plans would be onerousand drive up the costs of an HMO, reducing its competitiveposition. At the same time, public programs and employersare finding the data available from cavitation systemsinadequate to meet their needs.

● Many public programs and private insurers and their tradeorganizations currently are undertaking efforts to achievegreater uniformity within their own data systems.

HCFA reported that in the past 5 years the programhas made two major changes that have enhanced theuniformity of data collection in the Medicare program.First, the HCFA Common Procedure Coding System(HCPCS) was adopted for coding of procedures. Second,a common claims form, the HCFA 1500, was institutedfor all physician-submitted claims. Further uniformity isto be achieved as regulations have been published torequire Medicaid State agencies to use both HCPCS andthe 1500 by the end of 1986. The Medicaid programhas also undertaken a tape-to-tape project with the goalof standardizing unit record data in a common formatacross States.

The Group Health Association of America (GHAA)is developing a summary level database for Health Mainte-nance Organizations using standardized definitions. Thiseffort is in partial response to the lack of uniformityin source documents and definitions used for encounterlevel data even within the same HMO system.

The Veterans Administration is in the process of auto-mating its clinical data system and moving toward collec-tion of more detailed information on demographic vari-ables and services rendered on 100 percent of its patients.The CHAMPUS and DOD programs are also examiningchanges in their outpatient data systems.

● Presenters also supported the concept of standard defini-tions to facilitate comparison of different databases acrosssystems and to achieve as much uniformity as possibleat the national level. There is particularly wide supportfor standardization of coding systems and definitions ofunits of patient care. Less support was expressed forstandardized claim forms, although many data collectorsand users recognize their advantages.

The United Hospital Fund reported the difficulty incollecting data on outpatient department and emergencyroom visits in New York City due to inconsistent defini-tions used by hospitals across the city. A self-insuredcorporation actively involved in a business coalition de-scribed the significant challenge of developing a standard,uniform data base from a variety of indemnity carriers,Blue Cross Blue Shield plans, and HMO programs. TheCHAMPUS program cited as a major problem the incom-patibility between data from its system and data fromthe military hospital ‘system.For example, the CHAMPUSprogram defines a visit as a face-to-face encounter witha health care professional, whereas the military systemincludes telephone contacts as encounters. GHAA notedsimilar inconsistencies among HMO’s.

Most presenters concluded that the recent efforts un-dertaken by HCFA have fostered standardization in theindustry. The commercial insurers and Blue Cross BlueShield plans that testified have adopted HCPCS, whereasmost of the other public and private insurers reported cur-rently to be using CPT-4, on which HCPCS is based.A number of public and private insurers are using theHCFA 1500 or a variant of this form. The Uniform ClaimForm Task Force, chaired jointly by HCFA and the Ameri-can Medical Association with representation from the BlueCross and Blue Shield Association and the Health InsuranceAssociation of America, agreed in June 1985 to pursuegreater standardization of the claim form. At the sametime, the Task Force identified approximately 14 additionalnew items for consideration, many of which relate to deter-mining eligibility and coordination of benefits.

The Pennsylvania State Medicaid Program reportedconsiderable resistance in that State to use of the HCFA1500 and argued strongly for common definitions ratherthan common claim forms. The latter, it was noted, canbe very difficult to use, because they contain unnecessaryitems for some insurers and the same item can havemultiple definitions depending on the insurer. The HCFAstaff estimated that approximately 75 percent of the Statescan use the 1500. Representatives of the NationalMedicaid Program and the self-insured corporation bothnoted that flexibility and decentralization of public andprivate programs are often at odds with uniformity.

It was clear to the Subcommittee that there are signifi-cant benefits to achieving as much conformity as possiblebut that the variable administrative items must be accom-modated by any common form and preclude totaluniformity.

The Department of the Army has pursued encounterform development, because its needs are performance meas-urement and epidemiologic research rather than claims pay-ment. Under a pilot study to establish a new ambulatorycare data collection and reporting system, the Departmenthas developed a series of forms with common administrativeand demographic data but with diagnoses and procedures,coded in ICD-9-CM and ~-4, respectively, tailoredto particular specialties.

● The increasing demands of employers for data from thereimbursement systems may force greater standardizationof data.

Insurers reported that employers are becoming veryaggressive with cost containment measures and are provid-ing a major impetus for encouraging use of outpatientsettings, As a result, they are demanding comparabledata, which will permit them to make informed decisionsabout carriers and benefh packages as well as to assessthe use and outcome of outpatient services delivered totheir beneficiaries. In addition, the growing number ofbusiness coalitions are interested in sharing data frommany different employers and insurance companies. Theself-insured corporation, in its presentation, confirmedthese needs and demands of employers and business coali-

6

tions and described two related activities. First, the corpo-ration is developing a competitive medical arrangementrequest for proposal, with a management information sys-tem requirement, for local divisions to use in solicitingbids from Preferred Frovider Organizations. Second, theMinnesota Coalition on Health Care Costs is developinga uniform data reporting request from the employers tothe HMO’S.

. Even as public and private insurers move toward greateruse of cavitation, some forms of which require consider-ably less data for reimbursement, management informationfrom the encounter level will continue to be needed toassess quality, efficiency, cost, and accessibility of care.

Public and private insurers, the Peer Review Organi-zations, employers, and business coalitions are all con-cerned about the current and future availability of datafrom HMO’s and similar capitated systems. Presentersgenerally concluded that data similar to what is beingcol!ected currently from the fee-for-service systems willbe needed but that it will be more difficult to obtainoutside of a billing mechanism, There was general agree-ment that tying data to reimbursement does improve over-all reporting, even though it may influence the way somedata are reported, The representative from the GHAAacknowledged that Government and employers are de-manding more utilization information about their enrol-lees, but urged development of a data system specificto HMO’s rather than applying the fee-for-service systemto the HMO environment.

. Although data needs for operational purposes may becomemore decentralized under cavitation plans, there remainneeds for data at the national level to evaluate policyoptions, recommend program changes, and monitor per-formance. In a dynamic situation such as currently existsin the health care arena, data can provide a clue as towhether specific medical interventions and reimbursementmechanisms are having a positive, negative, or neutralimpact on the outcome of medical care.

It was suggested that encounter-level data will stillbe available at the carrier or local level, but it was clearthat this will not completely substitute for national leveldata, In addition to meeting needs for program develop-ment and monitoring responsibilities, data of adequatebreadth and coverage are needed to address outcomeissues.

. Some of the needs for policy-relevant information andhealth services research can be met by survey data, andsuch survey mechanisms must be supported. Current sur-vey data on ambulatory medical care are collected onan intermittent basis and are quite limited in their coverageof different types of outpatient facilities.

NCHS reported on the National Ambulatory MedicalCare Survey (NAMCS), which operated continuously asan annual survey from 1973 until 1981, when it wasplaced on a triennial basis due to resource constraints.After an additional 1-year delay, the NAMCS was mostrecently conducted in 1985, collecting information on

approximately 75,000-80,000 encounters from a sampleof 5,000 office-based physicians. Funding is not containedin the FY 1987 budget for the next cycle of NAMCS,originally scheduled to be conducted in 1988. AlthoughNCHS has done research on expanding the suwey beyondoffice-based physicians to include hospital-based physi-cians, funding also has not been obtained for this compo-nent. The Centers for Disease Control (CDC) reportedthat from their perspective NAMCS is an important na-tional resource and should be maintained at least on itscurrent triennial basis. The Subcommittee confirmed thatfurther survey delays will interrupt the ability to evaluatetrends and changes in ambulatory care. The Subcommitteealso noted that the information obtained by surveys differsfrom that acquired in remuneration data bases and, there-fore, is not redundant but complementary.

HCFA also reported on a number of surveys andspecial studies either underway or being planned to helpinform policy decisions. Presenters and Subcommitteemembers encouraged these activities to expand the infor-mation base concerning dhTerent policy options.

. It is timely to undertake a full review of the adequacyof the Uniform Ambulatory Medical Care Minimum DataSet. The current data set defines those items that shouldbe entered in the records of all ambulatory health care.A revision should bring the data set into accordance withthe Uniform Hospital Discharge Data Set by also specify-ing those items that should be abstracted uniformly fromexisting records.

The increasing interest in comparable ambulatory caredata bases with standardized definitions, evidenced bythe presentations made before the Subcommittee, supportsthe reassessment of the Uniform Ambulatory MedicalCare Minimum Data Set at this time. Significant changesin the care delivered in ambulatory settings and in therequirements for data have occumed since the last reviewand revision.

Although it was not the charge of the Subcommittee toundertake a systematic review of the Uniform AmbulatoryMedical Care Minimum Data Set nor to develop a full revision,the extensive presentations received by the Subcommittee didaddress many of the key items in the data set. The Subcommit-tee’s findings and conclusions related to these items follow:

. Patient identificationMuch of the testimony was supportive of a unique

patient identifier across the health care system whichwould facilitate managed care and linkage of records.Subcommittee members recognize both the advisabilityof a unique patient identifier and the diftlculty of im-plementation. Confidentiality and privacy concerns makethis an extremely complicated issue which requires moredeliberation.

All Medicare records contain the same unique iden-tifier for each individual served: the Social SecurityNumber with an associated beneficiary ID code. Medicaid,CHAMPUS, and private insurers collect the patient’s namebut do not have unique identifiers for the insured and

7

his or her dependents. Furthermore, other record systems,such as those used by institutions, frequently use theirown identifiers and do not retain the patient’s Medicare,Medicaid, or insurance number. The new nationalMedicaid reporting system will require all States reportingon tape to provide a unique recipient ID number. Althoughthe Social Security Number (SSN) is not mandated, theprogram notes that the SSN is a requirement for eligibilityand predicts its eventual availability on the data files.

Unique patient identifiers are desirable to track pa-tients, profile utilization patterns, link episodes of carein the same setting and across different settings, andcoordinate benefits. The multiple sources of care currentlyavailable to patients increase the importance of thiscapability.

. Provider identificationPresentations supported the need for a unique number

that distinguishes the provider from all other providersand is the same for the provider in all settings wherehe maybe in practice. The Subcommittee members concurwith this requirement.

In reimbursing physician services, Medicare identifiesthe billing entity, such as a group practice or outpatientclinic, rather than the individual practitioner providingthe services, except where the practitioner is in solo privatepractice where the practitioner is individually identified.Thus, the same practitioner may bill for services underseveral different identifiers, and a single identifier canrepresent many different practitioners. This system makeslinkage of an individual’s total practice patterns virtuallyimpossible, even within the Medicare program. HCFAis responding to this problem by developing a plan fora HCFA-assigned number (HAN) which would providea unique identifier for each physician rendering services.

The new national Medicaid reporting system containsa provider file but it was reported that the identifyinginformation on this file is not considered reliable andwill require a major cleanup. Unique provider ID numbersare considered particularly problematic in the ambulatorycare setting.

The CHAMPUS program also reported problems withidentifying individual providers. The program would liketo be able to identify specific individual providers byspecialty in its data system. The pilot Ambulatory Report-ing System under development by the Department of theArmy does include a unique provider identifier.

● Place of serviceThe presentations before the Subcommittee frequently

addressed the increasingly different settings for servicesdeveloping in the health care system. The Subcommitteeconcluded that a consensus group, such as the CPT Edito-rial Board or the NCVHS, should evaluate the definitionscurrently in use for place of service and determine theextent to which standardization is feasible. Surveymechanisms, such as the NCHS National AmbulatoryMedical Care Survey, should continue to be supportedto address some of the research questions, with attentionto more specificity of sites.

The distinctions between institutional and ambulatorycare and the definitions for different types of care settingsare frequently unclear. Services provided in a hospitalemergency room or out-patient clinic are often identicalto those provided in inpatient settings. Some free-standingfacilities substitute for acute inpatient care while othersare extensions of private physician offices.

In some cases, the specific place of service mustbe designated for reimbursement purposes. More broadly,reasonable specificity is needed for management and re-search activities. Place of service designations vary widelyamong insurers. Medicare, Medicaid, commercial insur-ers, and Blue Cross Blue Shield all have their own setsof definitions (see appendix III). The HCFA 1500 has16 different place of service codes, but the BMAD instruc-tions condense these down to 10 codes. The Blue Crossand Blue Shield Association reported 34 different placeof service designations a Plan can use in processing claims,although the place of service codes most commonly filledin by the person completing the claim form are morecomparable to those used on the HCFA 1500. A commer-cial insurer presented 19 facility designations with 3 clinicmodifiers.

CHAMPUS does not currently collect place of en-counter, but program officials feel it would be beneficial,particularly in the area of ambulatory surgery. TheMedicaid program reported that it has always collectedplace of service of physician visits but has encountereda great deal of difficulty with standard definitions acrossStates and has essentially dropped that requirement.

. DiagnosisPresenters recognized the value of accurate recording

of patient problems and diagnoses, but reported that thisinformation is frequently unavailable in ambulatory carereimbursement systems and often is of questionable qualitywhere available. The Subcommittee reaffirmed the impor-tance of capturing information on all patient problemsand diagnoses requiring attention at the encounter, soas to make judgments on medical necessity and appropri-ateness of services, plan resource allocation, and carryout other policy analyses. The members concluded thatthe development of guidance and instmctions on collectingand coding diagnoses in the ambulato~ setting shouldbe a consensus activity, and that the Federal government,as an interested participant, should be an organizer ofthis activity.

The Subcommittee fin-therdetermined:The International Classification of Diseases and its

clinical modification (currently ICD–9–CM) is widelyused by systems coding diagnoses and should continueas the standard coding convention for this purpose. Thereshould be a consensus process for developing interimcodes which can be utilized in a timely manner by indi-vidual countries between revisions to identify newly recog-nized conditions and reflect new knowledge.

Separate information on the patient’s stated reasonfor encounter can be useful for planning, administrative,

and research purposes. The ICD should be the basis forany reason for encounter classification system.

