managed care and competitive health care markets: the twin ... · care, in part in response to the...

Managed Care and Competitive HealthCare Markets: The Twin Cities Experience

July 1994

OTA-BP-H-130NTIS order #PB94-193240

LIBRARYOFFICE OF TECHNOLOGY ASSESSMENT

CONGRESS OF THE UNITED STATESWASHINGTON, D. C. 20

Recommended Citation: U.S. Congress, Office of Technology Assessment, Managed Careand Competitive Health Care Markets: The Twin Cities Experience, OTA-BP-H- 130(Washington, DC: U.S. Government Printing Office, July 1994).

Foreword

Reform of the health care system is at the top of the nation’s do-mestic policy agenda. As policy makers consider the manydirections the nation could take, they often look to the states aslaboratories.

The Twin Cities (Minneapolis and St. Paul, Minnesota) are often citedas a model of a competitive market and a potential case study for man-aged competition proposals. This background paper describes some ofthe developments and current characteristics of the Twin Cities’ healthcare system. The paper emphasizes recent changes in the market forhealth care and health insurance in the Twin Cities, including the growthof managed care organizations, the growth of integrated delivery sys-tems, the development of health insurance purchasing coalitions, and re-cent state health care reforms. The report concludes with a discussion ofpotential lessons from the Twin Cities for the health reform debate.

This paper was prepared as background for the Office of TechnologyAssessment study Understanding the Estimates Under Health Reform.The assessment as a whole was requested by the members of theTechnology Assessment Board (see inside front cover) and Senator TedStevens.

This background paper was prepared under contract to OTA by JonChristianson, Ph.D, Bryan Dowd, Ph. D., John Kralewski, Ph.D, andCatherine Wisner, Ph.D. It was reviewed by a number of experts in healthpolicy and health care delivery. OTA gratefully acknowledges the con-tribution of each of these individuals.

ROGER C. HERDMANDirector

. . .Ill

Project Staff.

Clyde J. BehneyAssistant Director

Sean TunisHealth Program Director

Denise DoughertyCo-Project Director

Tami MarkCo-Project Director

David GrabowskiResearch Assistant

ADMINISTRATIVE STAFFBeckie EricksonOffice Administrator

DanielB.CarsonPC Specialist

CONTRACTORS

Jon ChristiansonBryanDowdJohn KralewskiCatherine Wisner

Carolyn MartinWord Processing Specialist

iv

1

2

3

4

5

6Implications for Managed Competition Reform 50Concluding Comments 52

c ontents

Summary 1Managed Care and Integrated Delivery Systems in the

Twin Cities 1Development and Role of Purchasing Coalitions 2Minnesota Care 3Health Care Expenditures in the Twin Cities 3Relevance of the Twin Cities Experience 4

Introduction 5

Health Care Delivery in the Twin Cities 7The Twin Cities’ Health Care System 7Development of Minnesota Care 13Summary 16

Growth of Managed Care and IntegratedDelivery Systems 17Developmen[ of the HMO Market: 1970-1980 17Competition Among HMOs and Its Effects:

1980-1990 20The Consolidation of the Twin Cities Health Care

Market 25

Development and Role of PurchasingCoalitions 35Business Health Care Action Group 35The State of Minnesota Group Insurance Program 37The Minnesota Employees Insurance Program 44Employers’ Association Care Buyers Coalition 46Summary 47

Relevance of the Minnesota Experience 49

APPENDIX A: Acknowledgments 53

REFERENCES 54v

Summary

The Twin Cities (Minneapolis and St. Paul, Minnesota) arefrequently identified as a community where a competitivehealth care market has developed. Consequently, theyhave been the focus of a substantial number of empirical

studies through the years and have sometimes served as a ‘*mod-el” for various elements of cument “managed competition” ap-proaches to health care reform. If, indeed, the Twin Cities are atthe vanguard of managed competition in health care, it is impor-tant to understand how their health care delivery system hasevolved over the past two decades, why it is now undergoing arelatively dramatic transformation, and the effects of thesechanges. ]

MANAGED CARE AND INTEGRATEDDELlVERY SYSTEMS IN THE TWIN CITIES

The health care delivery system in the Twin Cities is bestknown national] y for its reliance on health maintenance organiza-tions (HMOs), and for the proportion of community residents en-rolled in HMOs. A variety of hypotheses have been offered in ex-planation of why HMOs were formed and prospered in the TwinCities, but no definitive answer to this question is possible. It isclear, however, that during the past two decades HMO enrollmentgrew rapidly. From .197.1. to.1978, HMO enrollment grew at anannual rate of 27 percent. Enrollment continued to grow duringthe 1980s, reaching 50 percent of the population by the end of thedecade.

1~1~ background paFr f(xuses on tie organization of health care delivery and on health care costs and al~(w4 it discusses Issues ‘Uch asconsumer satisfaction, health outcomes, and access, and recognizes that these are critical issues, it does not focus on these issues.

II

2 | Managed Care and Competitive Health Care Markets: The Twin Cities Experience

The HMOs in the Twin Cities encompass a va-riety of organizational forms and sponsorship ar-rangements, with m|st physicians affiliated withone or more HMO by the early 1980s. The 1980ssaw important changes in Twin Cities HMOs andtheir relationships with providers. These changesincluded the development of new products, suchas preferred provider organizations, and the insti-tution of more aggressive management strategies,such as concentrating patients at lower cost hospi-tals .

A number of HMOs merged during the 1980sand as a result of the merger activity it appearslikely that three or four large organizations willdominate health care delivery in the Twin Cities.Managers of Twin Cities’ health plans point to therecent passage of state health reform legislation(Minnesota Care), with its emphasis on the forma-tion of integrated delivery networks, as an impor-tant catalyst for consolidation. However, the trendtoward consolidation began prior to Minnesota-Care, in part in response to the demands of em-ployers as interpreted by health plans. Employersencouraged HMOs to develop a range of benefitoptions, and to broaden the geographic coverageof their networks. One way for an HMO to expandgeographically was to merge with another HMO.Some employers also believed that larger HMOshad greater potential to efficiently integrate serv-ice delivery. The responsiveness of HMOs to em-ployer concerns was heightened by the formationof a series of buyers’ coalitions among privatefirms and the adoption of a new buying approachby state government for its employees.

In addition to HMO consolidation, the 1980swi tnessed cons ide rab le conso l ida t ion amongTwin Cities’ hospitals. Four major, multihospitalsystems were formed in the late 1980s through aseries of consolidations and mergers. Some hospi-tals reported pursuing the development of multi-hospital organizations as a means of negotiatingmore effectively with HMOs over prices and ofpositioning themselves to offer broad geographiccoverage for HMO enrollees.

DEVELOPMENT AND ROLE OFPURCHASING COALITIONS

During the past decade, several private andpublic employers in the Twin Cities made signifi-cant changes in the way they purchase health care.One often-cited example is the formation of theBusiness Health Care Action Group (BHCAG), aconsortium of major private sector employers inthe Twin Cities. In 1991, these self-insured firmsjoined together to create a new health plan optionfor employees and dependents. Through an exten-sive negotiation process, one plan was selected byBHCAG. BHCAG has taken a very proactive ap-proach to the delivery of health care to its em-ployees. It is actively collaborating with the healthplan in the development of practice guidelines,and the institution of programs to improve qualityof care.

A second approach to buying is exemplified bythe State of Minnesota’s Group Insurance Pro-gram, which covers 144,000 individuals, includ-ing employees, dependents, and ret irees. Unti l1985, this program offered heal th benefi ts inmuch the same way as many other large employ-ers. Its contribution to premiums was tied to thepremium for the fee-for-service insurance optionin the program. In 1985, the state consolidated itsHMO offerings and instituted a new contributionformula under which employees were required topay the premium difference out-of-pocket if theydid not enroll in the low-cost plan. The 1985 re-forms were followed by a substantial shift in en-rollment from traditional indemnity plans to man-aged ca re p lans . Over t ime , HMO premiumsdeclined relative to premiums of other options,and recently the overall rate of increase in pre-miums has been quite low. Two other purchasingprograms, directed at individuals and smaller em-ployers,. have been- initiated recently. The statelegislature created the Minnesota Employees in-surance Program (MEIP) as part of the 1992 Min-nesotaCare legislat ion. Private businesses withtwo or more employees are eligible to enroll their

Chapter 1: Summary | 3

employees in MEIP. Four health plan options areavailable, and employers must pay at least 50 per-cent of the premium for single coverage, but can-not pay more than 100 percent of the cost of thelowest priced plan. The Minnesota Employers’Association is a nonprofit association of approxi-mately 1,300 businesses that offers a health insur-ance program to its members. Services are deliv-ered through a preferred provider networkcurrently managed by the Prudential InsuranceCompany and enrollment consists of approxi-mate y 5,000 employees and dependents. Becausethese programs are so new, it is difficult to predictthe ultimate impact of the MEIP and Buyers Co-alition efforts.

MINNESOTA CAREIn concert with the rapid developments in the

private sector, the Minnesota state legisture hasbeen pursuing major reforms which are likely tohave a significant effect on the Twin Cities mar-ket. In 1992, the Minnesota state legislature en-acted Minnesota Care. The general objective ofMinnesota Care is to enhance the availability of in-surance for uninsured people in the state, while atthe same time reducing health care cost increases.Although the legislation is still evolving, as it nowstands it aims to encourage the development of in-tegrated service networks (ISNs), to be formed byproviders or purchasers of medical care, with thecharge of providing a comprehensive set of healthservices to a designated population for a prospec-tively set budget. The State Health Commissionerhas the power to approve ISN arrangements andcan issue state exemptions from antitrust liabilitythat might arise in forming such relationships.Each ISN will be subject to an overall limit on therate of growth in its annual expenditures. A regu-lated all-payer option will apply to providers de-livering health care outside an ISN. In addition,Minnesota Care will create limits on total statehealth care spending, and the Commissioner ofHealth is charged with enforcing annual limits on

HEALTH CARE EXPENDITURESIN THE TWIN CITIES

Given the dramatic transformations in the TwinCities health care market, how do health care ex-penditures in the Twin Cities compare with othermetropolitan areas? Two recent studies havefound that the level of health care expenditures islower in the Twin Cities than in other metropolitanareas, while a third study found the opposite. Un-fortunately, all three of the studies have seriousflaws. None controlled for differences in benefitcoverage, nor the size and characteristics ofgroups in the metropolitan areas. Moreover, onestudy only looked at indemnity insurance and ex-cluded HMO coverage. Another only comparedcosts across selected cities (e.g., omitting Boston,Massachusetts, and Washington, DC).

Some data indicate that health care costs andexpenditures in the Twin Cities may be rising at aslower rate than in the nation as a whole. The med-ical price index in the Twin Cities was above thenational average from 1981 until 1987. However,since 1987 it has been below the national average.One study found that between 1971 and 1990 theannualized rate of increase in hospital costs percapita in the Twin Cities was 10.0 percent,compared with 11.2 percent nationwide. Overall,the evidence on health care costs in the Twin Ci-ties is limited and, in some cases, contradictory.Whether expenditures for health care in the TwinCities are higher or lower than in other metropoli-tan areas is unclear.

Another important question is how HMOs haveinfluenced the level of health care expenditures inthe Twin Cities. Several studies done during thelate 1970s and early 1980s examined how HMOenrollment in the Twin Cities affected health carecosts. Because they used data from differentsources and covering different time periods, theresults of these studies are sometimes difficult toreconcile. In general, however, it appears thatstudies based on data from the late 1970s and early1980s offer little support for the hypotheses that

the rate of increase in health care costs. HMO growth and competition among HMOs


would control premium increases or induce com-munity providers to contain their costs. For exam-ple, group and staff HMOs during this period ap-peared to benefit from a “favorable selection” ofrelatively healthy enrollees. As a result, one studyconcluded that employers who offered employeesa choice of HMOs and fee-for-service insuranceplans saw total health insurance costs increase.Other studies found that the hospital market forHMO enrollees was not price competitive in thelate 1970s.

Studies conducted using data from the mid-to-late 1980s present a somewhat different picture.Group and staff model HMOs appeared to concen-trate their patients at selected hospitals, with theprice of hospital services playing an importantrole in the selection of hospitals. Lengths of hospi-tal stays for enrollees in these plans were signifi-cantly shorter than for indemnity plans. Althoughthese results suggest that health care expendituresmay be reduced as plans deal more aggressivelywith providers, no recent studies have directly ex-amined the effect of these changes on health careexpenditures in the Twin Cities.

Some evidence, although imperfect, suggeststhat HMO enrollment may have contributed to thereduction in hospital beds in the Twin Cities. Al-though hospital capacity declined nationwide dur-ing the 1970s and 1980s, hospital capacity in theTwin Cities declined even more dramatically andhas continued to decline in the 1990s.