Different definitions currently exist in the inpatientand outpatient environments for designating the first-listeddiagnosis in a data system as either principal or primary.The terms are sometimes used interchangeably but inother cases have very different meanings. The selectionof a particuhm definition generally reflects the purposefor which the data element will be used. If definitionsare not clear or standardized, there is an impact on dataanalyses and trends. This problem is further compoundedby the fact that recording of secondary diagnoses alsovaries among data systems. If all relevant diagnoses orproblems are not captured by the data system, it maynot be possible to retrieve the patient’s most serious orlife-threatening condition or to identify comorbidity.

The presentations touched upon a number of importantuses of diagnostic information. Accurate diagnoses orproblem descriptions are essential for assessing the qualityof care delivered, They are needed to determine whattypes of health problems are being seen and treated inthe different types of ambulato~ care facilities and forassessing the appropriateness of the setting used to performthe services. Information on multiple diagnoses is impor-tant for developing severity indexes and assessing theresource requirements of individual cases. Further, severalproposed reimbursement reforms, such as bundling ofservices by episodes of care and prospective paymentbased on Ambulatory Visit Groups, a categorization beingdeveloped by the Health System Management Group atYale University, would require information on the pa-tient’s diagnosis. Finally, use of ambulatory care recordsfor epidemiologic surveillance, the unique function ofthe Centers for Disease Control, is dependent on therecording and reporting of diagnoses.

The presentations also revealed current limitationsand problems associated with collection of diagnoses.Under Medicare accurate designation of principal andsecondary diagnoses with ICD–9-CM codes are a neces-sity for reimbursing hospital services, whereas procedures,not diagnostic information, are the basis for reimbursingphysician services. Most Medicare carriers require thephysician or patient submitting a claim to provide a reasonfor encounter, nature of impairment, or diagnosis. TheMedicare program also plans to begin requiring carriersto have the capability to keep diagnosis in their claimsprocessing system. However, much of the diagnostic infor-mation obtained is not considered reliable by HCFA dueto the large number of beneficiary-submitted claims, andthus, carriers are not required to report the item to thenational level. Even with the new requirements, carriersprobably will only need to retain diagnosis where it isnecessary for specific reimbursement decisions. Over halfof the claims for physician services are for office visits,and the vast majority of claims are paid without referenceto diagnosis. Given the sheer volume of claims, approxi-mately 260 million in Fiscal Year 1985, the Medicare

program considers it prohibitive to match diagnoses withprocedures on a routine basis.

Diagnosis, coded in ICD-9-CM, is a “desirable”rather than required data item on the subfile for physicianclaims in the new national Medicaid reporting system.The Veterans Administration (VA) expressed considerableconcern about the accuracy of diagnostic information onoutpatient VA records and noted that this informationis not tracked. The CHAMPUS program reported thatonly the “primary diagnosis,” defined as the reason foroutpatient care, is captured by its data system and codedin ICD–9-CM. The forms include room for additionaldiagnoses but these are not transmitted on the tape. Thusthe coded diagnosis does not necessarily correspond tothe most expensive treatment or procedure, and thereis not a diagnosis associated with each procedure. Thepresenter stated that this seriously limits information forutilization review.

The current Medical Summary Reporting System usedby the Department of the Army is based on number andtype of visits and contains no coded diagnostic informa-tion. However, the Ambulatory Reporting System beinginitiated by the Department on a pilot basis requires aprimary diagnosis, coded in ICD-9-CM, and permits mul-tiple secondary diagnoses. The encounter forms used inthe system contain detailed lists of diagnoses tailoredto particular specialties as an aid to completion.

Private fee-for-service insurers report interest in col-lecting more specific diagnostic information for profes-sional services but encounter problems similar to the publicprograms. The GHAA indicated that very few HMO’srecord or code diagnosis on an encounter form in theambulatory setting and that, even on the hospital side,many HMO’s currently do not analyze diagnosis trends.

Finally, the Subcommittee members received infor-mation on the International Classification of Primary Care(ICPC) through their participation in the NCVHS Subcom-mittee on Disease Classification and Automated Codingof Medical Diagnoses. This classification was developedchiefly for the use of primary care providers, with particu-lar application to ambulatory settings, and is undergoinga number of field tests.

. ProceduresThe testimony reinforced the value of uniform coding

of procedures and the positive impact the HCFA CommonProcedure Coding System (HCPCS) has had on standardi-zation. The Subcommittee reaffirmed the NCVHS positionthat strong efforts should be made to develop a singleprocedure coding system for inpatient and ambulatorycare in the United States.

The Subcommittee further believes that HCPCS,which is a live system based on CPT4, should be thecore for a national common procedure code for physicianservices. Concerning coding of nonphysician services,the Subcommittee supports the considerable work carriedout for HCPCS and encourages continuation of this work.

Additional study on implementation of a common codingsystem for both physician and nonphysician servicesshould be undertaken.

The presentations confirmed that information is morewidely available on procedures than on diagnoses in am-bulatory care data systems, because procedures are thebasis for reimbursement in the fee-for-senice environ-ment. Accurate recording of procedures serves many ofthe same purposes described above for diagnoses, as wellas meeting reimbursement needs. Even when proceduresare recorded, however, a number of coding issues remain.

The differing Medicare reporting requirements forreimbursing hospital services, on the one hand, and forreimbursing physician inpatient and outpatient services,on the other, result in the use of two separate cod]ngsystems for the designation of procedures. Volume 3 ofICD-9-CM is required-for hospital reporting, whereas

the HCFA Common Procedure Coding System (HCPCS)is used for coding physician services, regardless of thesite in which they have been provided. In instances wherehospital billing offices also perform billing services fortheir staff physicians, both systems must be maintainedfor each procedure.

Until the recent implementation of HCPCS, therewere an estimated 120 different systems used in reportingphysician services for Medicare. Public and private insur-ers, for the most part, reported current use of either HCPCSor CPT-4 for physician services. If the latter is used,the insurer independently has extended the coding systemfor nonphysician services and to meet other unique require-ments. Coordination with HCPCS codes for nonphysicianservices has not occurred in these cases. As noted earlier,Medicaid State Agencies will be shifting to HCPCS bythe end of 1986.

Summary

The Subcommittee has addressed each aspect of its charge,as follows:

● The Subcommittee hasestablished excellent liaison withNCHS and HCFA, as well as between the public andprivate sectors, In many respects, the NCVHS seemsuniquely qualified to play this role for data issues thatcut across Departmental programs and heavily involvethe private sector.

● Through the series of meetings with representatives fromthe public and private sectors, the Subcommittee has madesignificant progress in identifying more clearly the specificneeds of users of data from patient-physician encountersin the ambulato~ care setting.

● The discussions with users have provided input to aschematic overview of the flow of data from variousambulatory settings into the various data bases in thefee-for-service environment (see appendix IV). Becausethe data systems differ considerably depending upon themethod of reimbursement, it is difficult to capture thevariations in one schema. However, the diagram doeshelp to demonstrate some of the data requirements, com-plexities, and interrelationships.

● Presentations by the various public and private insurershave described the multiple definitions and distinctionsfor ambulatory sites of care. Further work is requiredto evaluate the definitions currently in use and determinethe extent to which standardization is feasible. The Sub-committee has recommended that this activity be carried

out by a consensus group, such as the CPT EditorialBoard or the NCVHS, and be supplemented by surveymechanisms.

● The Subcommittee’s investigation has underscored thevalue of a Uniform Ambulatory Medical Care MinimumData Set. This report summarizes the Subcommittee’sfindings and conclusions related to specific items in thedata set addressed by the various presentations. Becausesome items require further study and others were notconsidered, the Subcommittee recommends a thoroughand systematic review of all items in the Uniform Ambula-tory Medical Care Minimum Data Set for the purposeof developing a revised version that meets current needs.This review and revision process should be carried outby the National Committee on Vital and Health Statisticsand, concurrently, by an interagency task force establishedwithin the Department. The interagency task force shouldalso have input to the definition of sites of care. Suchan approach would parallel the process used for thereexamination of the Uniform Hospital Discharge DataSet and the Long Term Care Minimum Data Set andshould be coordinated with these two data sets to assurecomprehensiveness, minimize redundancy, and facilitateepisode linkage. Attention should also be given to trainingin use of the data set, other approaches for assuring dataquality, and standards for reliable and meaningfulanalyses.

11

Appendixes

Contents

I. Uniform Ambulatory Medical Cme Minimum Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Section l. Summary of recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...13

Patient data items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...13Provider data items . . . . ...’. . . . . . . . . . . . . . . . . . . ... .. O.. ++. $OC$O C.. .13Encounter data items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...13

II. Minutes of the NCVHS Subcommittee on Statistical Aspects of Phvsician Payment Systems . . . . . . . . . . . . . ]4June 19,1985 meeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Medicare claims processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Medicare statistical system . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Commercial insurance data requirements . . . . . . . . . . . . . . . . . . . . . . .Ambulatory visit groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Blue Cross and Blue Shield Association (BCBSA) data needs . . . . . . . . . . . .

September 10,1985 meeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Health Maintenance Organizations (HMO’s) . . . . . . . . . . . . . . . . . . . . .

Figure I. HMO model types . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Medicare update . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Medicaid data requirements . . . . . . . . . . . . . . . . . . . . . . . . .-e-c

October 15,1985 meeting . . . . . . . . . . . . . . . . . . . . . . . ..O . . . . .National Medicaid reporting system. . . . . . . . . . . . . . . . . . . . . . . . .Dataneeds ofa Medicare cqn-ier. . . . . . . . . . . . . . . . . . . . . . . . . .Data needs of employers and business coalitions . . . . . . . . . . . . . . . . . . .Data needs of Peer Review Organizations (PRO’s) . . . . . . . , , . . . . . . . . .Dataneeds ofa Blue Cross Blue Shield plan . . . . . . . . . . . . . . . . . . . .

January 14,1986 meeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ambulatory care data needs of Veterans Administration (VA) programs . . . . . . .Ambulatory care data needs of Civilian Health and Medical Program of the Uniformedprogram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ambulato~ care data base project of the Department of the Army . . . . . . . . . .Collection and use of hospital outpatient data . . . . . . . . . . . . . . . . . . . .Ambulatory care data needs of Centers for Disease Control (CDC) programs . . . . .

111.Place of service designations (Figures II–VI) . . . . . . . . . . . . . . . . . . . . . .Figure II. HCFA 1500 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Figure III. Medicare Carriers Manual, Part 3-claims process . . . . . . . . . . . .Figure IV. Metropolitan Life Insurance Company place of service codes . . . . . . .Figure V. Blue Cross and Blue Shield Association place of service codes+laim formFigure VI. Blue Cross and Blue Shield Association place of service codes—processing

IV. Data flow for traditional fee-for-service physician reimbursement (Figure VII) . . . . .

. . . . . . . . . . . . . . 14

. . . . . . . . . . . . . . 14

. . . . . . . . . . . . . . 15

. . . . . . . . . . . . . . 15

. . . . . . . . . . . . . . 16

. . . . . . . . . . . . . . 16

. . . . . . . . . . . . . . 17

. . . . . . . . . . . . . . 17

. . . . . . . . . . . . . . 17

. . . . . . . . . . . . . . 18

. . . . . . . . . . . . . . 19

. . . . . . . . . . . . . . 20

. . . . . . . . . . . . . . 20

. . . . . . . . . . . . . . 21

. . . . . . . . . . . . . . 21

. . . . . . . . . . . . ., 22

. . . . . . . . . . . . . . 22

. . . . . . . . . . . . . . 2424

ie;ices-(CHAMPUi)” “ “. . . . . . . . . . . . . . 25. . . . . . . . . . . . . . 26. . . . . . . . . . . . . . 26. . . . . . . . . . . . . . 27

. . . . . . . . . . . . . . 29

. . . . . . . . . . . . . . 29

. . . . . . . . . . . . . . 30

. . . . . . . . . . . . . . 32

. . . . . . . . . . . . . . 32

. . . . . . . . . . . . . . 32

. . . . . . . . . . . . . . 33

Appendix LUniform Ambulatory MedicalCare Minimum Data Set[From lhe Report of the National Commitee on Vital and Health Statlstioa, April 1981]

Section 1. Summary of recommendations

The consultants of the Uniform Ambulatory Medical CareMinimum Data Set Technical Consultant Panel (TCP), Na-tional Committee on Vital and Health Statistics, recommendthat the following items constitute the minimum data set thatshould be entered in the records of all ambulatory healthcare. When so recorded, routinely and uniformly, the valueof health records to the patient care process, and the potentialof health records as an information source are greatly enhanced.

Patient data items

1. Personal identification2. Residence3, Date of birth4. Sex5. Race and ethnic background

Provider data items

6. Provider identification7, Location or address8, Type of practice9. Profession

Encounter data items

10. Date and place of encounter11. Patient’s reason for encounter12. Diagnostic services13. Problem, diagnosis, or assessment14. Therapeutic services15. Preventive services16, Disposition17. Expected principal source of payment18, Total charges

This content reflects several changes from the previousdata set that the consultants recommend be adopted by theNational Committee on Vital and Health Statistics. Thesemodifications are proposed in order to accommodate the chang-ing environment of ambulatory care. Specifically:

1, The consultants propose a data set that is patient orientedas opposed to provider oriented. The structure of thedata set is therefore modified to reflect all the servicesprovided for the patient during the course of an encounteras opposed to the services rendered by one provider.

2. In accordance with this concept of a patient-oriented dataset, the consultants recommend an expansion of the scopeof the data set. The definition of encounter is modified

to clarify the role of the provider and so capture a greaterproportion of those services provided by nonphysicians.

The following points concerning the Uniform AmbulatoryMedical Care Minimum Data Set (UAMCMDS) should beemphasized

1. The items constitute a minimum data set. As such, thedata set was designed to meet common data needs amongmultiple user groups, not to meet the total data needsof any one user group. Providers of ambulatory careand collectors of ambulato~ care data should feel freeto expand this minimum data set in accordance with theirparticular requirements.