RELEVANCE OF THETWIN CITIES EXPERIENCE

The well-known limitations of the ● “case-study” methodology suggest that drawing generalconclusions from the experience of the Twin Ci-ties health care market is difficult. However, sev-eral tentative conclusions are suggested by thehealth care delivery system’s evolution and per-formance in the Twin Cities. They are:

● Development of managed competition is likelyto be associated with reconfiguration of com-munity hospitals, such as the creation of multi-hospital systems.

Ž Managed care organizations will respond com-petitively to even moderately sized purchasingcoalitions, for example, by merging to providegreater geographic access.

● Organization of the demand side of the healthcare market under managed competition is like-ly to encourage the consolidation of providersand managed care plans, suggesting that spe-cific public and private sector strategies may beneeded to maintain a competitive market struc-ture.

Introduction

The effectiveness of managed competition as a strategy tocontain the growth in health care expenditures is now be-ing debated at both the state and national level. A “man-aged competition” approach has been proposed by the

Clinton Administration as the cornerstone of its plan for healthcare reform and serves as the basis for several legislative propos-als for reform as well. In discussions of “managed competition,”the Twin Cities market is sometimes offered as an example of anarea where competition is “working,” where competition has“failed,” or where the elements of “managed competition” arenow “in place.” The purpose of this background paper is to clarifythe features of managed competition as it has developed in theTwin Cities, to describe the relatively dramatic changes that arenow transforming the Twin Cities’ health care delivery system,and to discuss the implications of these developments for nationalhealth care reform.

The first section of this background paper provides an over-view of health care delivery in the Twin Cities, as well as the com-ponents of Minnesota Care, the state of Minnesota’s health carereform initiative. Understanding Minnesota Care is important forthe subsequent discussion because it appears to be a significantstimulus for the restructuring of the market that is now occurring.The second section of the background paper describes the evolu-tion of the health care market in the Twin Cities since the 1970s.The “conflicting evidence” about the results of competition in theTwin Cities appears to reflect, at least in part, the time periodchosen for examination. This section also highlights the rapidchanges that can occur in the configuration of health care deliverysystems and the difficulty this poses for making predictions of

2

6 I Managed Care and Competitive Health Care Markets: The Twin Cities Experience

future performance based on past studies. Thethird section focuses on changes that have oc-curred in the demand side of the Twin Cities’ mar-ket. An important element of the Twin Cities’*’story” is the recent evolution in the way privateemployers and state government have approachedthe purchase of health care. The concluding sec-tion summarizes the findings of the backgroundpaper and discusses their implications. Obvious-

ly, there are dangers involved in generalizing fromthe experience of a specific health care market.However, if that experience is interpreted cau-tiously, there may also be important lessons to belearned. It is hoped that this background paper willcontribute to a better understanding of what can belearned from the Twin Cities’ experience as thenation debates health care reform.

- Heal thCare

Deliveryin the

Twin Cities 3

The discussion in this section is presented in two parts. Thefirst part contains descriptive data on the health caresystem in the Twin Cities. In some cases, data are onlyavailable on a statewide basis, and this is noted in the dis-

cussion. The second part summarizes the recent health care re-form legislation passed by the state of Minnesota. It attempts toidentify the components of the legislation that have influencedthe recent reconfiguration of health care delivery in the TwinCities, as described in subsequent sections.

THE TWIN CITIES’ HEALTH CARE SYSTEMI Health Services Use and Expenditures

Personal health expenditures per capita for 1992 were reportedto be $3,166 for Minnesotans, compared with $3,286 nationally(19). Table 3-1 compares the public/private distribution of healthcare expenditures in 1991 for the state of Minnesota with the na-tional average. This comparison is based on estimates of personalhealth care expenditures for 1991 by payer category for Minneso-ta residents, using data supplied by the Minnesota Department ofHealth (MDH) as of April 1993. A higher percentage of healthcare dollars was spent by the public sector in Minnesota than inother states, with state and local spending constituting a muchhigher percent of public spending, indicating higher expendituresfor Medicaid and local assistance than the national average.

To assess Twin Cities’ expenditures on health care, it would bedesirable to compare health insurance premiums in the TwinCities with those in other metropolitan areas, focusing on levelsand changes over time. However, there are no published data


state ofNational (%) Minnesota (%)

Private 56 48

Public 44 52.Federal 30 30State/local 14 22

SOURCE Health Care Financing Admmistration, Off Ice of Actuary

Off Ice of National Health Statistics, “Standardized Per Capita Rates ofPayment, ” Baltimore, MD, 1994

series that permit such a comparison, while con-trolling for differences in standards of benefit cov-erage and size of groups by metropolitan areas.Two widely quoted studies that attempt to do sohave important limitations.

In assessing the relative cost to consumers ofhealth care in Minnesota, Milliman and Robert-son, Inc. reported that Minneapolis/St. Paul wasthe second to the lowest metropolitan area inhealth care costs, at 18 percent below the nationalaverage in 1991 (45). The highest metropolitanarea that year was Miami/Fort Lauderdale, withcosts that were 38 percent above the national aver-age, while the lowest area was Charlotte, NC, withcosts that were 22 percent below the national aver-age. The Milliman and Robertson, Inc. estimatesinclude costs related to hospital inpatient services,hospital outpatient services, surgery, office visitsand other medical encounters, radiology, pathol-ogy, and prescription drugs (45). Not included incalculating costs were wellness benefits, such asperiodic examinations and immunizations. TheMilliman and Robertson, Inc. report data arebased on indemnity insurers, not including Medi-care and Medicaid coverage or HMO coverage(45). They also report that their data are compiledfrom publications such as the American HospitalAssociation and the claims experience of severalmajor insurers. Therefore, it is difficult to deter-mine the actual sources of the Milliman andRobertson data, or whether they pertain to expen-ditures by insurers for care, or the costs of deliver-ing care.

Foster Higgins compared indemnity plan pre-miums and managed care premiums using datafrom 2,448 employers in selected U.S. cities for

1992. Their data are collected nonrandomly fromclients and large emplyed groups. As with Milli-man and Robertson, sample sizes generally are notadequate to make statistically valid comparisonsacross metropolitan areas. Moreover, the data arenot adjusted for systematic differences in benefitsand demographics across employers in differentcities. Foster Higgins reports that average pre-miums per employee for HMO plans are consis-tently lower in all the cities examined (table 3-2).Preferred provider organization premiums arehigher for some cities compared with indemnityplan premiums. Minneapolis/St. Paul has both thelowest average indemnity plan premiums and thelowest HMO average premium cost per employeecompared with other reported cities, subject to thecaveats noted above (29).

Data from the Cost of Living Index for SelectedMetropolitan Areas, compiled by the Associationfor U.S. Chambers of Commerce (ACCRA),shows the Minneapolis/St. Paul, MN-WI Metro-politan Statistical Area (MSA) in 1992 to have acomposite cost of living index equal to the aver-age for the nation, but a health care cost of livingindex 8 percent above the national average (’65).This contrasts sharply with the results reported byMilliman and Robertson, where the Twin Citiesranked 18 percent below the national average inmedical costs. Given these contradictory data,along with questions about the validity of the dataused in making comparisons, it is difficult to accu-rately assess whether expenditures for health carein the Twin Cities are higher or lower than in othermetropolitan areas.

Data supplied by the Health Care FinancingAdministration indicate that expenditures for fee-for-service Medicare beneficiaries are relativelylow in the Twin Cities, in comparison with othermetropolitan areas. Table 3-3 compares fee-for-service Medicare expenditures in the 20 U.S.counties with the largest enrollment in MedicareHMO risk contracts. These data show that onlyVolusia County, Florida, has lower AAPCC pay-ments than Hennepin and Ramsey counties in theTwin Cities MSA. The percentage change figuresshow

Chapter 3: Health Care Delivery in the Twin Cities | 9

Health Maintenance Preferred providerOrganizations Organizations

Indemnity plan Averagepremiums per premiums pew

city . employee employeeAtlanta $3,729 $3,311

Chicago 4,245 3,088

Cleveland 4,027 3,727

Dallas/Ft. Worth 3,917 3,330

Houston 3,627 3,575

Los Angeles 4,350 3,189

Minneapolis/St Paul 3,347 2,969

New York Metro 4,852 3,448

Orange County 4,276 3,124Philadelphia 4,696 3,319

Richmond 3,578 3,074

San Francisco 4,531 3,092

Seattle 3,554 3,092

HMOpremiums vs.

indermnity-11 .2%

-27.3

-7.4

-15.0

-1,4

-26.7-11,3

-28.9-26.9-29.3-14.1

-31.8-13.0

Averagepremiums per

employee$3,3633,6843,4593,8374,091

4,457

3,1213,8714,315

3,708

3,1834,459

3,114

PPO premiumsvs. indemnity

Cost-9 .8%

-132

-141

- 2 0

+12 8

tz 5

- 6 8-202+0,9

-21,0-11 0

-16

-124

SOURCE Foster Higgins, “1992 Health Care Benefits Survey Medical Plans, ” Medical Benefits, Mar 30, 1993

that Hennepin and Ramsey Counties had the low- inpatient hospital stays in the Twin Cities asest rates of increase from 1989 to 1994. compared with national rates for 1989 to 1991 are

An indication of trends in health care prices isprovided by the medical consumer price index.Figure 3-1 compares this index for Minneapolis/St. Paul with the United States’ city average.Since 1987, the Twin Cities has consistentlytracked below other cities, declining almost twopercentage points from 1990 to 1991, comparedwith a 0.3 percent decline for the U.S. city aver-age. It is also important to compare Minneapolis/St. Paul’s overall CPI to the United States’ cityaverage to determine if the area’s relatively favor-able performance with respect to the medical in-dex might be related to a favorable trend in theoverall cost of living. Table 3-4 presents thesecomparisons for 1970 to 1992. The rate of in-crease in the overall CPI for the Twin Cities isapproximately the same as the U.S. city averagerate of change.

With respect to inpatient utilization, Minneso-tans experienced 127.4 hospital admissions per1,000 in 1989, 8.5 admissions per 1,000 less thanthe national average, and had 98.2 per 1,000 fewervisits to the emergency room. Minnesotans alsohad fewer outpatient hospital visits (309 per 1,000less than the national average in 1989). Length of

shown in table 3-5. The Twin Cities had lengths ofstays approximately one half day shorter than thenational average for each year reported (49).

I I

- CPI-U National ‘

~ CPI-U Mpls-St. Paul

5 ‘i-J ,1 r I I 1 1

1981 82 83 84 85 86 87 88 89 90 91

SOURCE U S Department of Labor, Bureau of Labor Stallstlcs, Cf’/Detaded Report Dala for January 1994, J Mathery and TJ Moslmann

(eds ) (Washington, DC March 1994)


Part A 1994 Part B 1994 Total 19S9 Pefcent ChangaCounty State Part A 1989 Part B 1989 Total 1994 190994

Los Angeles

San Diego

Broward

Dade

Orange

Rwers{de

San Bernadmo

Mancopa

Cook

Palm Beach

Multnoma

Kmg

Hennepln

Plnellas

Volusla

Bexar

Monroe

Plma

Hlllsborough

Ramsey

. California

Callforma

Florida

Florida

Califorma

California

California

Arizona

Illmols

Florlda

Oregon

Washington

Minnesota

Florida

Florlda

Texas

New York

Arizona

Florida

Minnesota

317.53190.18256.52155.57277.4919940292.01218.87289.83180.96260.98152.80276.98154.41248.56160.57315.33209.91231.51153.68237.00187.22225.24179.63233.02191.98219.64151.77196.32140,65234.99139.91268.68140.81227.24150.20228.80159.68243.44192.61

213.48183.72177.83157.84238.76179.71282.64202.04208.23181,22181.22151.05167.08138.13169.06121.33145.67114.80220.04151.45119,1995.50

138.2599.35

118.3696.79

164.02111.95147.40102.69146.69119.62107,3687.46

158,32110.43164.13125.38114,3595.61

531.01373.90434.3531341516.25379.11574.65420.91498.06362.18444.46303.85444.06292.54417.62281.90461.00324.71451.55305.13356.19286.72363.49278.98351.38288.77383.66263.72343.72243.72381.68259.53376.04228.27385.56260.63392.93285.06357.79288.22

42,02

38.59

36.17

36.53

37.52

46,28

5179

4814

41.97

47.99

24.33

30.29

21.68

4548

41.25

47,07

64.73

47.93

37.84

24.14

SOURCE Health Care Fmancmg Admmlstratlon, Off Ice of the Actuary, Data from the Ofhce of Nahonal Health Statlstlcs, Baltlmore MD

Anderson and colleagues compared the Twin nity hospital characteristics were obtained fromCities, assumed to be a MSA with a competitive the 1972 to 1991 editions of Hospital Statisticsstrategy toward health care, to Baltimore, and the Guide to Zhe Health Care Field, both pub-assumed to be a MSA with a regulatory strategy lished by the American Hospital Associationtoward health care, using measures of hospital (1,2). Anderson and colleagues found that theproductivity, cost per discharge, and hospital uti- annualized rate of increase in hospital costs perlization (3). For this comparison, data on commu- capita in the Twin Cities was 10.0 percent, com-