2. The data set is recommended for inclusion in health rec-ords. Providers of care should record these data in atleast the detail specified using the recommended defini-tions. Although desirable, all items need not be recordedin the individual patient health record. Some items, forexample, may be included in registration or billing rec-ords. In such instances, however, the capability shouldexist to link data from the various record sources. Inaddhion, some data items need only be recorded onceand updated when necessary.

3. Neither a survey or other data collection system nor adata collection “form” is recommended. Data systemsand forms will generally vary according to the data needsof individual programs. It is recommended, however,that when UAMCMDS items are included in a data system,the recommended definitions and minimal classificationsbe followed.

The consultants recommend that the Uniform AmbulatoryMedical Care Minimum Data Set be accepted by the U.S.Department of Health and Human Services, and other Federalagencies that fund ambulatory health care programs. It isfurther recommended that the U.S. Department of Healthand Human Services designate an office to have responsibilityfor promulgation of the data set. Such promulgation wouldinclude the adoption of the Uniform Ambulatory MedicalCare Minimum Data Set and its definitions by those Federalprograms requiring reporting of ambulatory care events. Thiswould encourage uniformity in Federal reporting for the itemsdescribed in the minimum data set. The consultants also reiter-ate previous recommendations of the National Committee onVital and Health Statistics that private health care institutions,professional organizations, and insurance caniers endorse andagree to use the data set.

13

Appendix ILMinutes of the NCVHSSubcommittee onStatisticalPhysician

Aspects ofPayment Systems

June 19, 1985 meeting

The National Committee on Vital and Health Statistics(NCVHS) Subcommittee on Statistical Aspects of Physician,Payment Systems held its first meeting on June 19, 1985.The Subcommittee members had previously held three meet-ings as a work group between June and October 1984.

Participants were William R. Felts, M. D., Chairman,and Theodore Allison, NCVHS members; William F. Stewart,James Delozier, and Marjorie Greenberg, National Centerfor Health Statistics (NCHS); Alan Bradt and Jean Harris,Health Care Financing Administration (HCFA); David Zim-merman and David Rinaldo, Metropolitan Life InsuranceCompany; and Steven Culler, Blue Cross and Blue ShieldAssociation. In addition, approximately 8 persons attendedas observers, including a representative from the PennsylvaniaState Medicaid Program. The third NCVHS member of theSubcommittee, Carmault B. Jackson, Jr., M. D., was unableto attend due to illness.

Representatives of the Medicare program, MetropolitanLife Insurance Company, and the Blue Cross and Blue ShieldAssociation each provided an overview of ambulatory caredata flow for the patient-physician encounter in their respectiveorganizations. Current data procedures, requirements, andproblems, as well as anticipated future data needs, werediscussed.

Medicare claims processing

Jean Harris, Division of Carrier Procedures, providedextensive information on claims processing for MedicarePart B, which includes physician services, durable medicalequipment, and laboratory services. The program has approxi-mately 30 million beneficiaries. In FY 1985, HCFA anticipatesprocessing about 260 million claims, each of which will havetwo to three services per claim. Due to the size of the systemand the cost of operating it, the program tries to keep dataitems to a bare minimum.

Medicare Part B claims can be submitted by either thepatient or physician, which has enormous implications forthe quality of the data. In the past 5 years, the programhas made two major changes that have enhanced the uniformityof data collection. First, a common claims form, the HCFA1500, was instituted for all physician-submitted claims. Sec-ond, the HCFA Common Procedure Coding System (HCPCS)was adopted for coding of procedures.

The Medicare program also is actively encouraging sub-

mission of claims in electronic medi% currently between 10and 20 percent of claims are submitted in this form.

Carriers retain data for future use for four purposes:

. To maintain a history of the beneficiary (to check forduplicate claims and medical necessity).

. To update reasonable charge screens.● To perform post-payment utilization review.. To monitor fees of nonparticipating physicians.

Because Medicare is no longer always the primary payer,it has become necessary to collect more patient informationthan in the past for coordination of benefits.

Patient-submitted claims represent a significant problemfor reporting diagnosis or reason for encounter. On theseclaims only symptoms or complaints can be requested. Dueto the unreliability of much of the information, carriers donot routinely enter diagnosis into their data bases, unlessdata are submitted in electronic media. Currently HCFA isconsidering an instruction to carriers that will require themto have the capability to keep d]agnosis in their claims process-ing systems, However, carriers probably will only need toretain diagnosis for paper claims where it is necessary forspecific reimbursement decisions. Over half of the claimsare for office visits and the vast majority of claims are paidwithout reference to diagnosis. Given the volume of claims,it will probably be prohibitive to match diagnoses with proce-dures on a routine basis. Mr. Stewart suggested considerationshould be given to collecting the complaint or symptoms,as well as a diagnosis, on physician-submitted claims.

Availability of diagnostic information will improve withan increase in physician-submitted claims. About 29 percentof physicians and suppliers currently are classified as participat-ing physicians in the Medicare program, and in the past yearthere has been a dramatic rise in the assignment rate from53 percent to 68 percent of the claims. In some cases, physi-cians complete the claims form even when they do not acceptassignment. At the same time, valid questions still can beraised about the accuracy of some of the diagnostic and,to a lesser extent, procedure codes received, because codingusually is performed by office personnel who are not specifi-cally trained in medical coding. When codes are not recordedand must be entered by the carrier, addhional inaccuraciescan occur. Dr. Felts and others raised questions about theappropriateness of carriers changing codes for various reasons.

In the area of physician identifiers, most carriers are

14

able to distinguish the individual physician who performedthe service, even when the physician is in a group whichuses a single identifier. However, it is virtually impossibleto pull together all of the services rendered by a given physicianin a variety of practice settings. It would be very costlyand sensitive to give each provider a unique number.

The Uniform Claim Form Task Force met in early Juneand agreed to pursue greater standardization of the claimform (HCFA 1500) by eliminating local variants. The TaskForce is chaired jointly by HCFA and the American MedicalAssociation and includes representation from Blue Cross andBlue Shield Association and the Health Insurance Associationof America. The Task Force also reaffirmed the enormouscosts to third-party payers and providers alike of makingchanges in the claims form and concurred that we may beon the verge of significant reimbursement reform for physi-cians. The latter would provide the next opportunity to makemajor changes in the claims form. The Task Force identifiedapproximately 14 additional new items for consideration; manyrelate to determining eligibility and coordination of benefits.

Since the implementation of prospective payment for hos-pitals in October 1983, HCFA is requiring intermediaries tosend Part A data, which cover hospital services, to the carrierswho process the respective Part B data. In addition to prevent-ing double billing for nonphysician services during a hospitali-zation, there is also the concept that if a hospital stay ora portion thereof is denied for reasons of medical necessity,then perhaps some physician services should also be deniedor their intensity reduced. HCFA is also supporting a numberof research studies involving linked Part A and Part B records.

Medicare statistical system

Alan Bradt, Bureau of Data Management and Strategy,updated the Subcommittee on Medicare Part B data activities.There are basically four statistical systems:

. Bill summary record—5 percent sample of beneficiaries.

. Payment record—100 percent of payments.

. BMAD-Part B Medicare Data System, consisting offour files; it may make the previous two files archaic.

. Peer Review Organization (PRO) data--currently onlyPart A data, but the PRO’s may expand to reviewingPart B services as well.

Details of the different data sets are contained in the minutesto the work group meeting of June 14, 1984.

Standard data files and reports from the BMAD will bepublished and available. The 1984 data currently are beingprocessed; approximately 95 percent are reporting HCPCScodes, although some of the codes have been converted fromprevious coding systems. No information is available fromhealth maintenance organizations.

Uses of the statistical files include the following:

● Merging Part A and BMAD data for the report due toCongress July 1, 1985, on prospective payment for physi-cians. This report will also look at other possible physicianreimbursement reforms, such as fee schedules, relativevalue scales, and cavitation.

. HCPCS maintenance.

. Prevailing charge tables.

. Administrative cost containment analyses.

HCFA is developing an on-line access system which willgreatly facilitate special studies by the different components.This system will probably not be operational by October 1,as originally planned.

There was general agreement that the method for com-municating to physicians and carriers the rationale behinddeleting old procedure codes and adding new ones to HCPCSshould be improved.

Commercial insurance data requirements

David Zimmerman, Assistant Vice President, describedthe data requirements at Metropolitan Life Insurance Company,which he considered simiIar to other commercial carriers.

Claims are submitted to the insurer by individuals, provid-ers, employers, and electronically through a clearinghouse.Electronic submissions account for approximately 5 percentof the claims. The majority of claims come from the individualor the employer, who often must veritj eligibility. Claimsaverage three services and two providers per claim.

Basically, the HCFA 1500 satisfies most of the datarequirements. One new item which probably will be neededis the spouse’s month and day of birth to determine theprimary carrier for chiIdren of two insured parents. TheNational Association of Insurance Commissioners has recom-mended the use of whichever parent’s birthday comes earlierin the year as an unbiased decision rule. Greater detail onplace of service is also desired by many companies, whichare encouraging people to use outpatient care as much aspossible.

Employers ate becoming very aggressive with cost con-tainment measures and increasingly are making demands fordata, Two factors will force greater standardization of datacollection. First, employers are tending to switch insurancecompanies far more frequently than in the past. Second,business coalitions are interested in sharing data.

Following the lead of Medicare, the commercial insurersare moving to ICD-9-CM for diagnostic coding and CPT4for procedure coding. There are no immediate plans to useHCPCS, but the CPT modifiers eventually will beaccommodated.

David Rinaldo, Statistical Consultant at Metropolitan Life,reported on typical uses of the claims history data. Theseinclude:

. Reporting to policyholders on service utilization andexpenditures.

. Tracking of savings from cost containment programs.

. Identifying aberrant patterns of practice.

. Identifying efficient patterns of practice.

Mr. Rinaldo suggested that in the future, data needswill be greater than the items currently contained on healthinsurance forms. Specifically, there will be a need to linktreatment and charges data for a series of encounters toconstruct a defined episode of care. Questions on whether

15

the condition is chronic or acute, whether this is the firsttreatment for the illness, and on referrals would facilitatethis construction. The stimulus for these data will probablycome from employers. Although most of the focus has beenon inpatient costs, there is an increasing interest in the ambula-tory arena.

Ms. Harris noted that packaging services is attractivedue to the sheer volume of claims, but that to date effortsto bundle have had minimal success. It has been estimatedthat packaging could reduce the annual number of claimsfrom 260 million to 150 million. Dr. Felts pointed out thatpatient outcome must be considered when looking at episodesof care and expressed concern about the state-of-the-art forexamining and making judgments on outcomes from differenttreatment patterns.

Ambulatory visit groups

Jim Delozier, Ambulato~ Care Statistics Branch, NCHS,described the research currently being conducted at Yale withsupport from HCFA on Ambulatory Visit Groups (AVG’S).Data from the National Ambulatory Medical Care Survey(NAMCS) conducted by NCHS are being analyzed to developgroups with a common pattern of resource use. Patients areclassified primarily by diagnosis, but in some cases the pa-tient’s complaint is also a factor. The goal is to developan ambulatory patient classification capable of serving as thebasis for a per case prospective reimbursement system.

It was noted that, whereas certain reimbursement ap-proaches, such as AVG’S, might require data not currentlycollected, others such as cavitation would require practicallyno data for reimbursement. However, data for utilization re-view are still necessary in the HMO setting to identifyunderutilization.

Blue Cross and Blue Shield Association(BCBSA) data needs

Steven Culler, Manager of Professional and InstitutionalPayments, reported on BCBSA data requirements. TheBCBSA is basically a coordinating agent for 88 independentplans, providing research, actuarial marketing, and educationalservices. Claims processing is performed by the individualplans, and there is no national aggregation of claims data.The majority of claims are submitted by participating physi-cians, and eligibility is usually determined by the plan andnot the employer.

Blue Shield uses a claim form very similar to the HCFA1500. A collapsed CPT-4 coding system is used for reimburse-ment purposes. The extent to which the coding system iscollapsed varies among the plans.

Uses of the data include reimbursement, utilization review,auditing, and antifraud efforts. Hospital records have improvedconsiderably for all patients since the Prospective PaymentSystem was implemented for Medicare, and similar improve-ments are anticipated in ambulatory records as outpatient careis scrutinized. The emphasis is expected to be on improveddata quality and more analysis rather than major new datarequirements. Mr. Culler concurred with Mr. Rinaldo thatinterest in outpatient review is relatively recent but on therise. Employers are providing a major impetus for encouraginguse of outpatient settings.

For institutional review, the emphasis is on pre-admissionreview rather than post-admission denials. The plans are tryingto identify cost-effective physicians by making comparisonswith comparable peer groups. Mr. Culler acknowledged thatthere are problems defining the correct peer groupings anddetermining efficiency. Also more profiling of claims databy physicians will be needed in the outpatient area. In somecases, profiling of patients is also necessary to identify over-utilizers. BCBSA has made some preliminary efforts to developvisit categories, which would require better collection of thereason for encounter, but the individual plans have not beenenthusiastic about this approach. Fraud and abuse activitiesrequire evidence from outside the claims form, but there istremendous variation in the backup records available in am-bulatory settings.

The Subcommittee agreed that it would not be necessaryto meet again July 9 as tentatively scheduled. The next Sub-committee meeting will be September 10. Subsequently, ameeting was also scheduled for October 15.

Dr. Felts identified the following groups as possible pre-senters at fim.u-emeetings: other insurers, such as Medicaid,Kaiser, Worker’s Compensation, Civilian Health and MedicalProgram of the Uniformed Services (CHAMPUS), and theVeterans Administration; State personnel; chairmen of the rele-vant congressional committees or their staffs; Peer ReviewOrganizations; Institute of Medicine; employer groups; otherDHHS components; and the research-epidemiology community.

Prepared by: Marjorie S. Greenberg, OPPEC, NCHS.

September 10, 1985 meeting

The National Committee on Vital and Health Statistics

(NCVHS) Subcommittee on Statistical Aspects of PhysicianPayment Systems held its second meeting on September 10,1985.