——— —

Chapter3: Health Care Delivery in the Twin Cities I 11

CPI Avemges 1970 to 1992 0/0 Increase1970 19ao 19a5 19a7 19aa lm 1990 1991 1992 1970-1992

OverallU S city average 3 8 8 8 2 4 107,6 113,6 118.3 124.0 130.7 136.2 1403 262Twm Cdles 3 7 4 7 8 9 107.0 111,6 117,2 122,0 127,0 1304 135,0 261

SOURCE U S Department ot Labor Statistics. Bureau of Latxx Stallst!cs, CPI Deta/led Repor?, (Washington, DC January, 1994) U S Depart-ment of Labor, Bureau of Labor Stahstlcs, CPI Dlwsron, Sumnary Data, (Washington, DC 1994)

pared with 11.2 percent nationwide. Anderson andcolleagues concluded that both strategies had onlya minor effect on controlling hospital expendi-tures percapita from 1971 to 1990(3). They foundregulation to have a greater impact on hospitalproduction processes, primarily by controllingexpenditures per discharge and per day. Competi-tion was found to have a greater impact on utiliza-tion, mainly by lowering the number ofadmissions per capita.

| Insurance Coverage andUninsured PeopleIn 1990, fewer people were uninsured in Min-

nesota than the national average, ranking seventhlowest among states, in part reflecting a relativelygenerous Medicaid program. According to onesource, approximate] y 6.5 percent of Minnesotanswere uninsured for health care services at any giv-en time in 1990. Approximately 4.5 percent wereuninsured for the entire year and 8.6 percent wereuninsured for at least one month in 1990 (42). Thiscompares with approximately 14 percent (34.7million) uninsured nationally in 1990.1

More recent data from the March 1992 CurrentPopulation Survey indicates that approximately10.1 percent were uninsured in 1993 in the Minne-apolis/St. Paul metropolitan area, in comparisonwith 15.4 percent nationwide and 17.6 percentin metropolitan areas with over one millionpersons ( 18).

National 6.6 6.6 6 5Twin Cities 6.2 6 1 5 9

SOURCE Minnesota Department of Health, Health Economics Pro-gram unpublished data Minneapolis MN 1993

People without insurance in Minnesota aremore likely to be male and younger than peoplewho are insured. Uninsured adults in Minnesotaare less likely to have a high school education,more likely to have lower incomes, less likely tobe married, and more likely to be nonwhite thanthose insured by group plans. In 1990, only 28percent of the uninsured people in Minnesota hadincomes that were below the Federal Poverty Line(FPL). However, 71 percent had incomes thatwere below 200 percent of the FPL. Only 3 per-cent of the insured group were below the FPL,with 20 percent at 200 percent below the FPL (42).

| Enrollment in Managed Care PlansIn 1992, in the Twin Cities metropolitan area,

44 percent of the population were enrolled inHMOs. Total enrollment by HMO in Minnesota atthe beginning of 1992 is presented in table 3-6.Medica Choice was the largest HMO, followed byGroup Health and MedCenters. The majority ofenrollees, 82 percent, are in commercial plans,with 12 percent in Medicare and 6 percent in

] Est]mates from different surveys may differ due to the way the question is asked. Therefore this cornpans(m should be Interpreted cau -tl(msl}.

. . . . . . . - ---- -

Parsn~ owner Year 1991HMO Headquarters or manager opened Enrollment

Blue Plus Eagan Blue Cross and Blue 1974 69,884Shield of Minnesota

Group Health Minneapolis Group Health, Inc 1957 294,969

MedCenters St LouIs Park Aetna Health Plans 1973 258,839

Med{caMedica Choice Mlnnetonka Umted HealthCare 1975 352,378

Medlca Primary Mlnnetonka Umted HealthCare 1973 28,637Metropolitan Health Plan Minneapolis Hennepln County Bureau 1983 28,712

of Health

NWNL Health Network St Paul Northwestern National 1984 19,586Life Insurance Co

UCare Minneapolis Umversity of Minnesota, 1989 10,709Department of FamilyPractice

SOURCE Clttzen’s l-eague Research, “Mlnneso\a Managed Care Review 1992,” Mlnneapol!s, MN, August 1992

History/Status

Changed name from HMO Minnesota in 1988.Absorbed Coordinated Health Care HMO in1988 Affiliate Minnesota Health Plans, Inc.,merged into Blue Plus, effective Dec. 31, 1990

Includes Group Care, nonfederal!y qualifled HMO,In 1992, announced intent to merge withMedCenters Health Plan,

Formed by merger of MedCenter Health Plan andNlcollet-Eitel Health Plan in 1983 In 1992,announced intent to merge with Group Health.

Formerly known as Physicians Health Plan (PHP)Combied with Share Health Plan to formMedica l effective Jan. 1, 1991.

Formerly known as Share Health Plan.

Created for Medicaid Demonstration Project andVoluntary AFDC Managed Care Program. ,

Founded as Senior Health Plan Acquired andrenamed by NWNL in 1987.

Created for Medicaid Demonstration Project

Chapter 3: Health Care Delivery in the Twin Cities | 13

IntamalAll primary care Family practice rnadicine Pediatrics OB/GYN

Statewide 90.5 42,3 27.3 11,8 9 0

Nonmetro counties . 5 5 2 41.7 8.3 2.2 2.8

Nonmetro counhes( c 10,000 population) 48.8 43,2 3.4 1,1 11

SOURCE Area Resource Flleasanalyzed byJ Chmtlanson, B DoWd, J Kralewskl, et al , lnstllutetor Health services Research, Schoo[of PublicHealth, Unwersty of Minnesota, Mlnneapolls, MN, 1994

Medicaid. Medicaid enrollment increased by 25percent from 1990 to 1991 in contrast to a de-crease in enrollment in both commercial andMedicare plans during the same period (54).(More detail on trends in HMO enrollment is pro-vided in a later section of this background paper).

I Health Care ProvidersTable 3-7 displays the number of Minnesota

primary care physicians per 100,000 population,by geographic unit and specialty for 1988. Therewere 42.3 family practice physicians per 100,000population, and 27.3 internal medicine specialistsper 100,000. Statewide, Minnesota had 90.5 pri-mary care physicians per 100,000 population,which is close to the Bureau of Health Professionsmanpower requirements of 91.9. It should benoted, however, that Olmsted County (the loca-tion of the Mayo Clinic) has a high concentrationof primary care physicians. Excluding OlmstedCounty from the calculation, the ratio for Minne-sota is 83.1 primary care physicians per 100,000population (52).

Nationally, the rate of growth of physicians hasexceeded that of the population (table 3-8), withsimilar trends in Minnesota. Thus, physician topopulation ratios increased for both metro andnonmetro areas in Minnesota from 1975 to 1988(table 3-9).

Minnesota ranks 10th among the 50 states inactive physician-to-population ratio (66). Fornonmetropolitan counties, Minnesota ranks 37thamong states (64). Nonphysician providers are ac-tive in Minnesota, with approximately 700 nursepractitioners and 159 practicing physician assis-tants cunently practicing in the state (52).

| Availability of Inpatient CareThe Minnesota Department of Health reports

7,480 beds in 1990 in the Twin Cities’ metropoli-tan area, as compared with 10,193 beds in 1971(54). The Department of Health reports that thenumber of beds per 1,000 population droppedfrom 5.1 to 3.0 during this time period and occu-pancy rates fell from an average of 73.6 percent in1971 to 43.6 percent in 1990. Another data sourcereports a slight increase in the percent occupancyof hospital beds in the Twin Cities from 1985 to1991, as calculated using staffed beds and li-censed beds (table 10) (44). From 1982 throughAugust 1993, six Twin Cities’ hospitals contain-ing more than 2,021 beds closed. These data areconsistent with the decline in hospital beds experi-enced nationwide.

DEVELOPMENT OFMINNESOTACAREWithin the past decade, Minnesota has enacted anumber of reforms to improve access to healthcare services and control health care costs. In1987, the Minnesota legislature passed the Chil-drens Health Plan (CHP). This was followed by asecond major reform in 1992, originally called theHealth Right Act and later renamed the 1992 Min-nesota Care Act. The 1992 Act was followed bythe 1993 Minnesota Care Act and the 1994 Minne-sota Care Act.

Implemented in 1988, CHP provided outpa-tient acute care services to non-Medicaid-eligible,low-income pregnant women and children underage six. CHP was expanded in January of 1991 toinclude all low-income children through age 19.


1963 1973 1978 19s3 19a6

Total physicians 276,475 366,279 437,486 519,546 569,160

Physicians per 100,000 population 146 174 196 218 232

Average annual percent increase Inphysicians (from previous yearshown) 2.9 3.6 3.5 3 1

Average annual percent increase mpopulation (from previous yearshown) 11 1.2 1,3 1.0

SOURCE U.S Department of Health and Human Servces, Health Resources and Services Admmlstratlon, Bureau of Health Professions, SeventhRepoflfothe Prmden[ & Congress onV?e SWJS of Hea/th Persormdm the M/led Sra(es, DHHS Pub No HRS-P-OD-90-1 (Rockwlle, MD HF?SA

June 1990)

Then in 1992, the state of Minnesota enacted theHealth Right Act, now called Minnesota Care.This bill built on CHP and provided subsidizedhealth insurance coverage through a programknown as the Minnesota Care Program. A ciga-rette tax increase and a 2 percent provider tax wereused to finance the program and other health re-form-related activities (11 ).

Subsequently, the 1993 and 1994 Minnesota-Care Acts established the goal of achieving uni-versal health coverage of all Minnesotans by July1997 and beginning in July 1997 requires that allMinnesota residents obtain and maintain healthcoverage (53). However, the 1993 and 1994 billsdid not specify a financing mechanism for univer-sal coverage under the Minnesota Care Program(the subsidized insurance program) and universalcoverage is contingent upon the development of afinancing mechanism in the 1995 legislation. TheMinnesota Care Act did establish other reforms inan effort to expand coverage, including voluntarypurchasing pools, a prohibition on underwriting,restrictions on the use of preexisting conditionlimitations, and a requirement that health plancompanies offer plans that are issued on a guaran-teed basis.

The Minnesota legislature has also advanced anumber of reforms to try to slow the growth rateof health care spending. The 1992 Minnesota CareAct created the Minnesota Health Care Commis-sion (MHCC), consisting of 25 members repre-senting labor unions, consumers, providers,employers, health insurers and others, to developa cost-containment strategy for health care reform

to slow the rate of growth in total private and pub-lic health care spending in Minnesota by at least10 percent per year over the next five years(51 ).MHCC delivered a cost-containment plan to thelegislature in early 1993.

Based on the work of the MHCC, the 1993Minnesota Care Act established a comprehensivecost-containment plan. In the plan, a limit on totalhealth care spending was created and the Commis-sioner of Health was charged with enforc ing annu-al limits on the rate of increase in health care costs.The Minnesota Department of Health estimatedthat Minnesota Care would yield a total of $7 bil-lion in savings by 1998. The Department aimed(in consultation with the MHCC) to have detailedlegislation and regulations developed and to beginimplementing the plan by July 1994. However,this target date proved to be overly optimistic. In-terim controls are being used until final regula-tions are established.

T h e M i n n e s o t a e n t o f “ I n t e g r a t e d S e r v i c e N e t -works” (I SNS), to be formed by providers, payers,and/or purchasers of medical care. The intent wasthat ISNs would provide a comprehensive set ofpersonal health care services to a designated popu-lation of individuals, for a prospectively set budg-et. The Minnesota Commissioner of Health canestablish limits for total ISN budget increases, butcompetition among ISNs is also encouraged tocontrol costs (50). Under the 1993 law, the StateHealth Commissioner can approve ISN arrange-ments, exempting participants from state and fed-

— — .—

Chapter3: Health Care Delive~ in the Twin Cities 115

1975 19M Percert t Chanae 197S88

Actwe physiciansMetropolitan counties 195.5 268,4 37.3Nonmetropolltan couratles 64.3 80.4 25.1

Primary care physiciansMetropolitan counties 88.7 108.6 22.4Nonmetropolltan counties 44.3 55.2 2 4 6

SOURCE U S Department of Health and Human Services, Health Resources and Services Admlnistratlon, Bureau of Health Professions, SeveiMhF/epor?[o (he Pres/der?f& Congress on (he Status o/F/ea/th Persome/m fhe Ur?/ted SkNes, DHHS Pub No HRS-P-OD-90-1 (Rockvllle, MD HRSAJune 1990)

era] antitrust liability. A recent study identified 19organizations that are developing ISNs (37).2

In addition, the 1993 legislation establishes aRegulated All-Payer Option (RAPO) for pro-viders delivering health care outside of an ISN.Providers are required to accept reimbursement atthe all-payer level as payment in full for servicesprovided to: ( 1 ) Minnesota residents; (2) personscovered by all-payer insurance; and (3) out-of-network services provided to ISN enrollees.RAPO will provide an alternative to ISNs forthose who prefer to participate in a fee-for-servicesystem and is expected to be fully implemented byJuly 1, 1997.

Minnesota has also enacted a number of otherprograms to improve coverage and reduce healthcare costs. The 1994 Minnesota Care Act estab-lished voluntary purchasing pools to negotiateand purchase health care coverage for employers,groups, and individuals. By July 1, 1997, largepurchasing pools are expected to be available toall purchasers, regardless of employment status orgroup membership. Recommendations will besubmitted by the MHCC to the 1995 legislativesession regarding whether all or some purchasersshould be required to obtain coverage throughpurchasing pools. Recommendations also will bemade regarding the creation of a state-adminis-tered purchasing pool, which would serve all Min-

nesotans who do not have access to otherpurchasing pools (53).

A universal, comprehensive benefit set will bethe standard coverage for all Minnesotans in 1997.The benefit set will be the basis for coverage understate health care programs, with additional wrap-around provisions to meet the special needs ofpopulations served by government programs (53).

The national health care reform proposed bythe Clinton Administration is similar to Minneso-ta Care in several respects. Minnesota Care did notinitially propose to provide universal coverage, asdid the Clinton plan; however, the 1994 Minneso-ta Care legislation does support universal cover-age. Minnesota Care differs from the Clinton planin that it does not have employer mandates for thepurchase of health insurance. A financing mecha-nism for achieving universal coverage under Min -nesota Care is still being developed.

Under the Clinton proposal, regional or corpo-rate insurance purchasing alliances would becreated to “manage competition. ” The regional al-liances would be mandatory for firms with fewerthan 5,000 employees. Alliances would adminis-ter subsidies to eligible individuals, enforce thepremium limits, and have other administrative re-sponsibilities. Purchasing pools formed under theMinnesota legislation are currently voluntary andwill not be involved in the enforcement of cost

z~e ~ge[ &[e of ISN implementation (July 1994) established in the 1993 legislation was postponed. The 1994 MmnewtaCare Act allows

ISNS [o form voluntarily after July 1, 19%, and rules govemlng ISNS will be adopted by Jan. 1, 1997.


1985 4,677.3 62.1 45.8

1986 4,517.8 61.8 45.0

1987 . 4,450.4 61.9 44.9

1988 4,457,1 64.0 47.2

1989 4,463.1 68.2 49.9

1990 4,422,5 66.5 48.2

1991 4,303.7 65.7 47.6

● Table 3-10 shows aggregate occupancy rate and bed capacity figures for Twm Cmes hospitals The number of staffed and licensed beds were

obtained from the Minnesota Department of Health, Survey and Compliance SectIon Only those hospitals for which data was avadable from bothThe Health Care Council and the Minnesota Department of Health appear m occupancy and bed capacity analyses

Occupancy rates were calculated using a combination of Council discharge data and hospital bed statistics obtained from the Minnesota De-partment of Health To determme percent occupancy m a gwen year, the average dally census was first calculated bydwldmg the total number ofInpatient days for Twm Cities hospitals (excludlng newborns and neonates) by 365 days. Average dally census was then dwlded by the totalnumber of either hcensed or staffed beds to determme percent occupancy

SOURCE Metropohtan Health Care Council, Report on Twm Clhes Hospitals from the Counctl of Hospital Corporations Inpatient Utlllzatton DataBase, Mmneac)ols, MN, 1993

containment regulations or in the administrationof subsidies.

Under the Clinton proposal, a National HealthBoard would enforce the health alliance averagepremium targets. A similar function under Minne-sota Care will be provided by the Minnesota Com-missioner of Health, who will be responsible forenforcing the cost controls that apply to prov idersunder the Regulated All-Payer Option (RAPO)and to the ISNs. Both the Clinton plan and Mone-sota Care include the development of standards forquality of care and monitoring of provider com-pliance. Consumer empowerment through moreinformed decision making is also part of both ef-forts.

SUMMARYThere are no published data that would allow a

comparison of health care expenditures in theTwin Cities to other metropolitan areas while con-trolling for differences in benefit. coverage andsize and characteristics of groups by metropolitanareas. Two uncontrolled studies have found that

health care expenditures are lower in the TwinCities, while a third study found the opposite.Some data indicate that health care costs in theTwin Cities may be rising at a slower rate than inthe nation as a whole. The medical price index inthe Twin Cities was above the national averagefrom 1981 until 1987. However, since 1987 it hasbeen below the national average. One study foundthat between 1971 and 1990 the annualized rate ofincrease in hospital costs per capita in the TwinCities was 10.0 percent, compared with 11.2 per-cent nationwide. There are fewer uninsured indi-viduals in the Twin Cities than the nationalaverage.

Minnesota is implementing an ambitioushealth care reform plan designed to improve ac-cess for the uninsured and control health carecosts. Like the Clinton Administration’s healthreform proposal, Minnesota Care relies on com-petition among health care organizations and gov-ernment regulations to control costs. However, itdiffers from the Clinton Administration’s plan inseveral important respects.

Growth ofManaged Careand Integrated

DeliverySystems 4

The health care delivery system in the Twin Cities is bestknown nationally for its reliance on HMOS, and for thehigh proportion of community residents enrolled inHMOs. It has been scrutinized as a community where

“competition” among health plans has occumed, although thereremains debate about the exact nature of that competition and itseffects. This section describes the evolution of the Twin Cities’health care market in three phases. The first phase covers the de-velopment and early growth of HMOs. The second phase spansthe 1980s, when a large number of studies sought to evaluate theimpact of that HMO development on various measures of marketperformance. The third phase focuses on the recent consolidationof the supply side of the Twin Cities’ health care market.

DEVELOPMENT OF THE HMO MARKET:1970-1980

The first health maintenance organization (HMO), GroupHealth, Inc. (GHI), was founded in the Twin Cities metropolitanarea in 1957 (6). This plan, which was managed as a consumercooperative, employed salaried physicians and purchased hospi-tal services by contractual arrangements with community hospi-tals. The strongest early advocates of HMOs in the Twin Citieswere union groups and public sector employees. Most physiciansviewed Group Health as inferior socialized medicine, and privateemployers were generally opposed to offering GHI as a healthplan option (28).

In January 1970, Dr. Paul Ellwood, a health care reformer inthe Twin Cities, coined the term HMO in a Fortune article dealingwith prepaid medical care. Ellwood advocated the development

I 17of a large number of HMOs nationwide to compete for patients


with each other and traditional insurers. The hopewas that the internal incentives associated withprepayment, together with competitive pressuresto contain premiums, would result in a more effi-cient health care delivery system and lower ratesof increase in health care expenditures. The ideasof Ellwood and his colleagues appealed to theNixon Administration to the extent that PresidentNixon, in his 1971 address to Congress, promotedHMOs as a national strategy to contain health carecosts.

In 1972, a highly respected multispecialtygroup practice in the Twin Cities, the St. LouisPark Medical Center (now Park-Nicollet MedicalCenter), created a prepaid alternative called Med-Centers Health Plan, thereby improving the imageof HMOs in the Twin Cities (28). St. Louis ParkMedical Center had begun to lose patients toGroup Health, and large employers in the commu-nity showed interest in offering a competingHMO. The launching of the HMO initiative by theNixon Administration provided further impetusfor the formation of MedCenters. MedCenterswas a group HMO, allowing physicians toprovide care to patients not enrolled in the HMO.In this respect. it differed from Group Health, Inc.,where physicians were salaried and treated onlyGroup Health enrollees.

In the increasingly competitive environment ofthe Twin Cities, physicians outside the HMO sys-tem sought to offer alternatives to fee-for-serviceand the existing staff and group model HMOs (5).In 1975, Physicians Health Plan (PHP) wasformed as an independent practice association(I PA ) model HMO. Independent physicians couldbe associated with PHP while maintaining theirfee-for-service practices. PHP was a much looserHMO model than MedCenters or Group Health,in that physicians in PHP were not salaried, andPHP enrollees had considerably more freedom tochoose their physicians, with no physician gate-keepers to determine if visits were necessary. En-rollees, therefore, had access to a system wherethey could use their own doctor and favoritehospital. with the added benefit of an improvedpayment mechanism. Without intense price com-

petition from indemnity plans, the HMOs couldbe generous in the benefits they offered (57).

From 1971 to 1978, HMO enrollment in theTwin Cities grew at an average annual rate of 27percent. By December 31, 1978, there were240,800 individuals (12.4 percent of the standardmetropolitan statistical area) enrolled in sevenHMOs, compared with 5 percent enrollment inHMOs nationally at that time (10). In 1980, theTwin Cities had an enrollment in HMOs per capitathat was three times larger than in the rest of thecountry. Table 4-1 shows HMO enrollmentgrowth in the Twin Cities from 1970 to 1981. Dur-ing the 1980s, HMO enrollment continued togrow, reaching almost 50 percent of the Twin Ci-ties’ population by the end of the decade (58). Thiswas attributed primarily to PHP entering theHMO market. Within six years of entering themarket, PHP’s enrollment grew to 95,141, maki-ng it almost equal in size to MedCenters and halfas large as Group Health. It has been speculatedthat this rapid growth was due largey to PHP pro-moting its policy of consumers being able tochoose their own providers (28).

Why did HMO development proceed more rap-idly in the Twin Cities during the 1970s than inother cities? Anderson and colleagues argued thatthree factors supported the development andgrowth of HMOs in the Twin Cities (6). The firstwas the “pre-existing environment.” Andersonand colleagues concluded that the social homo-geneity, political progressiveness, and economicstability in Minnesota were primary causes of theaccelerated development of HMOs (6). Thesecharacteristics were manifested in the large num-ber of multispecialty group practices existing inMinnesota (which facilitated the formation ofHMOs), employers with a track record of success-ful community leadership, and a communityproud of the quality and accessibility of its healthcare and concerned mainly with the cost of healthcare.

The second factor described by Anderson andcolleagues involved the “initiatives” taken by em-ployers (6). Because the primary concern in theTwin Cities regarding health care was cost, the

.

Enrollment 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 g

Group HealthPlan 35,996 42,879 52,230

CoordinatedHealth Plan 1,715

MedCenter 1,000Nlcollet-Eitel

Health PlanSHAREHMO MinnesotaPhysicians

Health PlanTOTAL 35,996 42,879 54,945% Growth 19 28

Metropolitanpopulation 1,874,440 1,883,100 1,891,600

% Metropolitanpopulation 1 9 2 3 2 9— —

a Seven-county metropolitan areab Includes Medicare Dernonstrahon prolect Enrollment

59,172 66,638 76,883 91,372

1,945 2,184 2,941 3,5784,233 7,049 10,090 17,591

441 1,853 2,370 3,1792,846 3,299 9,189 12,130

1,725 2,914 3,368

53 9,708

68,637 82,748 104,440 140,929

25 21 26 35

1,899,200 1,914,900 1,912,500 1,924,100

3 6 4 3 5 5 7 3

107,517

3,98531,797

5,49117,1216,400

14,227186,538

32

1,931,500

9 7

121,184

4,02546,706

8,48521,86212,170

26,422240,854

29

1,945,600

124

130,810

4,45961,278

14,95727,44926,195

45,240

310,38829

1,959,800

158—

153,869

4,92270,616

20,98433,898b48,309

85,173417,771 b

35

1,985,700

21.0

181,328

5,24390,282b

27,373 b

37,486 b

49,511 b

95,1 41b486,364 b

16

1,989,600

244 ‘

SOURCE O Anderson, T HeroId, B Butler, et al, HMO Development PaUerns and Prospechves (Chicago, IL Pkmbus Press, 1985)

A

co

20 | Managed Care and

payers, and particularly

Competitive Health Care Markets: The Twin Cities Experience

large corporate employ-ers, supported the development of alternative pay-ment mechanisms for health care. Other areas ofthe country were more concerned with access andquality; therefore, efforts in these other communi-ties were led by consumers and providers, and didnot focus on alternative payments.

The third factor identified by Anderson andcolleagues concerned the “responses” in the com-munity (6). Anderson and colleagues found excel-lent communication and responsiveness withinthe Twin Cities, in contrast to other communitiesthey studied. In the Twin Cities, when St. LouisPark Medical Center formed an HMO, the entirecommunity was aware of and interested in itsprogress. Other communities were either wary ofthe development of HMOs or viewed them as rela-tively unimportant experiments. When mandatedto include federally qualified HMOs in their bene-fit options, employers in the Twin Cities decidedto offer a selection of those not federally approvedas well. This greatly aided in the distribution andgrowth of HMOs in the Twin Cities.