Participants were William R. Felts, Jr,, M.D., Chairman,Theodore Allison and Carmault B. Jackson, Jr., M. D.,NCVHS members; James Delozier, Marjorie Greenberg, andWilliam F. Stewart, National Center for Health Statistics(NCHS); Alan Bradt and Bob Silva, Health Care FinancingAdministration (HCFA); David Plotnick, Group Health As-sociation of America (GHAA), and Gerald Radke, Pennsyl-vania Department of Public Welfare.

Dr. Felts opened the meeting by reviewing the ma. vconcerns of the Subcommittee. These concerns include iden-tifying data items that are not collected in a uniform mannerfrom the patient-physician encounter, anticipating changes inphysician payment systems which may require collection andanalysis of additional items of data, and examining the possibleimpact of these changes and additions on comparability ofdata and trend analysis.

The first Subcommittee meeting on June 19, 1985, focusedon Medicare and private insurer data requirements. This meet-ing emphasized data requirements of Health Maintenance Or-ganizations and a State Medicaid program.

Health Maintenance Organizations (HMO’s)

Mr. Plotnick, Director of Research and Development atGHAA, provided background on his organization and thetypes of HMO’s it represents. The GHAA was founded in1959 as a National Trade Association of prepaid group practiceand staff model HMO’s. The organization currently represents130 HMO’s, which account for about 12–13 million enrolleesor approximately 75 percent of the Nation’s enrollment inHMO’s, At present about one quarter of the HMO’s enrollpersons on Medicare but many more are contemplating itdue to the new Tax Equity Fiscal Responsibility Act of 1982(TEFRA) requirements.

In addition to group and staff model HMO’s, there areother models, including independent practice associations(IPA’s) which are represented by a separate organization.

Mr. Plotnick, who is responsible for developing a com-

parative data base on HMO’s, emphasized that the structureof an HMO affects the organization’s data collection andcontributes to the lack of uniformity among different typesof HMO’S. The GHAA recently examined the different HMOreporting requirements through a HCFA grant.

Basically, there are three different HMO models—staff,group, and IPA--(see figure I), plus mixed models. In eachcase the HMO is responsible for marketing the health plan,collecting premiums, and then contracting with other organiza-tions or hiring providers. The unique aspects of each modeland their impact on data requirements areas follows:

● Staff model-Group Health Association of Washington,D. C., is an example. Staff are salaried and represent

a fixed cost. No claim forms are generated, and thereis minimal examination of specific units of service orspecific encounters. The HMO contracts for inpatient carewith individual hospitals which prepare the HCFA uniform

bill. Specialty care is provided either on a fee-for-serviceor retainer basis.

● Group model—This model is exemplified by the KaiserMedical Care Program. The HMO either sets up its own

Staff model

I

HMO

I

Group model I HMOI

IPA model HMO

1

1

Hospital

Figure L HMO model types

17

multi-specialty medical group or contracts with an existingmedical group. In the latter case the HMO is dependentupon the data system used by the group. The HMO gener-ally pays by cavitation, putting the group at risk formedical services and, in some cases, ancillary and evenhospital services. Again it is not necessary to look atspecific units of service for reimbursement. Some largergroup models own their own hospitals, while others con-tract for hospital services as in the staff model. Somespecialty services may be purchased.

● IPA model—The IPA model is set up to contract witha number of providers covering a broad geographic area.This model may use cavitation or fee for service. Claimforms are used for each encounter, and thus generallythere is more information than in the other tyo models.Community hospitals are used, as in the other models,but there are usually very few referrals for specialty careoutside the IPA.

In recent years many HMO’s are mixing the model types,and the lines are becoming less clearcut.

The source documents used for encounter data lack uni-formity, even within the same HMO system, such as thedecentralized Kaiser Program. Some HMO’s use appointmentbooks rather than encounter forms, and even the definitionof an encounter may vary. The recommended definition ofan encounter is “a face-to-face contact between an HMOmember and a provider of health care services who exercisesindependent judgment in the care and provision of healthservice(s) to the member.” However, some HMO’s counta telephone contact as an encounter, and others may counta series of services ordered by a physician as separate encoun-ters. HMO’s also differ on whether they can document theactual number of encounters or types of services providedby specialty providers.

The reason for the encounter may appear in the medicalrecord but rarely is recorded on an encounter form. Veryfew HMO’s code diagnosis in the ambulatory setting, andeven on the hospital side many HMO’s currently do not analyzediagnosis trends.

HMO’s routinely track provider productivity, queuingtime and waiting time, Quality of care and underutilizationare monitored more frequently through chart review thanthrough the data systems. HMO’s will be subject to reviewby Peer Review Organizations under procedures currentlybeing negotiated with HCFA.

Although HMO’s, to the extent that they are communityrated, don’t need group specific experience data for administra-tive purposes, government and employers are demanding moreutilization information about their enrollees. As these demandsincrease and as the HMO indust~ becomes more competitivein negotiating with providers, HMO’s are under pressure todevelop more sophisticated data systems, to improve codingof procedures and diagnoses, and to capture more information.

However, it will probably be 5 to 10 years before theHMO data system capabilities are significantly improved. Eventhen, the type of data system used in the fee-for-service envi-ronment would be onerous and drive up the costs of an HMO,reducing its competitive advantage. A system specific to

HMO’s would be preferable. Only one HMO, the HarvardCommunity Health Plan, uses an automated medical record,and this is said to be generally disliked by the physicians.

Currently the IPA’s, which are the fastest growing modeltype, have much better utilization data than staff or groupmodels because payment is usually tied to a claim form.Since IPA’s have to manage utilization over a whole geo-graphic area to providers who may not be seeing many HMOpatients, the claims are examined fairly carefully. Larger cen-tralized systems established by chains and management com-panies will naturally have better and more uniform data collec-tion systems.

Although there are data supporting both adverse and favor-able selection by HMO enrollees, there are very inadequatedata, in general, on why people choose certain health plansover others. Another area where more data are needed istechnology utilization. The GHAA recently conducted twosurveys on organ transplants. Perhaps 15 to 20 HMO’s havetheir own research departments, and Kaiser Health Plan inOregon has its own health services research center. For re-search projects, these plans typically generate their own dataset, drawing directly from the medical records. The NationalAmbulatory Medical Care Survey (NAMCS) conducted byNCHS, which uses its own data collection form, includesHMO physicians in its survey of office-based physicians andis trying to tighten its definitions in this area in the 1985survey.

In partial response to the overall lack of uniformity indata collection, GHAA is developing a summary level database, using standardized definitions. The data base shouldbe able to provide numbers of encounters by different typesof providers for the whole industry. Funded by the HenryJ. Kaiser Foundation, the project currently is collecting datafrom 14 HMO’s across the country. Eventually HMO’s willparticipate on a subscription basis and receive comparativedata by which they can measure their own performance.

Medicare update

Alan Bradt, Bureau of Data Management and Strategy,reported on Medicare data activities. The issue of a uniquephysician supplier identifier is being addressed at an internalHCFA meeting on September 11. Currently a single physiciancan have multiple I.D. numbers within the same carrier servicearea, as well as differing numbers in different carrier areas.The probable approach will be a HCFA-assigned number(HAN), which would be unique to HCFA and not tie intoany other existing systems. The carrier would use the HANto replace any identifiers currently in use.

Coding maintenance is continuing on the HCFA CommonProcedure Coding System (HCPCS). The carriers basicallyhave converted to HCPCS, but there are communication prob-lems. Not all carriers are using all the codes and not allcodes are being paid properly. Staff will be working on achiev-ing more uniformity in coding applications across the country.

Coding standardization activities underway include theMandex contract to develop an index for HCPCS and CPT,BMAD data analysis to identify trends or patterns of services

which will help standardize coding applications, and validatingand reviewing the two adapters that have been developedfor placing Volume 3 ICD-9-CM and CPT codes into commonDRG’s. Mandex is also conducting the latter effort and isencouraged by preliminary results.

The 1984 BMAD data are in house, with approximately50 percent edited and accepted. The data base should bereudy for HCFA use and analysis by October 1, 1985, andthe on-line system should become available by February orMarch 1986. Access to the latter will be menu driven, permit-ting users to work with manageable data files that meet theirspecific needs. All carriers will have converted to HCPCSby the end of September and corrections of erroneous conver-sions are moving along slowly.

One BMAD application currently under study by HCFAis development of national fee schedules. Other groups withinHCFA are examining relative value scales and extension ofDRG’s to cover physician payment. Each of these researchefforts relates to the report to Congress on Physician DRG’s,which was due July 1, 1985, and is currently undergoingDepartmental clearance.

Finally, HCFA is looking at ways to streamline thePart B data system to reduce carrier burden and expense.One approach would be to increase the BMAD submittaland eliminate other components, such as the payment record.

Medicaid data requirements

Gerald Radke, Deputy Secretary for Medical Assistanceimd Medicaid Director in the Pennsylvania Department ofPublic Welfare, reported on Medicaid data requirements fromthe perspective of his State. Mr. Radke emphasized that thereis little central direction to the 54 separate Medicaid programsat the operational level and that, to some extent, each Statemakes its own decisions on coverage and reimbursement.

Data requirements are driven by both political consider-ations, such as sensitivity to public perceptions of fraud andabuse, and the following administrative-structural problems:

. Since the legal contract is with the provider, not’ thebeneficiary, all the controls are on the provider side.

. Further, because it is an open-ended contract, the controlsare after-the-fact controls which seek to examine medicalnecessity, This is the focus of most of the informationcollected, although the diagnostic information is not verygood, The most frequent procedure is an office visit.

As the payor of last resort, the Medicaid programmust spend a lot of time making sure it doesn’t pay

bills covered by another payment source. Trauma codes,in particular, are closely monitored for alternativecompensation.

In addition, there are Federal requirements related to time-liness and accuracy. At the same time, there is pressure atthe State level to reduce administrative costs. Pennsylvaniahas reacted to these cross pressures by putting the burdenon the provider to fill out the invoice correctly or have itreturned.

Like most large States, Pennsylvania uses a fee schedule.Since 1979, the State has been making a concerted effortto upgrade the schedule, beginning with the primary carearea. As a result, statewide payments to the physician commu-nity have doubled over a two-year period.

The Medicaid program currently uses ICD-9 for codingdiagnoses and a variation of CPT-3 for coding procedures,but will probably convert to HCPCS by July 1986. Mr. Radkenoted that the Medicaid data bases have tremendous potentialfor research, but the primary focus must be on collectinginformation to administer the respective State Medicaidprogram.

Although supporting uniform minimum data sets, Mr.Radke stated he was strongly against common claim forms.The latter are very difficult to use, because they contain un-necessary items and the same item can have multiple definitionsdepending on the insurer. As a result, it is diff]cult to trainpeople to fill them out and there are potential problems withdata accuracy. The physicians in Pennsylvania have refusedto use the HCFA 1500 for Medicaid reimbursement, preferringa simplified State invoice. The 1500 is totally inadequatefor determining coordination of benefits, an important issuefor Medicaid programs. Hospitals may start using the UB82,even though it is considered inappropriate for DRG’s, butthe UB82 will never be used for outpatient services in Pennsyl-vania. Every time the State changes invoices it costs theMedicaid program about $1 million.

Pennsylvania Medicaid is in the process of implementinga prepaid cavitation project for 100,000 Philadelphians. Theproviders will submit encounter forms to a contractor, whowill feed information back to the physicians and the Medicaidprogram.

Following the presentations, the Subcommittee reviewedthe agenda for the October 15 meeting, which will beginwith Subcommittee business at 9:00 a.m. It was tentativelyagreed that an additional meeting would be held in January.

Prepared by: Mmjorie S. Greenberg, OPPEC, NCHS.

19

October 15, 1985 meeting

The National Committee on Vital and Health Statistics(NCVHS) Subcommittee on Statistical Aspects of PhysicianPayment Systems held its third meeting on October 15, 1985.

Participants were William R. Felts, Jr., M. D., Chairman,Theodore Allison and Carmault B. Jackson, Jr., M. D.,NCVHS members; James Delozier and Marjorie Greenberg,National Center for Health Statistics (NCHS); Alan Bradt,Richard Bale, Ph. D., Richard Beisel, Donald Muse, Ph. D.,and Donald Sikora, Health Care Financing Administration(HCFA); Thomas J. Beatty and Brenda Freitag, PrudentialInsurance Company of America; Susan McAllister, Hon-eywell, Inc.; and Barry Wilson, Group Hospitalization andMedical Services, Inc.

The meeting began with discussion of Subcommittee busi-ness, including preparation of the Subcommittee’s report tothe full Committee. Mrs. Greenberg agreed to draft an interimreport for review by Subcommittee members and distributionto all NCVHS members prior to the November 7–8 NCVHSmeeting. The Subcommittee decided to hold an additionalmeeting on January 14, 1986, to hear presentations fromother Federal agencies, such as the Centers for Disease Control,the Veterans Administration, and the Department of Defense,and to obtain more information on hospital outpatient dataneeds. Subcommittee members and staff will also meet thefollowing day, January 15, to prepare the Subcommittee’sfinal report.

Following Subcommittee business, presentations wereheard on the following topics related to data needs and currentdata practices in the ambulatory care arena:

. The new person-based national Medicaid reporting sys-tem.

. Data needs of a Medicare carrier.

. Data needs of employers and business coalitions.

. Data needs of Peer Review Organizations.

. Data needs of a Blue Cross Blue Shield plan.

National Medicaid reporting system

Dr. Donald Muse, Director, Division of Medicaid Esti-mates, Office of Actuary, described the evolution of the na-tional Medicaid reporting system. Between 1965 and 1981,Medicaid had a paper system based on a 12-page requiredform. Beginning in 1981, the form was expanded to 47 pagesand a quality review system was instituted with financialpenalties for noncompliance. More importantly, a tape-to-tapeproject was undertaken with the goal of standardizing unitrecord data in a common format across States.

The tape project determined that certain variables couldbe collected uniformly. As a result, States were given theoption either to continue submitting the aggregate paper reportor to begin submitting unit record data on tape. Thirty Stateshave chosen the tape option, with eight selected initially toprovide data on FY 1985 payments. Reports will be available3 months after the close of the fiscal year. Currently theMedicaid fiscal year reports are based on the date-of-payment rather than the date-of-service. Within a year,HCFA may require all States to submit Medicaid data ontape in response to pending Congressional action regardingspecial adjustments in Medicare payments to hospitals thatserve large numbers of poor patients. In order to implementthis requirement, it would be necessary to link Medicaid andMedicare data on hospital services.

The tape system consists of the following three files:

● Paid claims.. Eligibility.. Provider.

The paid claims file contains three separate subfiles forhospital inpatient claims, long-term care institution claims,and other provider or physician claims. Each subfile has re-quired data elements, desirable data elements, and State-specific codes. Since the paid claims file and the eligibilityfile both are used extensively by the States for payment andutilization control, the data on these files are considered morereliable than the data on the provider file. The latter willrequire a major clean-up overtime as use of the file increases.

Although the long-term goal is to have the Medicaiddata comparable to Medicare data, the initial objective wasto make the system operational. Thus it was not possibleto achieve standardization for certain elements, such as diag-nosis and procedure codes, where widespread variation existsat the State level. Diagnoses, coded in ICD–9–CM, are “desir-able” rather than required data items on all three claim subfiles.Procedure codes are “desirable” on the inpatient claim subfileand required on the “other” or physician claim subfile. Inboth cases a number of different coding systems are acceptedfor procedures, with each State designating which systemit is using. No procedure codes are requested on the long-termcare subfile. In addition, the inpatient subfile includes as“desirable” a principal procedure category, which is to beselected from a standard list of very broad surgical categories,such as “Surgery Cardiovascular System.” If these tiles becomeinvolved with payment mechanisms for Medicaid, as manyexpect, significant efforts will be made to increase standardiza-

20

ticm. However, the Medicaid program has a history of flexibil-ity at the individual State level, which is often at odds withuniformity.

Other items of interest to the Subcommittee included placeof service and personal identifiers. Dr. Muse noted that theMedicaid program has always collected place of service ofphysician visits but has encountered a great deal of difficultywith standard definitions across States and has essentiallydropped that requirement. Concerning personal identifiers,every State reporting on tape must provide a unique recipientID number, which will probably be the person’s Social Securitynumber by the end of the decade. Un;que provider ID numbersw-econsiderably more problematic, particularly in the ambula-tory care setting.

The first priority for data production will be the basicreporting by service types and types of eligibles. Initial analysiswill probably focus on long-term care data. Eventually, publicuse data files should be available, but procedures must bedeveloped to draw useful samples from the approximately600 million records submitted per year by the thirty States.

Data needs of a Medicare carrier

Mr. Thomas Beatty, Vice President for Federal Programs,reviewed Prudential Insurance Company’s experience convert-ing three States and four programs from California RelativeValue (CRV) to the HCFA Common Procedure Coding System(HCPCS), Prudential is the Medicare (Part B) carrier for NewJersey, North Carolina, and Georgia as well as the Medicaidcontractor in New Jersey, processing nearly 25 million claimsper year. Prudential is also the Medicare (Part A) fiscal inter-mediary for approximately 500 providers .’The conversion diffi-culties Prudential encountered were not with the actual cross-walk between the two coding systems but with the inconsis-tency in definitions for related data elements across the threeStates. These elements included followup days, surgical proce-dures which required an assistant, method for measuring oxy-gen, etc. Since the payment safeguards are related to theseelements, Prudential felt it was crucial to standardize themacross the States. After considerable effort, all three Statesnow are recording the HCPCS elements uniformly, with theexception of risk on anesthesia. In addition, the dentists par-ticipating in the New Jersey Medicaid program did not agreeto the conversion to HCPCS and were permitted to retaintheir current coding system. Although the conversion fromCRV to HCPCS was fairly direct, Mr. Beatty expressed limitedconfidence in conducting trend analyses using data from thetwo coding systems.

Mr. Beatty and Ms. Freitag also described the work Pru-dential currently is conducting under a 2-year HCFA grantto examine patient-induced overutilization. Prudential becameinterested in this area during the 1970’s after analyzing datafor visits by one patient to numerous physicians or excessivevisits by one patient to one or two physicians. The lattertype of overutilization was labeled “patient-induced and pro-vider-condoned.” The HCFA grant will permit interviews withover 1200 patients who had 47 or more office visits peryear, and with their physicians. In addition, clinical informa-

tion on patients with 2546 visits per year will be requestedfrom their physicians by mail. After all the cases are reviewed,some patients may be placed on a form of pre-careauthorization.

The review of Medicaid utilization patterns by Prudentialis facilitated by the availability of diagnosis on the automatedrecord. Mr. Beatty noted that if physicians could access thisinformation base to review a patient’s previous treatment andprescriptions, it might be possible to reduce expenses andincrease the quality of patient care.

Prudential discontinued inputting diagnosis on the Medi-care Part B records when the Department of Health, Education,and Welfare decided not to pay for the item and to chargethe carrier with an error if the item were erroneous. Mr.Beatty stated that it would be helpful to have a diagnosison the Part B claim for hospital visits but was less sureof its value for ambulatory care visits. In particular, he ex-pressed concern about the quality of the information sincethe majority of claims are submitted by the patients. MatchingPart A and Part B records is considered a valuable tool.

Prudential estimates that in the State of New Jersey, forevery dollar spent on monitoring, $25 are saved. The ratioin the South is approximately 1:15.

Data needs of employers and businesscoalitions

Susan McAllister, Corporate Manager for Health Dataand Administration, presented an overview of the Honeywellhealth benefits program, which encompasses 68,000 employ-ees in 40 different business units and 200,000 dependents.Honeywell has been self-insured for 13 years and offers 400different medical plans. It is administered through two in-demnity carriers and four Blue Cross Blue Shield plans. Onebusiness unit, located in Phoenix, Arizona, is self-adminis-tered. The company is committed to self-insurance but hasmore questions about self administration. Although the lattermay reduce the costs of claims processing and produce moretimely data, it requires considerable expertise of staff.

The divergence of plans and a corporate culture whichencourages flexibility and decentralization make it extremelydifficult to develop a standard, uniform data base. Many ofthe plans in the field operations with very few employeesare HMO’s. In addition, 75 percent of the employees in thelargest unit, located in Minneapolis, are enrolled in HMO’S.

Honeywell has worked very strongly with the MinnesotaCoalition on Health Care Costs to develop a uniform datareporting request from the employers to the HMO’S. Therequest represents a compromise but should help alleviatesome of the current inadequacies in the HMO data. Ms. McAl-lister stated that HMO’s are beginning to respond to the pres-sure from employers to produce utilization data. Inpatientdata will become available first, followed by outpatient data.The IPA’s currently are in the best position to provide thedata requested.

The indemnity carriers also differ in their ability to producedata. The private carriers which deal nationally across manydifferent plans have been much more willing and able to

21

provide adequate data than the local Blue Cross Blue Shieldplans which have a monopoly in their area.

Currently, Honeywell is developing a competitive medicalarrangement request for proposal (RFP), which includes amanagement information system requirement. The RFP willbe made available to the local divisions to put out to PreferredProvider Organizations (PPO’S) forbids or to evaluate offeringscoming to them.

Honeywell also has an arrangement with McGraw HiIl-Systemetrics, which is developing a standard data base fromthe six different carriers. The data are on a mainframe fileto which Honeywell has online access. One analysis has at-tempted to attach DRG’s to the inpatient claims data. Interestin analyzing ambulatory care data is also growing since ex-penditures for ambulatory care have shifted from 30% oftotal dollars spent to 5570 in the past two years.

In addition to using data for purchasing decisions, Hon-eywell is also interested in educating its employees abouttheir health care. At the same time, the Company must besensitive to contldentiality and privacy issues.

Ms. McAllister noted that the imposition of a uniformbill by the Medicare program has helped foster standardization,resulting in improved data. State-wide data bases also offersome promise since many Honeywell divisions outside ofMinneapolis are quite small and cannot, themselves, produceadequate data on providers. The ability to request profilesfrom the Peer Review Organizations at the institutional levelwill also be pursued. Ms. McAllister supports strong researchdata bases for developing standards and benchmarks of care.

The Minnesota Coalition is interested in acquiring dataon diagnoses and procedures but has not finalized thetaxonomies. Currently Honeywell is dealing with ICD–9 codesfor diagnoses and ICD-9 or CPT4 codes for procedures.

Data needs of Peer ReviewOrganizations (PRO’s)

Donald Sikora, Chief of the Systems Management Branch,HCFA, presented a paper on the data needed by the PROprogram to carry out review functions. Mr. Sikora emphasizedthat many of the policies concerning physician review andrelated data requirements are evolving and subject to change.

Planned uses of the PRO data areas follows:

. Track patient health care histories and build a scenarioof the patient’s multiple encounters with the health caresystem.

. Generate statistics on patient care by DRG, geographicallocation of service, age, type of care, charges, etc.

. Examine changes in practice patterns and in the healthcare delivery system.

Currently, the PRO’s are reviewing hospital care, butthe need to examine pre-care and after-care in the ambulatorysetting is strongly recognized. Plans are underway for eightm nine PRO’s to review ambulatory care on an experimentalbasis. Mr. Sikora presented two record formats for datacollection to facilitate hospital and ambulatory review. Theinpatient PRO record format (FI Unibill), which was promul-

gated on January 1, 1985, is an outgrowth of the PSROHospital Discharge Data Set (PHDDS). The proposed recordfor PRO review of ambulatory services is in the preliminarydiscussion phase, and no decisions have been made on codingsystems and how the data will be collected. Mr. Sikoraemphasized that a good ambulatory record would make acontribution to the health care delivery system.

Review of Health Maintenance Organizations by thePRO’s is still under consideration. They will probably reviewthe hospital stays of HMO patients, but review of the totalambulatory care is questionable. Although the proposed recordformats should allow the PRO to establish whether a patientwas an HMO enrollee, identifying enrollees in PPO’S orself-insured groups would be more difficult. This limitationwould need to be addressed if a PRO moved into privatereview, which is encouraged by the legislation.

Standardizing the definitions for sites of care will alsobe necessary for the PRO’s to carry out their responsibilityto deny services or the cost of services, where indicated.Alan Bradt confirmed that the information Medicare collectson place of service is adequate for current processing butprobably will not meet future needs or the needs of otherprograms.

Data needs of a Blue Cross Blue Shieldplan

Barry Wilson, Vice President for Public Affairs, describedthe data flow at the Bhte Cross Blue Shield Plan of theNational Capitol Area. As is the case with all health insurers,the major focus of the Plan’s data systems is to process andpay claims rather than perform utilization analyses. The Planserves approximately 1.1 million participants, contracting with35 hospitals, 6000 physicians, 1500 dentists, over 30 homehealth care agencies, and a wide range of other types ofinstitutional and professional providers.

Currently 20 to 25 percent of the physician claims aresubmitted electronically, and a number of hospitals submittapes for use in processing and payment. The Plan is imple-menting a new claims processing system which will graduallyreplace the multiple electronic and manual systems and willbring together the Blue Cross and Blue Shield data. Thenew system employs ICD-9-CM coding for diagnoses andHCPCS coding for procedures, but analysis of this data willstill be limited. Because Blue Shield pays on the basis ofprocedures and Blue Cross pays for patient days of care,only three digits of the diagnosis are collected on professionalclaims and procedural data are collected only incidentallyon hospital claims. Thus date of service must be used tomatch hospital and professional services rather than diagnosisor procedure codes. Up to 60 percent of the data can belost in this kind of matching process. The importance ofcollecting more specific diagnostic information for professionalservices is recognized in house, but it has had a low priorityin the claims payment process. Recording of place of serviceis also limited to three categories needed for payment: inpa-tient, outpatient, and physician’s office.

The situation is compounded by the UB82 claim form,

22

which all hospitals in the Plan are adopting, and the InterplanData Record required by the National Blue Cross and BlueShield Association. Neither of these formats breaks out ancil-lary services in sufficient detail to be useful for studyingutilization, particularly of specific laboratory and radiologicalexams and other related services. Finally, patient-physicianencounters outside of the hospital and many other ambulatorycare services are reimbursed only under the major medicalcoverage for which claims are submitted by the subscriberrather than the provider. Early next year the Plan intendsto accept physician-submitted claims for all outpatient services,including office visits, which should improve the qualityof data received and the ability to construct patient carehistories. Within the Plan there is no difficulty identifyingindividual participants because the patient name and birthdate are available, upon enrollment and on each claim.

Mr. Wilson noted that even within the current constraints,the Plan is able to conduct a number of analyses with thedata base derived from the paid claims file. These analysesare used to redesign benefit structures and to consider im-plementation of cost containment features, such as pre-

admission authorization and second surgical opinions. Thedata are also used in the aggregate to identify potentiallyaberrant patterns of practice which then must be investigatedon a manual basis. At the same time the Plan has conducteddemonstration or pilot programs to acquire the data neededto evaluate health care services, such as home care andhospices, in terms of their appropriateness for coverage. ThePlan’s data support a 5-percent savings from pre-admissionauthorization, but there is less evidence of cost savings fromsecond surgical opinion programs.

Analysis of data on trends and practice patterns is com-plemented by a strong educational approach. The Plan believesthat the key to quality, cost effective care is obtaining profes-sional consensus for given protocols and then developingthe benefits structure to support these protocols. This approachis in the embryonic stage, but is viewed as promising bythe medical community as well as the Plan. Good qualitydata are useful both in developing the protocols and in trackingcompliance with them.

Prepared by: Mmjorie S. Greenberg, OPPEC, NCHS.

23

January 14, 1986 meeting

The National Committee on Vital and Health Statistics(NCVHS) Subcommittee on Statistical Aspects of PhysicianPayment Systems held its fourth meeting on January 14, 1986.