In summary, the decade of the 1970s was a timeof rapid growth of HMO enrollment in the TwinCities. This appeared to have occurred in part as aresponse to the rising cost of health care, sup-ported by strong interest on the part of Twin Ci-ties’ employers, and the general progressivenature of the Twin Cities’ political climate. TheHMOs spanned a variety of ‘*models,” with mostTwin Cities’ physicians affiliated with one ormore HMOs by the early 1980s (6).

COMPETITION AMONG HMOS ANDITS EFFECTS: 1980-1990

The seven Twin Cities HMOs that began the1980s had two opportunities for growth. Theycould gain new enrollees from the fee-for-servicesector, or they could capture business from eachother. The HMOs’ emphasis on the former strate-gy led to accusations that they “shadow priced”the fee-for-service sector and to a perceptionamong employers that competition among HMOshad failed. The first criticism appears to havesome validity, but the extent to which competition

failed, succeeded, or indeed was ever really triedduring this period is a much more complicated is-sue.

In 1980, about 20 percent of Twin Cities’ resi-dents were enrolled in HMOs. Clearly, HMOs hadan opportunity to grow rapidly by underpricingcompeting fee-for-service insurance plans. Thereare several reasons why the HMOs may not havepursued this strategy more aggressively. The firstand perhaps most important reason relates to theconditions under which employers offered theseoption to their employees. Paul Ellwood, an earlyproponent of HMOs in the Twin Cities, in a 1984interview with John Iglehart, in the New EnglandJournal ofitledicine(31 ) said:

His ‘biggest disappointment’ about health caredevelopments in the Twin Cities is the failure ofcorporations to take advantage of their purchas-ing power in the market. Major national corpo-rations based here (in the Twin Cities) . . . havebeen unwilling to go out and buy care on the ba-sis of price.

In the same article, Walter McClure, another TwinCities’ health policy analyst, noted that:

Employers and unions have been willing to offerworkers health care coverage through the high-cost, traditional insurance plan, which almosttotally lacks incentives to make the consumerprice-sensitive, and then make that sameamount available to HMOs. HMOs have beendelighted to pick up that money.

In the early 1980s, few employers that offeredmore than one health plan set a “defined,” or fixed,contribution at or below the premium of thelowest-cost plan, which would have required em-ployees to pay the additional cost of more expen-sive plans out of their own pockets. Feldman andcolleagues examined 44 Twin Cities’ firms offer-ing multiple health - plans and found that fewerthan half had adopted a level-dollar contributionmethod for the family coverage premium and onlyabout one-third paid a level dollar contribution to-ward the single coverage premium (27). Othercontribution formulae, such as a level percentagecontribution, or explicit subsidy of the traditionalfee-for-service plan, mitigated the incentives of

Chapter 4: Growth of Managed Care and Integrated Delivery Systems | 21

HMOs to reduce their premiums, since a $1.00 de-crease in an HMO’s premium would not necessar-ily increase the premium differential between theHMO and its competitors by $1.00.

Also during the 1980s, some policy analysts ar-gued that high-risk. fee-for-service enrollees weremore likely to have a long-standing relationshipwith their fee-for-service physicians and thereforewere less likely to join a staff or group modelHMO. The ● ’favorable selection” of relativelyhealthy employees into group and staff modelHMOs in the Twin Cities was documented byJackson-Beeck and Kleinman and Dowd andFeldman (14,32).

Feldman and Dowd modeled HMO enrollmentgrowth, assuming that HMOs experienced initialfavorable risk selection that decreased over time(22). They further assumed that, in a two-plan em-ployee benefits offering consisting of one HMOand one fee-for-service plan, the fee-for-serviceplan would experience-rate, charging premiumsthat equaled the average cost of care for its enroll-ees, plus an administrative fee, while the HMOwould be free to set its premium, subject to theconstraint of employee demand. Under these as-sumptions, Feldman and Dowd showed that theHMO could maximize its profits by setting pre-miums at levels that would capture only a portionof the fee-for-service sector’s enrollees, ratherthan driving the fee-for-service plan from the mar-ket. In order to capture enrollees from otherHMOs, the HMO might have to set premiums solow that fee-for-service plans would be drivenfrom the market (22). From the HMO’s point ofview, this would not be a profit-maximizing strat-egy in the longer run.

In addition to experiencing favorable selection,HMOs appeared to enjoy a “technological advan-tage” over the fee-for-service sector. HMOs wereable to produce “output” (i.e., treatment of theirenrollees) using fewer or lower cost “inputs” (e.g.,hospital days and physician visits) than the fee-for-service sector. If HMOs had competed fiercelyamong themselves on the basis of price, the pre-mium for HMO enrollees should have drivendown the cost of producing treatment using theHMOs improved “technology.” As noted above,

however, excessively low premiums might havereduced fee-for-service market share below theprofit-maximizing market share.

This HMO pricing strategy. coupled with a rel-ative lack of employer information on the healthstatus of their employees, resulted in disappoint-ing effects of HMOs on employer health insurancecosts. Employers who offered their employees achoice of HMOs and the fee-for-service sectorsometimes saw their total health insurance costsincrease as the relatively healthy employees leftthe experience-rated, self-insured, fee-for-sexviceplan to join HMOs (26).

If employers had known the health expendi-tures of their HMO enrollees, they might havebeen able to prevent some of the losses associatedwith this selection process. Unfortunately, how-ever, early attempts by employers to obtain in-formation on the actual health expenditures oftheir employees who were enrolled in HMOs weregenerally not successful. Because their premiumswere community-rated, HMOs were able to tellemployers that they kept no data on the experienceof employees by firm. The ability of employers tothreaten HMOs with expulsion from their benefitplans was limited because the national HMO Actof 1973 required employers to offer at least onefederally qualified HMO of each “type” availablein a market area, if “mandated” by a federallyqualified HMO. Throughout the 1980s. employerpressures for experience-rated products and moredata on utilization of services by HMO enrolleesincreased, and HMOs began to offer products thatwere not federally qualified in order to meet thesedemands.

When employers offered a choice among healthplans, employees were quite sensitive to out-of-pocket premium differentials. Feldman and col-leagues studied the choice of health plans byemployees in 17 large Twin Cities’ firms in 1984(28). Unlike previous studies, the authors wereable to identify precisely the health plan choice set(i.e., single versus family coverage) for each indi-vidual, and incorporate information on the avail-ability of coverage through the spouse. Theelasticities estimated by Feldman and colleaguesare considerably higher than those found in pre-


vious studies, as large as -8.6 for choice amongsingle coverage plans. This means that a one per-cent increase in the out-of-pocket premium differ-ential between two plans reduces the enrollmentshare of the higher cost plan by 8.6 percent.

Although the HMOs’ incentive to cut prices toconsumers was limited by employer premiumcontribution methods and (for some HMOs) fa-vorable selection, the HMOs’ incentive to reducetheir costs was not so impaired. Cost-cutting ef-forts on the part of health plans precipitated a sig-nificant reorganization of the Twin Cities’ healthcare market during the latter 1980s from a rela-tively close, collaborative relationship betweenplans and providers to a distinct division betweenfinancing and service delivery functions. Thechange was often slow and subtle, but sometimesit was abrupt, contentious, and played out on thefront pages of the local press. Blue Cross and BlueShield of Minnesota evolved from a traditional in-surance plan to an aggressively managed healthplan engaged in outcomes research and the devel-opment of preferred provider networks in both ur-ban and rural areas (see discussion below).Physicians Health Plan, started by physicians forphysicians, experienced a bitter dispute betweenphysicians and the health plan’s management overfees and administrative practices (38). The samefate befell MedCenters Health Plan in its relation-ship with the Park-Nicollet Clinic.

The changes in the relationship between healthplans and providers that occurred in the latter partof the 1980s were driven largely by consumer de-mand. In the Twin Cities’ health plan market, con-sumers had grown accustomed to being offered achoice of health plans and recognized that not allhealth plans offered access to all providers. Be-cause the majority of consumers are in goodhealth, the choice of health plan, based on factorssuch as coverage, clinic locations, and out-of-pocket premiums, tended to override consumerloyalty to specific health care providers. Sincepremiums were an important determinant ofhealth plan choice, even in the face of employerpremium contribution policies that reduced out-of-pocket price d inferences, consumer willingness

to change health plans gradually produced pres-sure on plans to restrain premium increases (28).That pressure eventually was transmitted to pro-viders in negotiations over contracts.

I Empirical Studies

Hospital Finances and Demandfor Hospita/ Services

During the 1980s there were several attempts toevaluate the competitiveness of the Twin Cities’hospital market and the changing relationship be-tween health plans and hospitals. In a case study ofthe Twin Cities, using data primarily drawn fromthe 1970s, Luft and colleagues found no convinc-ing evidence that the growth of HMOs had af-fected hospital use (40). Feldman and Dowdestimated the price elastic it y of demand for hospi-tal services from 1981 data on 31 Twin Cities hos-pitals (23). They found that price sensitivity at thattime was either totally lacking, as in the case ofMedicare patients, or fell far short of the competi-tive ideal. In another study, Feldman and col-leagues examined the effect of HMO discounts onhospital revenue, cost, and profits (25). Thisstudy, based on Twin Cities hospital data from1979 to 1981, found that neither HMO discounts,nor a larger share of HMO, Medicare, or Medicaidpatients, were associated with lower hospitalcosts. Furthermore, neither HMO market sharenor HMO discounts adversely affected hospitalprofits. The authors concluded that, if competitionamong health plans was to reduce hospital costs orprofits, it would have to encompass more than justgrowth of HMO market share. Kralewski and col-leagues also found that HMOs were not usingcompetitive bidding in their contractual relation-ships with hospitals during the period 1977 to1980 (36).

By 1986, however, the pattern of HMO-hospi-tal relationships had begun to change. In a study ofsix HMOs in four large metropolitan areas (one ofwhich was the Twin Cities), Feldman and col-leagues found that HMOs, especially staff andnetwork HMOs, were beginning to concentratetheir patients at hospitals and that price played an

—.——


19a28282 Iw 1982 Length 1988 LengthSefvice group Diachar’ges Discharges of stay (days) of stay (days)

Oncology 10,856 8,835 8.82 5,92

Cardiology 26,741 28,302 8.22 5,38

Psychiatry 13,328 13,924 17.84 1195

Chemical dependency 7,906 7,422 1545 11,31

Ophthalmology 7,528 1,742 3.10 2.24

ENT 10,830 6,938 2,51 1.97

Neurology 15,188 11,751 8.69 6 0 4

Orthopedics 31,769 21,833 7 8 5 5.38

Urology 12,650 9,884 5.85 4 0 7

Gynecology 11,036 7,525 5.22 4.05

Obstetrics 39,002 41,635 3.52 2 6 9

Newborns 35,438 37,482 4,23 3 2 4

General medlcme 78,369 66,542 7,02 5 6 3Total 300,641 263,727 6.91 6 1 3

SOURCE Council of Hospital Corporahons (renamed the Metropolitan Health Care Council), Trends m Twm C/(/es HospIm/UMIzaOon 1982-1988St Paul MN, 1989

important part, not so much in the HMO’s choiceto affiliate with a particular hospital, but in thevolume of services demanded from the hospital(21 ). The estimated price elasticit y of demand foradmissions in HMO-affiliated hospitals was -3.0,indicating a considerable degree of price sensitiv-ity (a 3 percent reduction in admissions associatedwith a 1 percent increase in price). Independentpractice association HMOs were not found to ex-hibit the same degree of price sensitivity. The esti-mated price elasticity of demand for independentpractice associations was -1.0, similar to the elas-ticity estimate in Feldman and Dowd’s studybased on 1981 data (22).

Inpatient Resource UseDuring the 1980s, Twin Cities’ hospitals faced

declining discharges and lengths of stay acrossvirtually all types of services that were tracked bythe Metropolitan Health Care Council, which isthe Twin Cities’ hospital trade association (table4-2). Even among the service groups experiencingsome increase in discharges (i.e., cardiology, psy-chiatry, obstetrics, and newborns), lengths of stayfell precipitously. Part of the declining use of inpa-tient resources mirrored a trend in national data.

Dowd estimated the proportion of reduced admis-sions from 1977 to 1982 that could be attributed toHMOs in the Twin Cities’ market (13). That esti-mate depends crucially on the amount of creditthat HMOs receive for reducing length of stay inthe non-HMO sector. If HMOs are given credit fornone of this “spillover” effect, then the reducedadmissions among HMO enrollees, plus thegrowth in HMO market share, would imply thatHMOs were responsible for one-third of the de-crease in admissions over that time period. IfHMOs are given credit for the entire decline indischarges in the non-HMO sector, then HMOscould be responsible for 85 percent of the totaldrop in admissions.