Participants were William R. Felts, Jr., M. D., Chairman,Theodore Allison and Carmault B. Jackson, Jr., M. D.,NCVHS members; James Delozier, Gail F. Fisher, Ph. D.,and Mmjorie Greenberg, National Center for Health Statistics(NCHS); Alan Bradt, Health Care Financing Administration(HCFA); Major Stuart Baker and Colonel Donald Rosenberg,M.D., Department of the Army; Kenneth Zimmerman, Civil-ian Health and Medical Program of the Uniformed Services(CHAMPUS); Emily Goodwin, United Hospital Fund of NewYork; Michael M. Lawson, Veterans Administration (VA),and John Llvengood, M. D., Centers for Disease Control(CDC).

The meeting began with an update on HCFA and NCHSdata activities relevant to the Subcommittee’s work. AlanBradt reported on the following HCFA issues:

●

●

●

●

Staff are researching current proposals for physician pay-ment reform but are awaiting further congressional guid-ance. The Harvard AMA study on relative value scaleswas funded, and projects on vouchers and carrier cavita-tion are among those under consideration.The Medicare data bases are being strengthened, with1984 calendar year data in house and instructions forsubmittal of 1985 data under development. The contractorcurrently is testing the on-line system and training forthe system will begin in February.Regulations have been published requiring Medicaid Stateagencies to use the HCFA 1500 and the HCFA CommonProcedure Coding System (HCPCS) by the end of 1986.Work continues on the HCFA assigned number (HAN)which will provide a unique identifier for each physicianrendering services. The associated data base will indicatethe physician’s specialty and related credentials and de-mographic variables. A staff paper on the HAN will beavailable for administrative consideration by the end ofFebruary, and a decision is expected by the end of theyear. Implementation would proceed state by state andtake approximately a year to a year and a half to complete.

Dr. Felts asked about the impact on data if the cavitationapproach predominates in the Medicare program. It was notedthat incentives inherent in cavitation may lead to underutiliza-tion, and to assure quality of care may require different ap-proaches to data collection and analysis. Mr. Bradt stated

that data similar to what is currently being collected willstill be needed, but that it will be more difficult to obtainoutside of the billing mechanism. Dr. Jackson suggested thatmore data will reside at the carrier and local level but agreedthat data at the national level for policy decisions may suffer.Mr. Bradt also observed that reliability of data is more ques-tionable when the data are not tied to a billing mechanism.

Jim Delozier reported that data collection for the 1985National Ambulatory Medical Care Survey (NAMCS) con-ducted by NCHS is being completed and that data fromthe survey should be available by the end of 1986. The1985 NAMCS collected information on approximately 75,000to 80,000 encounters from a sample of 5,000 physicians.Funding in FY 1987 for the next cycle of NAMCS, scheduledto be conducted in 1988, currently is in doubt. AlthoughNCHS has done research on expanding the survey beyondoffice-based physicians to include hospital-based physicians,funding for this component also has not been obtained. Dr.Felts noted that trend data from the 1981 and 1985 NAMCSshould be of interest, because the Medicare prospective pay-ment system was instituted between the two time periods.Mr. Delozier agreed but pointed out that expected sourceof payment, indicating Medicare or Medicaid status, willbe available only for the 1985 survey.

Following the NCHS and HCFA updates, presentationswere received on the following topics:

●

●

●

●

●

Ambulatory care data needs of Veterans Administration(VA) programs.Ambulato~ care data needs of Civilian Health and Medi-cal Program of the Uniformed Services (CHAMPUS)program.Ambulatory care data base project of the Departmentof the Army.Collection and use of hospital outpatient data.Ambulatory care data needs of Centers for Disease Control(CDC) programs.

Ambulatory care data needs of VeteransAdministration (VA) programs

Mr. Michael Lawson, Director of the Medical Administra-tion Service, and members of his staff described ambulatorycare data requirements of the VA programs. The VA deliversambulato~ care through the following two distinct approaches:

24

Two hundred and twenty-six outpatient VA clinics as-sociated with 172 hospitals and accounting for 17 millionoutpatient visits per year.Direct contract with Private Providers, resulting in approxi-mately $250 million in expenditures per yea-. Eligibilityfor contract care is much more restricted than for cliniccare and usually must be authorized in advance by theVA. There is a $75/month limitation, although this canbe exceeded after approval of a treatment plan.

The data systems supporting the two approaches are totallyindependent and currently are not compatible. The outpatientclinics use a manual system and collect only two items on100 percent of the visits in order to perform work load analy-sis—the patient’s eligibility category and whether or not thevisit is scheduled, is for a compensation claim, or is to applyfor medical care. More detailed information on demographicvariables and services rendered is collected on a 20 percentsample and on special categories of patients. The VA is inthe process of automating the clinic data system and movingtowards collection of all items on 100percent of the patients.

There is no mechanism at this time within the clinicsystem for identifying and tracking any services by an indi-vidual provider. Mr. Lawson has considerable concern aboutthe accuracy of diagnostic information on outpatient VArecords, and this information also is not tracked. Concernstibout resource allocation are creating pressures to code am-bulatory surgeries, but questions also exist about the accuracyof this data. General categories of surgery may be usedas Qcompromise,

The data system for contract care is a fee system whichbasically enrolls vendors, receives and keypunches bills, andmakes payments. Providers include on their usual bill a briefdescription of the reason for encounter and the servicerendered, but no medical coding of diagnoses or proceduresis performed or retained by the data system. Clerks manuallycheck the condition treated on the bill against the veteran’seligibility for treatment. Although the VA fully recognizesthe limitation of this data system, there is reluctance to burdenthe fee-basis physicians with additional data requirements.

Ambulatory care data needs of CWilianHealth and Medical Program of the UniformedServices (CHAMPUS) program

Mr. Kenneth Zimmerman, Health Care Data Analyst inthe CHAMPUS Statistics Branch, provided an overview ofthe CHAMPUS program and ambulatory care data require-ments, The mission of CHAMPUS is to augment the militarydirect health care system for the approximately 6.5 milliondependents of active duty members, retirees, and their de-pendents, whenever care cannot be provided by a militaryhospital. The actual users of CHAMPUS in a given yearcomprise fewer than 20 percent of the eligible population,or 1,1 million people worldwide.

Approximately 875,000 of the CHAMPUS users receivesome reimbursed care in the ambulatory setting, accountingfor 4.5 million visits in Fiscal Year 1984, and 85 percent

of the claims. Outpatient expenditures are approximately16 percent of the total CHAMPUS budget of $1.4 billionor about $225 million. Although the majority of resourcesand cost controls are on the inpatient side, utilization andexpenditures are increasing in the ambulatory sector.

The program is set up very similarly to a Blue CrossBlue Shield high option plan. Five fiscal intermediaries areunder contract with CHAMPUS to process claims for thebeneficiaries, which are submitted on one of several forms.About 65 percent of the billing comes in on the CHAMPUS500 form, with another 15 percent submitted on the HCFA1500 (CHAMPUS 501) form. Outpatient claims are also sub-mitted on the UB 82. The fiscal intermedhries are responsiblefor coordinating benefits, since CHAMPUS is always thesecondary payer, except to Medicaid. Charges are profiledand paid on a “usual, customary, and reasonable” basis atthe 80th percentile for outpatient services. Payment tapes,containing some of the data items abstracted from the claimsform, are submitted to CHAMPUS. Mr. Zimmerman reviewedthe specific items of interest to the Subcommittee:

●

●

●

●

●

Patient identification---considerable information is re-quired to determine eligibility, including the sponsor’ssocial security number. The patient’s name is collectedbut no unique number is assigned.Provider identi&ation-The IRS tax number is used foridentifying the provider. Since there are some inconsisten-cies and duplications across fiscal intermediaries, the pro-gram is trying to move towards a central provider database to be updated and maintained by the intermediaries.Problems will still exist with multiple providers withina group or clinic sharing the same ID number. The programwould like to be able to identify in the data system specificindividual providers by specialty, particularly in the psy-chiatric area.Diagnoses-Only the “primary diagnosis,” defined as thereason for outpatient care, is captured by the data systemand coded in ICD-9-CM. The forms include room foradditional diagnoses but these are not transmitted on thetape. Thus the coded diagnosis does not necessarily corre-spond to the most expensive treatment or procedure, andthere is not a diagnosis associated with each procedure.This seriously limits information for utilization review.Procedures-Coding of procedures is by CPT4 with sev-eral hundred additional CHAMPUS-assigned codes fornonphysician services and mental health services. Thelatter codes have not been coordinated with the HCFACommon Procedure Coding System (HCPCS). If othercoding systems are submitted by providers, the fiscalintermediaries must convert to CPT-4 and CHAMPUScodes.Encounter data—The data system captures date of en-counter, but when there are multiple encounters for thesame procedure, only the number of times the procedureis billed on the claim is recorded, not the specific datesof care. Place of encounter is not collected, but programofficials feel it would be beneficial, particularly in thearea of ambulatory surgery.

25

A limited data audit function currently exists atCHAMPUS, but the program is hoping to implement a dataintegrity study, which will be the first in-depth examinationof data quality and accuracy.

Utilization review is confined at this point primarily tomental health care. A small Office of Program Integrity doesselected reviews of suspected provider fraud and abuse.

Because the CHAMPUS budget has risen from under$1 billion in Fiscal Year 1981 to over $1.4 billion at thepresent time, there is a great deal of pressure to implementcost-efficient payment mechanisms which take advantage ofcompetitive forces in the marketplace. A number of HealthMaintenance Organization (HMO) and Preferred Provider Or-ganization (PPO) demonstrations currently are underway. Mr.Zimmerman noted that it will be important in these arrange-ments to define clearly the essential set of data items required.

A major data problem for CHAMPUS is the incompatibil-ity between the CHAMPUS data system and the militaryhospital data system. Definitions are not standardized, forexample, on patient units of care. Whereas the CHAMPUSprogram defines a visit as a face-to-face encounter with ahealth care professional, the military system includes telephonecontacts as encounters. Also, although a patient may receiveonly a portion of his care under CHAMPUS, the programdoes not have ready access to the data for the care providedin the military health system.

The CHAMPUS program developed a strategic informa-tion plan almost a year ago to revise totally the way dataare summarized from a claims-based approach to an encounter-based system. This plan is on hold pending other decisionson fi.uure CHAMPUS directions for reimbursing inpatientand outpatient care.

Ambulatory care data base project of theDepartment of the Army

Major Stuart Baker and Colonel Donald Rosenberg pre-sented a briefing on the Army Medical Department (AMEDD)Performance Measurement Study, which is now a Departmentof Defense, Tri-Service initiative. The AMEDD PerformanceStudy was initiated in 1983 in response to a need for betterepidemiologic and performance information (particularly inthe outpatient environment) in the Army Medical Department(AMEDD) Health Care Delivery system. The Army alreadyhad a specific inpatient reporting system which collected dis-charge data largely consistent with the data elements rep-resented in the Uniform Hospital Discharge Data Set.

The contribution of the AMEDD Performance Measure-ment Study to the inpatient system has been to examine theapplication of the Diagnosis Related Group (DRG) classifica-tion to the military beneficiary population in an attempt toexplain variations in practice patterns and hospital perform-ance. On the ambulatory side, the purpose is to establisha major new data collection and reporting system to replacethe current Medical Summary Reporting System which, likethe VA system, is based on the number and type of visits(e.g., cardiology, OB, etc. ) and cannot describe the natureor complexity of the ambulatory practice and services from

a diagnostic, procedural or demographic perspective. Boththe inpatient and outpatient components of the study are beingcoordinated with the current work on DRG’s and AmbulatoryVisit Group (AVG) research, which is ongoing in the HealthSystem Management Group (HSMG) at Yale University’sSchool of Organization and Management. From an ambulatoryperspective, one of the main purposes in maintaining a col-laborative relationship with HSMG was to insure compatibilitywith the newly proposed AVG “GROUPER.”

The Ambulatory Reporting System was designed to ad-dress three responsibilities—measuring performance, gather-ing occupational health medical information, and meetingepidemiologic data requirements. The system currently is beingused by six sites within the Army Medical Department. Thescope of the study extends the collection and managementof the ambulatory data through October 1987.

The system included a patient registration process, patientencounter (visit) forms, and a provider registration process.The patient is identified by the social security number ofthe sponsor and a two-digit prefix referred to as the FamilyMember Prefix. A unique number has been developed toidentify each provider. These two elements are the “keys”to the database.

The system employs microcomputers, mark sense forms,and scanners to update the data base. Each remote site canproduce data locally or address the main data base at FortDetrick, where the central data base is maintained. Encounterforms were developed by panels of physicians on a consensusbasis for nearly three dozen specialties. The forms includecommon administrative and demographic data as well as diag-noses and procedures tailored to the particular specialty. Diag-noses are coded in ICD–9–CM and procedures in CYf-4.Both coding conventions have been “extended” to describeprovider-demanded uniqueness. The form requires a primarydiagnosis, which is the primary reason for the visit, and permitsmultiple secondary diagnoses.

Physician acceptance of the reporting system has varied.Some who were initially resistant are recognizing the potentialof the data for special studies and practice profiling, Themajor complaint is that it takes about a minute to completethe encounter form for every patient.

The future of the reporting system after the demonstrationends has not been decided. Consideration is being givento promulgating it throughout the Department of the Army,which has some 400 separate reporting outpatient facilitiesor clinics (not hospital-based) accounting for the majorityof the 24 million outpatients treated annually by the AMEDD.Interest also exists in implementation across the three services.

Collection and use of hospitaloutpatient data

Emily Goodwin described an ambulatory care data projectcarried out in 1983–84 by the United Hospital Fund, whichprovides philanthropic support to the voluntary hospitals inNew York City. The project centered on patient origin informa-tion from hospital-based ambulatory care visits. There arebetween 10 and 12 million outpatient department and

26

emergency room visits annually in New York City, and manyhospitals have attributed their poor financial condition to theprovision of these services. Success with a patient-origin studyon inpatient care in 1978 provided the impetus for thisundertaking.