The HMO effect on length of inpatient hospitalstays provides an interesting example of the re-finement of resource management techniquesused by health plans. In early studies of HMOs inthe Twin Cities, HMO membership wasassociated with a 40 percent reduction in hospitaladmissions but no reduction in length of stay (seeDowd, et al., 1986 (16) and Johnson, et al., 1989(33) for reviews of the literature). However, by themid- 1980s, Dowd and colleagues found that en-rollees in group practice HMOs in the Twin Cities


had significantly shorter lengths of stay than in-demnity-insured patients in five of seven diagnos-tic groups examined, while enrollees in IPAHMOs had significantly shorter lengths of stay inthree of these groups ( 16). Johnson and colleaguesalso examined data from Twin Cities hospitalsover the three-year period 1982 to 1984 and foundthat lengths of stay for group practice HMO en-rollees were significantly shorter than stays for ei-ther indemnity-insured patients or IPA enrollees(33).

Why were Twin Cities’ HMOs, and particular-ly group and staff model plans, able to reducelength of stay, relative to indemnity insurers,when HMOs in other study sites were not? Allhealth plans should want to minimize their costs,whether those savings are passed on to consumersor not, but some cost-saving techniques may havehigher payoffs and be less costly to implementthan others. Dowd and colleagues suggested thatreductions in admissions were easier for healthplans to achieve than reductions in length of stay,since reduced admissions can occur simply byswitching treatment to the outpatient setting (16).Length of stay reductions, however, involved di-rect intervention in the physician’s onsite treat-ment decisions. Thus, the initial focus of HMOson reducing admission rates is not surprising.Once HMOs had reduced admissions rates, how-ever, the competitive advantage to be gained byreducing length of stay made that task worth pur-suing, although more difficult. The maturity of theHMOs in the Twin Cities’ market also may havehad some effect, but competition from otherHMOs does appear to have been an important fac-tor. In three of the studies that preceded Johnsonand colleagues’ study, the HMOs that had notachieved reductions in length-of-stay relative tothe fee-for-service sector had no close HMO rivalsin their market areas (33).

Access and Health OutcomesThree studies used data from the mid- 1980s to

compare the health outcomes of subpopulations ofMedicare or Medicaid beneficiaries in the Twin

Cities enrolled in HMOs with beneficiaries re-

ceiving care from providers under normal pro-gram managements. One examined the differencein physical functioning and perceived generalhealth status between Medicare beneficiaries en-rolled in HMOs and in traditional Medicare (69).A second assessed the effects of HMO enrollmenton the health and functional status of “*dual eligi-ble” Medicaid/Medicare beneficiaries, while athird assessed the effect of HMO enrollment onhealth, functional status, and service utilizationamong severely mentally ill Medicaid beneficia-ries (9,41 ,43,55).

The study of Medicare beneficiaries found nosignificant difference in predicted health status asmeasured by physical functioning between thoseenrolled in HMOs and traditional Medicare (69).However, there was a difference between the twogroups in predicted health status, as measured byperceived general health status, with those en-rolled in HMOs having a significantly higher levelof perceived health status. For a subgroup of lowerincome enrollees, no significant differences werefound in predicted health status. This does notsupport Ware and colleagues’ 1986 findings thatlow-income individuals have worse outcomes inHMOs, as compared with fee-for-service care(67).

Lurie and colleagues examined the effect onhealth and functional status measures of enrollingnoninstitutionalized elderly Medicaid recipientsin prepaid plans as compared with traditional fee-for-service Medicaid (41 ). Beneficiaries were ran-domly assigned to a group receiving prepaid carefrom one of seven health plans, with only theMedicaid proportion of their care being capitated.A sample of beneficiaries (400 in prepaid care and400 in traditional Medicaid) were interviewed atbaseline and one year later. Major outcome meas-ures in the study included general health status,activities of daily living, instrumental activities ofdaily living, corrected visual acuity, and bloodpressure and glycosylated hemoglobin for hyper-tensive and diabetic persons, respectively. Theanalysis found no significant difference betweenthe two groups in number of deaths or any of the1isted outcome measures, thus providing no evi-


dence, in the short-term, of harmful effects of en-rolling elderly Medicaid patients in Twin Cities’HMOs.

Lurie and colleagues also studied the effect ofHMO enrollment on chronically mentally illMedicaid recipients (43). Of 739 clients identifiedas chronically mentally ill, half were chosen atrandom to remain in traditional Medicaid, andhalf were permitted to choose among four capi-tated health plans. The beneficiaries were fol-lowed for an average of 11 months. Outcomemeasures consisted of general health status,physical functioning, social functioning and psy-chiatric symptoms. No significant differenceswere found in general health or mental health be-tween beneficiaries in traditional Medicaid versusHMOs. However, among the subgroup of subjectswith schizophrenia, scores on the Global Assess-ment Scale, a measure of community function,were 7.6 points lower for the HMO group than thetraditional Medicaid. The authors concluded thatthere was “no consistent evidence of short-termadverse health effects” among HMO enrollees rel-ative to traditional Medicaid enrollees.

Access to services and utilization of services bythe same group of chronically mentally ill Medic-aid recipients also was analyzed (9,55). Therewere slight improvements in the majority of ac-cess measures studied for HMO enrollees, al-though they were not statistically significant.Thus, enrollment in HMOs did not reduce accessto physical or mental health care for this group.There also were no significant decreases in the useof inpatient or outpatient services for the HMO en-rollees (55). In particular, there was no statistical-ly significant evidence that Medicaid enrolleeswith severe mental illness used community-basedtreatment programs differently than beneficiariesin fee-for-service Medicaid. However, there wasevidence that HMOs reimbursed these programsat a lower percentage of their charges (9).

| SummaryIn summary, the 1980s saw important changes

in Twin Cities’ health plans and their relationshipswith providers. These changes included the insti-

tution of more aggressive management strategiesby health plans. The aggressive management ofprovider relations was made possible by the grow-ing willingness of consumers to choose healthplans based on characteristics such as requiredout-of-pocket premium contributions, with aweakening of loyalty to specific providers. Thepressure on premiums experienced by healthplans caused some plans to be more sensitive tothe prices they paid for hospital care, leading togreater price shopping in the hospital market.

The empirical evidence also suggests that hos-pital lengths-of-stay were lower in HMOs, rela-tive to traditional insurers in the 1980s. However.there have been no studies measuring the direct ef-fect of the recent growth of managed care and inte-grated delivery systems on the growth rate ofhealth care expenditures.

There is very limited information regarding theeffect of increased HMO enrollment and competi-tion among providers for HMO contracts on ac-cess to services and the health status of TwinCities residents. The available evidence applies tosubgroups of the population, and not to HMO en-rollees from private employed groups. Thesestudies do not find significant differences inhealth status and access measures for HMO versusnon-HMO enrollees.

THE CONSOLIDATION OF THE TWINCITIES HEALTH CARE MARKET

Three recent mergers involving Twin Cities’HMOs have captured national attention. The firstwas a merger of two large HMOs, Group Health,Inc. and MedCenters. According to one policyanalyst, this merger is unique in that “we’ve neverhad a merger...in the national HMO market be-tween two equal partners of this size” serving thesame community (35). .The second major con soli-dation involved the merger of an HMO and a hos-pital, and the third was a merger between an HMOand a hospital system, creating the first verticallyintegrated health care organization in the Twin Ci-ties. In this section we begin by providing an his-torical context for understanding the importanceof these mergers and the public policy issues they


Hospitat Admissions HMO (~0) M e d i c a r e (’%0) Medicaid (?/0) Cxher (’?/0)

Abbott-Northwestern 30,504 15.0 34.0 8.0 43,0

Farvlew-Southdale 18,927 46.3 28.4 2.8 22.5

t-iennepm County 19,031 10.1 24.8 43.0 22.1

Mercy 11,555 51.8 15.8 6.1 26,3

Methodist ● 20,012 0.0 21.6 2.5 76.0

Metropolitan-Mt. Sinai 6,200 32.0 44.0 9.0 15.0Mlnneapohs Children’s 5,972 35.2 0.0 22.6 42.2North Memorial 22,367 44.3 20.0 8 1 27,7Rwerslde 23,855 67.6 17.1 8.6 6 7St. Joseph’s 13,208 44,0 25.2 134 174

St. Luke’s 8,505 8.6 33.0 16.7 41,7St Paul-Ramsey 13,989 0.0 28.6 34.4 371United 18,900 38.4 32.1 10.1 194Unty 10,944 54.3 15,1 6.8 23.8Unwerslty of Minnesota 7,848 10,7 26.3 11.1 51.9

● Melhodlst (and possibly other hospitals) groups HMO admissions with other payers

SOURCE Clt~zens League Research, “Minnesota Managed Care Review 1992, ” Mlnneapolls, MN, August 1992

raise. We do this by documenting merger activityover time in the Twin Cities involving hospitalsand health plans. We then describe each of thethree recent mergers noted above, focusing on themotivations for the mergers and the expectationsof the merger parties. We conclude by discussingseveral ongoing developments in the reconfigura-tion of the Twin Cities’ health care delivery sys-tem.

| Hospital ConsolidationIn 1976 there were 35 hospitals in the Twin Ci-

ties with approximately 10,000 acute care beds atan average of 70 percent occupancy (8). By 1992,as described in the previous section, the number ofacute care hospitals had declined dramatically, ashad the total number of beds and hospital occu-pancy rates. Also, by 1992 almost all hospitals inthe Twin Cities were owned by one of four multi-hospital systems: Fairview, HealthOne, Health-East, and LifeSpan. The hospitals that wereindependent of these systems at that time in-cluded: University Hospital, St. Paul RamseyHospital, Methodist Hospital, North MemorialHospital, Hennepin County Medical Center, andchildren’s hospitals in St. Paul and Minneapolis.

Table 4-3 contains data on admissions in majorTwin Cities’ hospitals in 1991 by payer.

The four major, multihospital systems wereformed in the 1980s through a series of mergersand acquisitions. In 1986, five different hospitalsin St. Paul came together to form HealthEast (59).In 1987, the Fairview system, which existed priorto the 1980s, added St. Mary’s Hospital through apartnership with the Carondelett Catholic order.In 1987, two existing multihospital systemsHealthOne and HealthCentral merged to form anexpanded HealthOne Corporation that includedhospitals in the northern suburbs of the TwinCities as well as facilities in both downtown St.Paul and Minneapolis. The downtown Minneapo-lis facilities were subsequently reduced in scaleand sold to Hennepin County to augment Henne-pin County Medical Center’s capacity. LifeSpan,a four-hospital urban/rural system, was repre-sented in the Twin Cities’ metropolitan hospitalmarket area primarily by its flagship, Abbott-Northwestern Hospital, a tertiary care facility lo-cated near downtown Minneapolis.

Several different but interrelated motivationsfor the “horizontal mergers” that occurred in theTwin Cities’ hospital market have been offered by


hospital and HMO executives. The early 1980swas a period of creation and expansion of multi-hospital systems nationwide, and the aggregationof hospitals in the Twin Cities could be viewed aspart of this general trend. The perceived advan-tages of multihospital systems included improvedaccess to capital, the potential sharing of manage-ment expertise, and cost savings from the con soli-dation of certain administrative functions and theaggregation of purchasing power. It was also be-lieved that the downsizing of individual hospitalsor the conversion of facilities to other missions(e.g., psychiatric care) could be more readily ac-complished under the umbrella of a multihospitalorganization. All of these motivations have beenidentified by Twin Cities hospital administratorsas important factors in the hospital mergers thatoccurred during the 1980s in their community. Inaddition, a motivation more closely tied to the de-velopment of the HMO market in the Twin Citieswas identified by some hospital administrators.

During the mid-1 980s, HMOs TwinCities were able to take advantage of substantialovercapacity in the Twin Cities hospital market tonegotiate relatively low prices for hospital care fortheir members. As enrollment grew in some plans,so did the potential for these plans to shift a sub-stantial number of admissions from one hospitalto another through renegotiation of hospital con-tracts. Anticipating further HMO enrollmentgrowth in the future, hospitals pursued the devel-opment of multi-hospital organizations as a meansof negotiating more effectively with HMOs overprices and to position themselves to offer broadergeographic coverage for HMO enrollees. The hos-pital organizations hoped that by offering broadgeographic coverage they could secure long-termexclusive contracts with HMOs that would gener-ate more predictable streams of patients and reve-nues for their facilities.

The consolidation of the hospital market in theTwin Cities continued in the 1990s when, in 1992,HealthOne and LifeSpan merged to form Health-Span. This was the first merger that generatedpublic debate over whether the consolidation ofthe hospital market in the Twin Cities had gonetoo far (35). The Minnesota State Attorney Gener-

al’s office brought suit in federal court, chargingthat the HealthOne/LifeSpan merger violated fed-eral antitrust laws. It argued that because themerged organization (HealthSpan) would control28 percent of the Twin Cities hospital market. itcould exercise undue market power in negoti-ations with payers. The state ultimately negotiatedan out-of-court settlement that required Health-Span to freeze its revenues for 1993 and documentsubsequent revenue reductions.

| HMO ConsolidationApproximately the same number of HMOs ex-

isted in the Twin Cities in 1991 as in 1977. In1977, five plans contained the great majority ofHMO enrollees: Group Health, MedCenters,PHP, Share, and HMO-Minnesota (Blue Cross/Blue Shield). Group Health was the dominantplan in terms of market share. In 1991, Medica hadthe largest HMO market share, followed by GroupHealth and Medcenters. The other HMOsTwin Cities had relatively small enrollments.