Overall goals of the study were to allow public policydecisianmaking to be based on information, to provide newhospital management planning tools, to improve and providefor consistent information practices at the hospitals, and toprovide opportunities for health services research.

The first step was a data capacity study, which revealeda number of inconsistencies across hospitals, including thedefinition of a visit. Because the Fund had to rely on voluntaryhospital cooperation and could not reimburse the clinics forsubmitting data, the original goals of obtaining a year’s worthof data from all 83 voluntary and municipal proprietary hospi-tals in machine-readable form had to be modified. The finalstudy focused on the second quarters of 1983 and 1984, withall hospitals participating in the 1984 data collection. Specificdata problems and issues were as follows:

The study’s intention to collect a unique patient identifierwithin each hospital and a primary payer patient ID numberfor linkage across hospitals did not prove feasible dueto confidentiality concerns.Separate hierarchical classification schemes had to be de-veloped to accommodate the detail submitted on paymentsource and clinic type. The scheme for the latter wasquite complex but turned out to be more of an administra-tive categorization than a reflection of medical specializa-tion, as originally desired.The Fund experienced tremendous problems getting hospi-tals to follow simple formats and coding instructions.Reliance on the hospitals’ outpatient billing system forin-scope visits created some difficulties. For example,same-day admissions were not contained in this system,whereas ancillary service visits were included.Recording of primary diagnosis, or the reason the patientpresented for the visit, was subject to considerable clericalerror, After correcting the most obvious problems, thedata appeared consistent with other information.The volume of data—approximately 1.6 million outpatientdepartment visits and 650,000 emergency room visits—initially proved overwhelming for analysis. Developmentof special condensed and summarized files alleviated thisproblem.

The Fund concluded that the information collected wasreasonably accurate and had considerable value for planningpurposes. If the study were repeated, the Fund would definitelyconsider sampling visits so that more detail could be collected.Patient ID continues to be of interest, as are procedures.

Mrs. Goodwin endorsed the need for standard definitionsbut noted that voluntary compliance has limitations, and theexistence of clout usually improves reporting.

Ambulatory care data needs of Centersfor Disease Control (CDC) programs

Dr. John Livengood provided the Centers for DiseaseControl’s perspective as a user of ambulatory care data. Theunique function of CDC is epidemiologic surveillance, whichrequires a systematic approach to collection, analysis, andinterpretation of health data. This process must be closelyintegrated with the dissemination of knowledge to the peoplewho need to know, which is linked to CDC’S primary focuson prevention.

Although an explicit purpose of a surveillance systemis to detect and control epidemics, in Dr. Llvengood’s experi-ence this is seldom accomplished through routine data report-ing. Ambulatory records, however, have the potential ofidentifying an epidemic several weeks before it shows upin mortality records. Ambulatory surveillance data also canbe used to test hypotheses about modes of transmission ofillness, to report trends in illness, to evaluate control measuresand public health interventions, to monitor changes in infec-tious agents, and to detect changes in the pattern of generalhealth practice.

Increased attention to the ambulato~ care setting reflectsa growing interest in describing the total effect of illness,not only on costs but on overall quality of life. Dramaticshifts of care to the ambulatory setting have also increasedthe importance of data from this sector.

Sources of data used by CDC include vital statistics,hospital discharge surveys, physician visit surveys, and surveysof the general population, such as the National Health andNutrition Examination Surveys. Dr. Llvengood noted thatthe National Ambulatory Medical Care Survey conductedby NCHS is an important national resource and should bemaintained at least on its current triennial basis. The CDC’Smost basic surveillance is through the Notifiable DiseaseReports, which are designed to trigger a local response andsubsequently are transmitted to CDC on a State-by-Statebasis. These reports suffer from underreporting but do permitdetection of trends.

Three other data sources that CDC uses are the NationalDrug and Therapeutic Index (NDTI), the Ambulatory SentinelPractice Network (ASPN), and the National Electronic InjurySurveillance System (NEISS). The NDTI has been operatedby IMS America since 1960 and collects utilization datafrom office-based physicians with an emphasis on prescriptiondrugs. The ASPN began in 1981 and is a voluntary systemof office-based family practice physicians which focuses onspecific conditions over a defined time period. The NEISSis a stratified random sample of hospital emergency rooms,initiated in 1972, and providing particularly valuable informa-tion on involvement of consumer products in injuries.

Dr. Livengood also described the Behavioral Risk FactorTelephone Surveys, which are being conducted by more than23 States with CDC involvement.

27

Major problems with the data sources include lack oftimeliness, limitations in scope, concerns about data validity,and insufficient detail at the geographic and condition level.

Dr. Livengood noted a number of areas where additionaloutpatient data would be desirable, including diabetes andhypertension control, technology assessment, and use of self-testing devices. An ideal list of data items would containinformation on patient disability, occupation, risk factors,circumstances of injury, product involvement, and onset andduration of illness, as well as the more standard items.

For data improvement, Dr. Livengood would supportprocedures for validation of diagnosis and development ofan algorithm for the underlying cause of visit. Better methods

for tracking patients and following back on patients are alsoneeded.

Dr. Felts concluded the meeting by noting that the numer-ous presentations heard by the Subcommittee demonstrateboth the commonalities and differences in user data needs.They all support, however, the essential need for informationregardless of the reimbursement approach adopted by publicand private programs. These needs include monitoring thesystem and maldng judgments about adequacy of utilizationand medical necessity. Maintaining comparability of datasets will bean ongoing challenge.

Prepared by: Marjorie S. Greenberg, OPPEC, NCHS.

28

Appendix Ill.Place of service designations(Figures H/l)

FNurG Il. HCFA 1500

BECAUSE THIS FORM IS USED BY VARIOUS GOVERNMENTANO PRIVATE HEALTH PROGRAMS, SEE SEPARATEINSTRUCTIONS ISSUED BY APPLICABLE PROGRAM.

REFERS TO GOVERNMENT PROGRAMS ONLY

MEDICARE AND CHAMPUS PAYMENTS A patienf’a signaturerequests that payment be made and authorizes release of medicalinformation necessary to pay the claim. If item 9 is completed, thepatient’s signature authorizes releasing of the information to the insureror agency shown. In Medicare assigned or CHAMPUS participationcases, the physician agrees to accept the charge detemrmalion of theMedicare carrier or CHAMPUS fiscal intermediary as the full charge,and the petient is responsible only for the deductible, coinsurance.and noncovered services. Coinsurance and the deductible are basedupon the cherge determination of the Medicare carrier or CHAMPUS

fiscal intermediary if this is less than the charge submitted. CHAMPUSis not a health insurance program and rendera payment for health ben-efits provided through membership and affiliation with the Uniformedservices. Information on the patient’s sponsor should be provided inthose items captioned “lnsurecK i.e.. items 3, 6, 7, 6, 9 and 11.

BLACK LUNG ANO FECA CLAIMS The provider agrees to accept theamount paid by the Government as payment in full. See BlackLung FECA instructions regarding required procedure and diagnosiscoding systems.

SIGNATUREOF PHYSICIANOR SUPPLIER(MEDICARE, CHAMPUS,FECA AND BLACK LUNG)

I cerfify that the services shown on this form were medically indicated although mcidantal perl of a covered physician’s service, 3) they must

and necessary for the health of the patient and were personally be of kinds commonly furnished in physician’s offices, and 4) therendered by me or were rendered incident to my professional service services of nonphysicians must be included on the physician’s bills.by my employee under immediate personal supewiaion. except asotherwise expressly permitted by Medicare or CHAMPUS regulations. For CHAMPUS claims. I further certify that neither I nor any employee

For sewtces to be considered a ‘incident” to a physician’s professionalwho rendered the serwces are employees or membets of the Uniformedservices (refer to 5 USC 5536). For Black Lung clalms, I further certify

service, 1) they must be rendered under the physician’s immediate that the services periormed were for a Black Lung related disorder.personal supervision by hisr’her employee, 2) they must be an integral,

No Part B Medicare benefits may be paid unless this form is recewed as required by existing law and regulations (20 CFR 422 510)

NOTICE: Any one who misrepresents or Ialslfies essential reformation to recaive payment from Federal funds requested by this formmay upon conviction be subject to fine and imprisonment under applicable Federal laws.

NOTICE TO PATlENT ABOUT THE COLLECTION ANO USE OF MEDICARE. CHAMPUS. FECA. AND BLACK LUNG INFORMATION

We are authorized by HCFA, CHAMPUS and OWCP to aak you forinformation needed in the administration of fhe Meelcare. CHAMPUS.FECA, and BLACK LUNG programs. Authority to collect informationis in section 205(a). 1672 and 1875 of the Social Security Act asamended and 44 USC 3101, 41 CFR 101 et seq and 10 USC 1079and 1066.5 USC 6101 et seq: and 30 USC 901 et seq.

The mformafion we obtain to complete claims under these programsIS used 10 identity you and to determine your eligibllify. It is also used

to decide if the services and supplies you received are covered bythese programs and to insure thaf proper payment is made.

The Information may also be given to other providers of services,carriers, intermediaries, medical review boards and other organiza-

tions or Federal agences as necessa~ to admmtister these programs.For example. it may be necessary to disclose mformatlon about thebenefits you have used to a hosp!tal or doctor.

With the one exception discussed below. there are no penaltes underthese programs for refusing to supply information However, failureto furnish information regarding the medical sewices rendered or theamount charged would prevent payment of claims under these pro-grams. FaJure to furnish any other information. such as name or claimnumber, would delay payment of the claim

It is mandalo~ that you tell us if you are being treated for a work re-lated injury so we can determine whether, workers’ compensation willpay for treatment. Section 1877 (a) (3) of the Social Security Act pro-vides criminal penalties for withholding this in formahon

MEDICAID PAYMENTS(PROVIDER CERTIFICATION)

I hereby agree to keep such records as are necessary to disclose fullyfhe extent of services provided to individuals under the State’s Title

SIGNATURE OF PHYSICIAN (OR SUPPLIER)’ I certify that the

XIX plan and to furnish information regarding any payments claimedservices listed above were medically ind)cated and necessa~ to the

for providing such aetvicee as the State Agency or Oept. of Health andhealth of this patient and were personally rendered by me or my

Human Services may request I further agree to accept, as payment inemployee under my personal direction.

full, the amount paid by the Medicaid program for those claimsNOTICE: This is to certify that the foregoing information is true.

submitted for payment under that program, with the exception ofaccurate and complete.

authorized deductibles and coinsurance.

I understand that payment and satisfaction of this claim will be from Federal and State funds, and that any false claims, statements, or

documents, or concealment of a material fact, may be prosecuted under applicable Federal or State laws.

PLACE OF SERVICE CODES: TYPE OF SERVICE CODES:

1- (IH) - Inpatient Hospital 1 - Medical Care

2- (OH) - Outpatient Hospital 2 - Surgery3- (o) - Doctor’s Office4- (H)

3 - Consultation- Patient’s Home

5-

4 - Diagnostic X-Ray

- Day Care Facility (PSY)

6-

5 - Diagnostic Laboratory

- Night Care Facility (PSY) 6 - Radiation Therapy

7- (NH) - Nursing Home 7 - Anesthesia

6- (SNF) - Skilled Nursing Facility9- - Ambulance

6 - Assistance at Surgery

O - (OL) - Other Locations9 - Other Medical Service

A - (IL) - independent LaboratoryO - Blood or Packed Red CellsA - Used DME

B - (ASC) - Ambulatory Surgical CenterC - (FITC) - Residential Treatment Center

F - Ambulatory Surgical CenterH - Hospice

O - (STF) - Specialized Treatment FacilityE - (COR) - Comprehenswe Outpatient

L - Renal Supplies tn the HomeM - Alternate Paymenl

Rehabilitation Factlity for Maintenance Dtalysls

F - (KDC) - Independent Kidney Disease N - Kidney Donor

Treatment Center V - Pneumococcal VaccineY - Second Opinion on Electwe SurgeryZ - Third Opinion on Electwe Surgery

29

Figure Ill. Medicare Carriers Manual, Part 3-clsims process

13710 (cont.) REPORTS AND STATISTICS 03-85

Field Position Number of Item Description CommentNumber Columns

4. 10 (1) HCPCS Indicate the procedure code 1) HCPCSIndicator reported as follows: codes include:

1 = HCPCS All CPT-4codes,HCFAnational

2 = Other codes andlocal codes(WXYZ)assigned byyou.2) Othercodesinclude: anycodeusedbefore con-version toHCPCS.

5. 11-12 (2) Modifier Show HCPCS 2-digit modi-fiers received on theclaim. If multiplemodifiers are reported,show 99. If no modifierwas used Ieave blank. Also~w any modifier you addfor administrative purposes.If you have not convertedto HCPCS, show the %pecialaction code” or other codeused to modify the procedurecode. Provide documentationof these special codes.

6. 13 (1) Place of This field denotes theService place of service. Use

the following codes:‘~1”= Office“2” = Home‘!3” = Inpatient Hospital114!1= SNF

These codesare consistentwith the codesand defini-tion used forthe PaymentRecord,s 13030.4

13-398

30

Rev. 1084

03-85 REPORTS AND STATISTICS 13710 (cont.)

Field Position Number of Item Description CommentNumber Columns

(6. cont.) “5” = Outpatient Hospital“6” = Independent Laboratory“7” = Other118!!= Independent Kidney

Disease Treat mentCenter

“9” = Ambulatory SurgeryCenter

‘~H1t= Hospice

7. 14 (1)

8. 15-20 (6)

Qpe of This field represents theService type of service. Use the

following codes:“1” = Medical Care“2’! = Surgery“3” = Consultation“4” = Diagnostic X-Ray~151~= Diagnostic

Laboratory‘?6tf= Radiation Therapy“7’?= Anesthesia“8” = Assistance at Surgery‘!9” = Other Medical Service“O” = Whole Blood or Packed

Red Cells“A” = Used DME“F” = Ambulatory Surgical

Center (FacilityUsage)

‘?L’t= Renal supplier in thehome

‘WI”= Monthly Cavitation Payment(Dialysis)

‘~N”= Kidney DonorT = Pneumococcal Vaccine“Y” = Second Opinion on

Elective SurgeryZ“ = Third Opinion on Elective

Surgery

These codesare consistentwith the codesand definitionused for thePaymentRecord,S1303.4.

Frequency Show the total number of timesthat this procedure code/modifier,occurred within “this locality,specialty, T/S, P/S sequence andwas recorded in History. This

Rev. 1084 13-399

31

Figure IV. MetropoMsn Life Insursnce Compsny placeof serviceCo&!s

Typa of Facility

01- Hospital (Rsgular-f3eneral)02- Clinic03- Paychiatnc Hospital04- Christian Science Hospital05- osteopathic Hospital06- Nursing Home07- Extendad Care Facility03- sanitarium09- School Custodial/Developmental10- School-Teaching11- Health Resort12- Ambulatory Surgicenter13- Drug/Alcohol Rehabilitation Center14- Home Health Agency15- Infirmary16- Veterana Administration17- Medical Rehabilitation Facility18- Hospice (terminally ill)19- Sirfhing Cantera

Clinic Modifiera

1- Hemodialysis Treatment Center2- Hemophilia Treatment Canter3- Freestanding Emergency Treatment Center

~um‘~~~= ;~~ue *M A-w pkitxt ofSenrics

[Placa of sarvicscedesfilledin bypersonccmplafingclaim form]

Plac8 of service code=

1 - (IH) - Inpatient Hospital2 - (OH) - Outpatient Hospital3- (0) - Doctor’s Office4 - (H) - Patient’s Home5- - Day Care Facility (PSY)6- Night Care Facility - (PSY)7 - (NH) : Nursing Home

- (SNF) - Skilled Nursing FaciMy:- Ambulance0 - (OL) = Other LocationsA - (IL) - Independent Laboratory

Other Medical Surgical Facility; ~ (RTC) = Residential Treatment Center

- (STF) - Specialized Treatment Facility

NOTES This &Me is on the back of claim form.

The number and fist may vary fmm plan to plan but this is the moat emmnen list uaad byWe Shafd p!arm

Figuremt31M.t313ar&f3hM3 Shiskl AssocMh pkiceof-

[Pfssa of semica filled in by gwson prwaasing the claim at fhs PM]

Description

1. Hospital, inpatient2. Hospital-affiliated hospice3. Rehabilitation hospital, inpatient

4. Hospital, outpatient5. Hoapifal-baaed ambulatory surgical facilii6. Hospital, outpatient hospice services7. Rehabilitation hospital, outpatient

8. Provider’s office (includes Professional Other Provider)9. Hospice services rendered in the prwider’s office

10. Patient’s home11. Hospice services rendered in the patient’s home

12. Psychiatric facilii, inpatiinf13. Psychiatric facilii, outpatient14. Psychiatric day-care facifii15. Psychiatric night-care facifity16. & 17. Residential substance abuse treatment

facility (includes A!cohotism TreatmentFacilii and Drug Abuse Treatment Facility)

18. & 19. Outpatient substance abuse treatmentfacilii (includes Alcoholism TreatmentFacility and Drug Abuse Treatment Facility)

20. Independent clinical laboratory

21. Numlng home

22. Skilled nursing faciiity/exfendad care facility

23. Ambulancq ground24, Ambulancq air25. Ambulancq sea

26. Ofher unlisted licensed facilii (Facilii Ofher Provider)27. Hemophilia treatment canter

28. Freestanding ambulatory mediial facility29. Freestanding dialysis facility30. Freestanding ambulatory surgical facility31. Freestanding alternate birthing canter32. Freestanding cardiac rehabilitation facility33. Freestanding hospice facility

34. Pharmacy

NOTES Moat plans enfy mcettl the detailed servioa sits fw National Aexunts.

fnfomsatkm pmv”@ cm tie dmgnda and pmcadure Ox#as alfcw pfan parsonnal todetermine tfw app+epdate place of service frem the claim form.

The actualbeationof fhawvice sitecannotb3 determined for every claim.

U.S. GWERNNENT PRINTINGOFFICE:1987- 181- 33 q 40022

32

Appendix IV.Data flow for traditionalfee-for-service physicianreimbursement(Figure VII)

FigureW. Dataflow fortfadwal fee-for-servicephyaklan raifnburaefnent

‘~Patient-physician encounter

Preparation of claim

or encounter form

v i

(usually with symptoms (If hospital outpatient site of

or diagnosis describad service, procedures described

by ICD-94M codee, by vol. 3 of ICD.9-CM codes

and services [procedures] for institutional reimbursement)

by CPT4 codes)

i J i i I i i 11

Blue Cross/Commercial

Medicare MadiceraState

Employers Blue Shield part B Chamrxmintarmadisry

insurer(e) aganc”msetc.

plan carrier[for cmtpatiantPart B procssebg)

>

‘-~

-

- -----~ ,f’’:’s -

3Resaarch

vH.C.F.A.

“B-MAD’U.S.A.

detsfilsCongrass

Policy

H.C.F.A. & D.H.H.S.Functions

. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Statistical : Adminiatretiva :Menegfmat end

. Policy

1 ) L-

~1 ~, ~.earth...........................................

Note: The schemetic doss not reflect the dsts flow for cepitation end HMO physicien reimbursement. Oate collection and date flow for cavitation

systems, such sa come parts of DOO, VA, CHAMPUS, and other public or private sector organizatiena, ara Iimitad to sdminktrative naads

with local application, ratention, and distribution.

33

Our warehouses here at the Government Talents, and The Back-Yard Mechanic.Printing Office contain more than 16,000 Books on subjects ranging fromdifferent Government publications. Now agriculture, business, children,we’ve put together a catalog of nearly and diet to science, space exploration,1,000 of the most popular books in our transportation, and vacations. Find outinventory. Books like Infant Care, what the Government’s books are all

Nationul Park Guide and Map, The about. For your free copy of our

Space Shuttle at Work, Federal Benefits ‘ew bestseller catalOg’ ‘rite—for Veterans and Dependents,

F-w !!%2?%. q,().fl

Merchandising Your Job y~y i&lMEl,E-awi3

Bestsellers

KEEPINGPACEISN’T EASY!

In todayss health world, nohealth professional can affordto fall behind. Let us help.Pub//c Ifedth Reports- the journal of theU.S. Public Health Service - brings together inone convenient source informationyou need onFederal health policies, innovative programs andservices of public and private agencies, researchin health fields and public health around the world.

Your subscription will bring you six wide-ranging issues a year-each one includingmore than a dozen peer-reviewed papers byrecognized authorities in health and relatedfields, thought-provoking commentary on healthissues and timely information on the preventionand control of disease.

The cost of all this? Only $21 a year.Don’t miss an issuel Subscribe now. Justcomplete and mail the coupon below.

TO:Superintendent of DocumentsDept. 3$KKU.S.GovernmentPrintingOfficeWashington, DC. 20402

Name- F-t L@I I 11111 I Ill

streetWkJmss

I 111111 II IllGmIXWnmeor additicaalWdres lme

1111 111111111C@11111111111111 Ill mm(CMcouni@11111111111111111111111111

•l YES!$?%idme PWJC HEALTHREPORTS (HSNIHA)for oneyear,

•l Here’s my check for $ 1made oui to “Superintendentof Documents.”($hd?scriptions we $21 per year domestic; $26.75 foreign.)

•l Charge my— GPO DepositAccount UuIJnu ExpirationDate

—visa — Mastercard 111111111111111111111

NiaHl$3IhdMlkaukm$s!Awdihdblh tilhmmgjh GPO

skmilfmqg OmMm’AkxImnrtm

Standing Order Accounts with the Government Printing Office--an easy way toorder and receive al I new publications from the National Center for HealthStatistics.

I

Vital and Health Statistics series

❑ ZHS1

O ZHS2

❑ ZHS3

❑ ZHS4

❑ ZHS5

❑ ZHS1O

❑ ZHS1l

•1

•1❑

•1

•1•1

•1•1•1

❑

•1

ZHS13

ZHS14

ZHS15

ZHS20

ZHS21

ZHS23

Program and Collection Procedures

Data Evaluation and Methods Research

Analytical Studies

Documents and Committee Reports

Comparative International Vital

Charge:

and Health Statistics Report

Data From the National Health Interview Survey Credit

Data From the National Health Examination Survey and Card No. I II 1 1 I 11 I I I 1 I I I 11 [111

National Health and Nutrition Examination Survey

Data on Health Resources Utilization Expiration Date

Data on Health ResourcesMonth/Year 1111

Data From Special Surveys

Data on Mortality Expiration Date

Data on Natality, Marriage and DivorceGPO Deposit Account:

Data From the National Survey of Family Growth

Vital Statistics of the United States Decsosit Account No.

ZVSN1 Natality

ZVSM2 Mortality, Parts A&B

ZVSM3 Marriage and Divorce

Health, United States

ZHUS

Bibhhgraphyon Hwtth

ZBHI

~–aNew Account ($50.00 minimum)

Indexes <i! Authorizing signature

(Orders not valid unless signed)

Ship to:

Company or personal name

11111 I 1 I 1 1 1 1 1 1 I I 1 1 I I 1 I I 1 I 1 I 1 I IAddltlonal address/attention line

11111111111111111111111 111111]Street address

I I I 1 I 1 1 I 1 I I I I 1 I 1 I 1 I I 1 1 I ICity State ZIP Code

1111111111111111111111 I 1 1 I I 1 I I Customer Phone Number

(or Country)Home Office

I

11111111111111111111111 111111] I I IPLEASE PRINT OR TYPE

(Please #nclude Area Code)

Mail ORDER FORM to: Cspt. 40A, Superintendent of Documents, U.S. Government Printing Office,Washington, DC 20402

Vital and Health Statisticsseries descriptions

SERIES1.

sEFfl Es2.

SE Ffl ES3,

SE RIES4.

SERIES 5.

SERIES 10.

SERIES 11.

SERIES 12.

SERIES 13.

Programs and Collection Procedures–Reports describing

the general programs of the National Center for Health

Statistics and its offices and divisions and the data col-

It?ction methods used. They also include definitions and

other material necessary for understanding the data.

Data Evaluation and Methods Research –Studies of new

statistical methodology including experimental tests of

nc!w survey methods, studies of vital statistics collection

methods, new analytical techniques, objective evaluations

of reliability of collected data, and contributions to

statistical theory. Studies also include comparison of

U.S. methodology with those of other countries.

Analytical and Epidemiological Studies–Reports pre-

senting analytical or interpretive studies based on vital

and health statistics, carrying the analysis further than

the expository types of reports in the other series.

Documents and Committae Reports–Final reports of

major committees concerned with vital and health sta-

tistics and documents such as recommended model vitalregistration laws and revised birth and death certificates.

Comparative International Vital and Health Statistics

Reports-Analytical and descriptive reports comparing

U.S. vital and health statistics with those of other countries.

Data From the National Health Interviaw Survey –Statis-

tics on illness, accidental injuries, disability, use of hos-

pital, medical, dental, and other services, and other

health-related topics, all based on data collected in the

continuing national household interview survey.

Data From the National Health Examination Survey and

the National Health and Nutrition Examination Survey–

Data from direct examination, testing, and measurement

of national samples of the civilian noninstitutional ized

population provide the basis for (1) estimates of the

medically defined prevalence of specific diseases in the

United States and the distributions of the population

with respect to physical, physiological. and psycho.

logical characteristics and (2) analysis of relationships

among the various measurements without reference to

an explicit finite universe of persons.

Data From the Institutionalized Population Surveys-Dis-

continued in 1975. Reports from these surveys are in-

cluded in Series 13.

Deta on Heelth Resources Utilization–Statistics on the

utilization of health manpower and facilities providing

Iongterm care, ambulatory care, hospital care, and family

planning services.

---, -,... -–. ,, ...,- . . . . .....>Emlc> 14.

SERIES 15.

SERIES 20.

SERIES 21.

SERIES 22.

SERIES 23.

uala on rneamr nesources: rwanpower and hactlmes-

Statistics on the numbers, geographic distribution, and

characteristics of health resources including physicians,

dentists, nurses, other health occupations, hospitals,

nursing homes, and outpatient facilities,

Data From Special Suweys–Statistics on health and

health-related topics collected in special surveys that

are not a part of the continuing data systems of the

National Center for Health Statistics.

Data on Mortality-Various statistics on mortality other

than as included in regular annual or monthly reports.

Special analyses by cause of death, age, and other demo-

graphic variables; geographic and time series analyses;

and statistics on characteristics of deaths not availablefrom the vital records based on sample surveys of those

records.

Data on Natality, Marriage, and Divorce-Various sta-

tistics on natal ity, marrtage, and divorce other than as

included in regular annual or monthly reports. Special

analyses by demographic variables; geographic and time

series analyses; studies of fertility; and statistics on

characteristics of births not available from the vitalrecords based on sample surveys of those records.

Data From the National Mortality and Natality Survs?ys-

Discontlnued in 1975. Reports from these sample surveys

based on vital records are Included in Series 20 and 21,

respectively,

Data From the National Survey of Family Grov#th-

Statlstlcs on fertility, family formation and dissolution,

family planning, and related maternal and infant health

topics derived from a periodic survey of a nationwide

probability sample of ever-married women 1544 years

of age,

For answers to questions about this report or for a list of titles of

reports published in these series, contact:

Scientific and Technical Information BranchNational Center for Health Statistics

Public Healtfr Sewice

Hyattsville, Md. 20782

301-436-8500

U.S. DEPARTMENT OF HEALTH ANDHUMAN SERVICES

Public Health SewiceNational Center for Health Statistics3700 East-West HighwayHyattsville, Maryland 20782

OFFICIAL BUSINESSPENALTY’ FOR PRIVATE USE, $300

BULK RATEPOSTAGE & FEES PAID

PHS/NCHSPERMIT NO. ~

DHHS Publication No. (PHS) 87-1461, Series 4, No. 24