In the early 1980s, some reorganization tookplace in the HMO market, but most of that reorga-nization did not have a substantial impact on over-all market structure. In 1983. the St. Louis ParkMedical Clinic acquired the Nicollet MedicalClinic, precipitating the incorporation of Nicollet-Eitel Health Plan into MedCenters. HMO Minne-sota, the Blue Cross/Blue Shield HMO, changedits name to Blue Plus in 1988 and absorbed Coor-dinated Health Care HMO in the same year. TwoHMOs were created in the 1980s as the result ofMedicaid demonstration projects: MetropolitanHealth Plan sponsored by Hennepin County andUCare sponsored by the University of Minnesota.Both, however, have attracted relatively limitednumbers of enrollees. An HMO managed byNorthwestern National Life Insurance Companyalso entered the market in the 1980s but had fewerthan 20,000 enrollees by 1991.

One of the most important consolidations in theHMO market occurred when SHARE HealthPlan, a group model HMO, merged with Physi-cian Health Plan, a physician-sponsored I PA mod-el plan. The resulting entity, renamed Medica, had


480,000 enrollees in 1991, making it the largestHMO in the community. Medica continued to of-fer both the group and IPA model plans separatelyto employers, renaming them Medica Primary andMedica Choice, respectively. The ability to offer asingle employer both a group and an IPA modelmanaged by the same entity was one of the majormotivations for the merger. During the 1980s,many employers felt that biased selection intosome health plans made it difficult for them torealize cost savings from offering HMOs to theiremployees. They believed they could avoid prob-lems associated with biased selection by contract-ing with a single firm that was able to offermultiple plan options (see discussion of buyer co-alitions below). The merger allowed Medica to bersponsive to these employer demands and there-by strengthened its competitive position relativeto other HMOs in the market.

While the SHARE/Physician Health Planmerger represented a major consolidation in theHMO market, prior to this merger the most signif-icant market developments involved product di-versification on the part of the HMOs and theemergence of Preferred Provider Organizations asclose competitors to HMOs. Product diversifica-tion was pursued by the HMOs primarily throughthe development of “open-ended” (also calledpoint-of-service) options to the traditional closed-panel HMO product and through the sponsorshipof Preferred Provider Organizations. The "bopen-ended” product allowed enrollees to seek carefrom providers that were not part of the HMO net-work but, if they chose to do so, they were re-quired to pay for a greater portion of their care‘“out -of-pocket .“ Premiums for these plans typi-cally were set somewhat higher than for the stan-dard HMO product, and some types of serviceswere excluded from coverage. PHP was particu-larly agggrssive in marketing its open-ended planand by 1992 it had enrolled 228,000 members (al-most two-thirds of all enrollees) in this option(figure 4-1 ).

Some HMOs also established Preferred Pro-vider Organizations, in collaboration with insur-

MedicaChoice

GroupHealth

Blue Plus

MedCenters

NWNL

Metropolitan

L I

59,05659,822

45,34325,568

D 6/894,802 _ 1/923,403

0 50,000 100,000 150,000200,000250,000

SOURCE Interstudy, The Inyrtdyufy Edge, Excelsion, MN, 1990, 1992

ers, so that they could offer a broader range ofinsurance alternatives to employers. The providernetworks for these products general] y were broad-er than the networks offered as part of the basicHMO product. Under a PPO, enrollees typicallyreceive more comprehensive benefit coveragewith lower out-of-pocket costs, if they obtain carefrom the Preferred Provider Network rather thanfrom the general provider community. The PPOsdeveloped by HMOs offered broader provider net-works to self-insured firms. Over the past fewyears, enrollment in self-insured plans in the TwinCities has steadily increased, reaching almost900,000 at the end of 1993 ( 12). Their PPO net-works permitted HMOs to serve this market.

In addition to HMOs, sponsors of PPOs in Min-nesota during the 1980s and early 1990s includedBlue Cross/Blue Shield (which reconfigured itsstandard insurance product as a PPO), hospitalsystems (including Fairview and LifeSpan), andindemnity insurers (table 4-4). Hospital systemsdeveloped PPOs to reduce their dependence onHMOs for patients and as a means of developingcloser ties with their physicians.

—.

Chapter4: Growth of Managed Care and Integrated Delivery Systems 129

1991 Eligiblas/PPO Headquarter Paren$ owner or managar enrollmentAetna PPO Ed ma Aetna Health Plans 9,600Blue Cross and Blue Shield of Blue Cross and Blue Shield of

Minnesota Eagan Minnesota 1,058,805Ethlx-Midwest “ Bloomington Investor-owned 78,202Group Health PPO Minneapolis Group Health 4,453Medlca Choice PPO Mlnnetonka Medlca 54,551

Northwestern Nahonal LifeNWNL PPA St Paul Insurance o*

HealthOne, Fa[rwew and NorthPreferred One Minneapolis Memorial hospdals 237,000Prudential Plus and PruNetwork Minneapolis Prudential o’Select Care Bloomington LlfeSpan hospitals 170,000

● Both NWNL and Prudential reported fhaf they first began PPO enrollment on Jan 1, 1992Enrollment Includes Indlvlduals Ilvlng outside the Twin Cltles metropolitan area

SOURCE Cmzens League Research, “’Minnesota Managed Care Review 1992 “ Mlnneapolls, MN August 1992

I Recent MergersThe consolidation of health care providers in

the Twin Cities occurred in three phases. Duringthe 1980s, the consolidation largely centered onhorizontal integration of hospitals to form multi-hospital systems. During the 1980s mergersamong HMOs began, but the absorption of rela-tively small HMOs by larger ones had little effecton the overall HMO market. The beginning of the1990s saw a substantial shift in the scale of merg-ers for both hospitals and HMOs. The merger ofHealth One and LifeSpan was the most significantmerger among hospital systems to that point and,for the first time, raised serious concerns about ag-gregation of market power relating to the provi-sion of inpatient services. The merger of SHAREand PHP was the first merger of large HMOs in theTwin Cities and resulted in a substantial consoli-dation of enrollment in the HMO market. As itturned out, these mergers represented only theleading edge of a series of consolidations and or-ganizational reconfiguration that fundamentallychanged the nature of the health care delivery sys-tem in the Twin Cities.

The Merger of Group Healthand MedCenters

Merger discussions first began between Med-Centers and Group Health in 1991 in the wake of

the SHARE/Physician Health Plan merger dis-cussed above. There was a concern on the part ofboth organizations that they would not be able tocompete effectively with Medica and Blue Cross/Blue Shield for employer contracts when employ-ers demanded a “total replacement” productoffered by a single health care organization. Nei-ther MedCenters nor Group Health offered pro-vider networks with the comprehensivegeographic coverage that could be offered by BlueCross/Blue Shield and Medica, and both HMOshad lost employer contracts because of this. Theimmediate precipitating factor for the merger,however, was the development of the BusinessHealth Care Action Group and the Request ForProposals (RFP) that it issued in the Spring of1992. (The development of the BHCAG is dis-cussed in detail in the next section.) In order to of-fer a competitive bid, both MedCenters and GroupHealth believed that they needed to increase thesize and geographic coverage of their providernetworks. Therefore, they began to discuss thepossibility of submitting a joint bid in response tothe BHCAG RFP. This led naturally into discus-sions of a more formal merger between the two or-ganizations.

A merger agreement was signed by MedCent-ers and Group Health in August 1992. The agree-ment created a holding company, HealthPartners,

30 I Managed Care and Competitive Health care Markets: The Twin Cities Experience

to be governed by a board drawn from both orga-nizations. The 17-member board consisted of 13consumer representatives and four providers. Thisreflected an insistence on the part of Group Healththat the merged entity maintain a governancestructure dominated by consumers. HealthPart-ners continued to offer both HMOs as separateproducts with separate governing boards but de-veloped a new joint product as part of its responseto the BHCAG RFP. HealthPartners was subse-quently chosen by the BHCAG as the winningbidder. At the time of the merger, HealthPartnershad 40 medical clinics (24 owned and 18 underlong-term contracts) and contracts with four hos-pitals. Most of HealthPartners enrollees receiveinpatient care at FairView-Riverside Hospital andat Methodist Hospital, the hospitals that histori-cally have provided the majority of care to GroupHealth and MedCenters enrollees, respectively. In1992, the merged organizations reported about580,000 enrollees and revenues of approximately$860 million, making HealthPartners slightlylarger than Medica at that time. After the merger,Medica and HealthPartners accounted for about90 percent of HMO enrollees in the Twin Cities( 1 2).

In addition to the precipitating factors alreadymentioned, there were several other consider-ations on the part of both parties that supported adecision to merge. For example, during the late1980s, MedCenters had experienced substantialdiscord in its relations with its major physiciangroup. the Park-Nicollet Clinic. Relationships be-tween the Clinic and the HMO, while greatly im-proved. were still somewhat unsettled whenGroup Health initiated merger discussions. Thesestrained relationships made it difficult for Med-Centers to respond quickly to changes in thehealth care market and to engage in longer-rangeplanning efforts. Also, MedCenters suffered a de-cline in its enrollment beginning in the late 1990sthat was linked to price competition and limita-tions in the geogaphical coverage of its providernetwork, and this made the board of MedCentersreceptive to examining a wide range of altern-atives for the health plan.

From Group Health’s standpoint, the mergeroffered several advantages in addition to those al-ready described. The strength of Group Health’sphysician group was in primary care, and Med-Centers’ brought a strong multispecialty grouppractice to the program. Affiliation with Med-Centers’ specialty physicians was seen as an assetin enhancing Group Health’s image in the com-munity as a provider of quality medical services.Also, while Group Health had strong penetrationof public employee groups, MedCenters’ enroil-ees were drawn primarily from the private sector,with a substantial number of enrollees from firmsthat participated in the BHCAG. Thus, from bothan employer group and a physician network stand-point, the merger offered advantages to GroupHealth.

The Merger of HealthPartnersand Ramsey HealthCare

The creation of HealthPartners set the stage forthe merger of that organization and RamseyHealthCare (48). The merger of MedCenters andGroup Health caused a reassessment of hospitalrelationships for the combined entity. WhenHealthPartners was awarded the BHCAG con-tract, it became necessary for HealthPartners todevelop new hospital relationships relativelyquickly in order to be able to deliver services with-in the premium offered to the BHCAG. Health-Partners issued an RFP to all hospitals in the TwinCities asking for proposals for new long-termrelationships with the health plan. The intent ofthese new relationships was to develop closer col-laboration between the health plan and the hospi-tals used by its members, while at the same timeholding increases in hospital expenditures to zeroin the near term. This process stimulated initialdiscussions between HealthPartners and RamseyHealthCare in August of 1993 that converged veryquickly on the possibility of merger of the two or-ganizations. A letter of intent to merge was signedon September 15, with the formal merger com-pleted on December 2, 1993. Under the terms ofthe merger, the two nonprofit organizations com-


bined their assets. HealthPartners assumed man-agement control of the three different componentsof Ramsey HealthCare: the hospital with 325staffed (435 licensed) beds, a 200-member multi-specialty physician group (the Ramsey Clinic),and an educational and research unit (the RamseyFoundation). Since Ramsey HealthCare wascreated as a public benefit corporation by statelegislation in 1986, the merger cannot actually becompleted without action by the legislature thatwould change the status of Ramsey HealthCare.Legislation will be introduced in 1994 to accom-plish this.

From the perspective of Ramsey HealthCare,the merger with HealthPartners offered several ad-vantages. As a result of its own long-range plan-ning process, Ramsey HealthCare had concludedthat the Twin Cities’ health care market wouldsoon be dominated by a very small number of pro-vider organizations and purchasing groups. As anindependent organization, Ramsey HealthCarebelieved that a merger with an existing HMOwould be more desirable than an affiliation with ahospital group that was attempting to form anISN, because the HMO would possess greater ex-perience and expertise in performing the functionsof an ISN. Ramsey HealthCare believed that itwas well-positioned to be a partner in such a merg-er. It had generated operating surpluses over thepast few years, was in the process of renovatingand adding to its existing facility, was capable ofproviding primary and specialty care onsite, andpossessed substantial strength in the areas of trau-ma and bum care. A merger with HealthPartnerspromised a continued patient flow to supportRamsey HealthCare’s teaching mission and wasexpected to help the Ramsey Clinic in attractingand retaining physicians.

HealthPartners found the prospect of a mergerwith Ramsey HealthCare to be attractive for sev-eral reasons. Foremost was the belief that toachieve savings on inpatient cost in the future itwould need to develop a much closer managementrelationship with hospitals. A merger withRamsey HealthCare offered the potential for bet-ter integration of inpatient and outpatient servicesreceived by enrollees living in the east metropoli-

tan area and therefore greater cost control in thelong run. And, to maintain geographic coverage ofthe metropolitan area with respect to inpatientcare, HealthPartners needed a linkage with a hos-pital or set of hospitals in the eastern metropolitanarea. The merger with Ramsey HealthCare filledthis need.

The Merger of HealthSpan and MedicaThe merger of HealthSpan and Medica, an-

nounced on December 8, 1993, was the first merg-er in the Twin Cities between a hospital systemand a health plan. The assets of the existing or-ganization were merged under a new entity, Alli-na, that was itself organized as three divisions:delivery system (including hospitals, long-termcare facilities, and home health care agencies),physicians (employees of the hospital system orcontracting physician group practices), and man-aged care (insurance and managed care compo-nents, including integrated service networks thatwill be formed in the future). There are approxi-mately 750,000 members in existing Allina healthplans (550,000 in Medica products and 200,000 inSelectCare, a PPO sponsored by HealthSpan).The combined annual revenues of Medica andHealthSpan are $8 billion, and the organizationstogether employ about 16,000 individuals.HealthSpan owns or manages 17 hospitals in Min-nesota and Wisconsin as well as 45 clinics. It has3,200 affiliated physicians. Medica contracts with5,000 physicians and is managed by UnitedHealth Care Corporation under a long-term man-agement contract. (The merger will necessitate arenegotiation of the terms of this contract.) Thenew Allina will be the largest nonprofit firm inMinnesota and the eighth largest firm overall (61 ).To manage Allina, the existing boards of Medicaand HealthSpan will be dissolved and a new boardwill be created.

The merger partners point to the Minnesota-Care legislation as the motivation for the merger.Allina will form the basis for an Integrated ServiceNetwork (ISN) that will meet the requirements ofan integrated health care delivery system underthe new legislation. It is expected that some reduc-tion in the number of hospital beds will occur un-


der Allina, and there will be consolidation of otherservices as well. Allina also plans to create ahealth plan option it hopes members of the Busi-ness Health Care Action Group will offer to theiremployees as an alternative to HealthPartners.

I Ongoing DevelopmentsObviously, the Twin Cities’ health care market

is in a period of very rapid change with the ulti-mate configuration of the delivery system stillopen to debate. Most of the actors in this systembelieve there will be three to four large organiza-tions, formed through merger or contractual aflli-ation, that will dominate health care delivery inthe Twin Cities. The ultimate closure or conver-sion of four or five more hospitals is anticipatedwith a reduction in the number of acute inpatientbeds of as much as 50 percent. The current trendinvolving the purchase of physician practices byhospital systems and HMOs is expected to contin-ue; already, there are very few independent physi-cian practices in the eastern metropolitan area.

While there is agreement among key actors inthe Twin Cities’ market about the general formthat the community’s health care delivery systemwill take in the future, there are several ongoingdevelopments that will have an important impacton this form. These include the conditions of con-tractual arrangements and other agreementsamong provider groups, the definition of the Uni-versity Hospital and Clinics’ role in the new sys-tem, and the strategies adopted by BlueCross/Blue Shield.

Contractual Relations AmongProvider Organizations

In theory, ISNs will be fully integrated systemsthat will be able to rationalize the use of healthcare resources and increase the efficiency withwhich care is delivered through close collabora-tion among participating providers. The necessityto compete for patients will provide a stimulus forthe coordination of resources to achieve systemefficiencies. Presumably, provider incomes willbe closely tied to the success of their ISNs, so theywill have a strong motivation to work collabora-

tively to keep costs down while at the same timeproviding a product that is responsive to the de-sires of consumers. Again, in theory, competitiveincentives would be strongest if ISNs offered con-sumers and payers a clear choice among providersystems. Then providers that collaborated underISN organizational umbrellas would benefit fi-nancially if their organizations prospered relativeto competitors.

In general, the development of the Twin Citiesmarket to date has not resulted in close exclusiveties between provider groups and health care orga-nizations. Even HealthPartners, which may at thistime be the organization in the Twin Cities thatmost closely approaches this model, does not haveexclusive contractual arrangements with some ofits key providers. For instance, the Park-NicolletClinic, HealthPartners’ major provider group inthe western metropolitan area, has recently an-nounced a merger of assets with Methodist Hospi-tal, the major supplier of inpatient care toHealthPartners in the same area of the Twin Cities,to form Minnesota Health Systems. One of thepurposes of the merger is the development of anISN under MinnesotaCare. Minnesota HealthSystems is seeking an insurance partner for thispurpose, with Blue Cross/Blue Shield being thelogical candidate. Thus, HealthPartners faces thepossibility of being in direct competition for en-rollees in some markets with a major componentof its delivery system. Even if a non-competeclause were negotiated in its contracts with Min-nesota Health System providers, it could be diffi-cult for HealthPartners to establish sufficientorganizational loyalty on the part of providers toachieve delivery system rationalization underthese circumstances.

A second example is provided by relationshipsbetween hospital systems and emerging ISNs inthe Twin Cities. Most hospitals expect to continueto sell services to a range of purchasers regardlessof their own sponsorship or ownership positionsin ISNs. For example, Allina, which owns hospi-tals in the northern suburbs, expects to continue toprovide inpatient services to HealthPartners andB1ue Cross/Blue Shield enrollees living in thisarea. Fairview Hospital Systems, which has en-


tered into a collaborative arrangement with BlueCrosdBlue Shield and the University Hospitaland Clinics to form an ISN, expects to continue toprovide inpatient care to HealthPartners’ enroll-ees.

In short, an interlocking web of contractualrelationships among different provider organiza-tions continues to exist below the overlay of ISNformation, and these relationships often areamong entities that will presumably compete witheach other through their affiliations with or own-ership of ISNs. It is not clear at this time how orwhether these relationships will be modified overtime to link specific groups of providers moreclosely to specific ISNs. It is also not clear howthey might affect the process of health care ratio-nalization through competition among large inte-grated health care delivery systems as envisionedby the legislative architects of MinnesotaCare.

Blue Cross/Blue ShieldDuring the 1970s, Blue Cross and Blue Shield

of Minnesota (BCBSM) enrolled over 70 percentof the private health insurance market in the TwinCities. During the early 1980s, as HMOs gainedmarket share and reportedly attracted healthier en-rollees, BCBSM premium increases were oftensubstantial. Despite this, the company lost over$10 million during 1986 and nearly $20 millionduring 1987. These operating losses causedBCBSM to completely restructure its health in-surance product lines. A PPO (Blue Cross andBlue Shield of Minnesota) was formed that hadover one million enrollees by 1990. A networkmodel HMO (Blue Plus), featuring an open-endedoption, evolved from HMO Minnesota, a BlueCross/Blue Shield-sponsored HMO. This pro-gram grew slowly during the early 1980s, but by1990 had 70,000 enrollees. An indemnity healthinsurance plan (Aware Gold), was developed thatwas very similar to the traditional BCBSM pro-gram but offered a comprehensive benefit packagecompetitive with the HMO products and devoteda great deal of effort to cost controls. Aware Goldwas very popular in the early 1980s because itgave consumers a generous benefit package and

free choice of provider. By 1989. there were over600,000 enrollees in this plan.

To compete in the current Twin Cities healthcare market, BCBSM has developed a new man-aged care strategy with three components (47).The first and most prominent is an extensive on-going analysis of small area variation in physicianresource utilization. While this approach report-edly has been effective in changing the practicestyles of some physicians. it does not provide amajor competitive advantage for BCBSM. Anysavings resulting from changes in physician re-source use accrues to other health plans as well asBCBSM. The second component focuses on pro-vider payment systems. Here the main thrust is tolink payment to severity of illness, institute risk-sharing agreements with groups of physicians,and provide extra payments for preventive ser-vices. Again, the gains from this strategy are oftenshared by competing plans (e.g., prevention) andBCBSM patients are often a relatively small pro-portion of a physician’s practice, reducing the im-pact of these interventions. The final componentconsists of a series of initiatives relating to themanagement of resource use (e.g., use of primarycare physicians, gatekeepers. designation of refer-ral specialists, and drug formularies).

In addition to these strategies, BCBSM intendsto pursue ISN development. in May BCBSM an-nounced a partnership with Affiliated MedicalCenters in Willmar for the creation of an ISN toserve a 14-county area in southwestern Minneso-ta. In June it indicated that it would participate in apartnership with Park-Nicollet Medical Centerand Methodist Hospital to form an ISN in the westmetropolitan area. In July it announced a partner-ship with Aspen Medical Group to lay the ground-work for an ISN to serve residents in the TwinCities’ metropolitan area. This ISN would be de-veloped in cooperation with Fairview Health Sys-tems and University of Minnesota Hospital andClinics.

BCBSM also intends to create a package of ser-vices that can be marketed to hospitals and medi-cal groups that are developing ISNs. Theseservices will include actuarial services, claims


management, payment systems, information sys-tems, quality assurance programs, utilization re-view systems, and some clinical guidelines. Thispackage, or parts of it, will be available as a ser-vice component provided under subcontract toISNs or through partnerships to establish ISNs.

University of Minnesota Hospital and ClinicsIn a competitive health care market an obvious

issue is: can, or should, a university hospital andaffiliated clinics compete effectively for patients?In the Twin Cities, most actors acknowledge thatall would benefit from the continued presence of astrong medical school that can provide high-quali-ty training for new physicians. However, there isless consensus concerning the appropriate role forthe university of Minnesota hospital and clinics inthe evolving system. Some contend that theeducation and research components of the univer-sity’s mission seriously limit its ability to competewith more service-oriented community hospitalsfor the patients of large health plans. Others arguethat the university may be able to successfullycompete for tertiary care patients, but will not beable to maintain a broad enough patient base tosustain excellence in clinical training programs.Under these circumstances, one alternative is forthe University Hospital to be purchased by ormerged with one of the existing large health careorganizations under an agreement to continue tooperate it as a university teaching and researchinstitution. The acquiring organization could thenconcentrate the patients needed for that role at theUniversity Hospital and phase out duplicate ser-vices in other parts of its system. Acquiring theUniversity Hospital could result in a competitiveadvantage for the purchaser because of the pres-

tige of the university system. However, it might bevery difficult to preserve the role of the hospital, asenvisioned by the university, when control is giv-en over to others.

An alternative strategy which the UniversityHospital appears to be pursuing is to be an aggres-sive competitor in the evolving system. To do so,the hospital has formed a corporate structure thatbrings the clinical faculty and hospital together tocontract with ISNs and negotiate with health in-surance plans. This organization, the Universityof Minnesota Health System, has the capacity todevelop health services programs, create newhealth plans, and bid on contracts to provide ser-vices to employers or health plans. As noted pre-viously, the university has chosen to become apartner with Blue Cross/Blue Shield and FairviewHealth System in the formation of an ISN. TheUniversity Hospital will be the secondary and ter-tiary care facility in this system. It hopes that byparticipating in this ISN it will be able to assure acontinued flow of patients for its teaching and re-search programs. The danger in this approach isthat providers affiliated with competing organiza-tions could restrict referrals of patients to univer-sity physicians and withdraw from participationin its teaching programs. These losses could out-weigh the gains from the ISN partnership. Theuniversity has responded to this concern by re-maining available as a participant in other ISNs.According to the president of the University ofMinnesota Health System, “Not to join [an ISN]could mean being left without a patient base in thecompetitive Minnesota health care environment.At the same time, we won’t be an exclusive part-ner. Our mission makes it imperative that we beavailable to any Minnesotan who needs us.” (63)

I

Developmentand Role ofPurchasing

Coalitions 5

As was previously discussed, many employers in the TwinCities have begun to restructure the way in which theypurchase health care (60). Their actions have had a majorinfluence on health plans and have contributed to the on-

going reconfiguration of the health care delivery system in theTwin Cities. In this chapter, several of these efforts are addressed.First, the development of the Business Health Care Action Group(BHCAG), a consortium of major employers in the Twin Cities,and the implementation of its purchasing approach are discussed.Second, the ● ’managed competition” approach used by the Stateof Minnesota Group Insurance Program to purchase health carefor 144,000 individuals in Minnesota is described. This programhas been cited as one example of how a “health alliance” mightfunction under the Clinton Administration’s health care reformproposal. Finally, two approaches--one public and one private--tothe formation of insurance pools for health care purchasing areexamined.

BUSINESS HEALTH CARE ACTION GROUP| FormationDuring 1988, several large private-sector firms headquartered inMinneapolis/St. Paul formed a coalition called the BusinessHealth Care Action Group to lobby for health care reform. Duringthe fall of 1991, the coalition decided to create a health plan fortheir employees and dependents. All of the firms were self-in-sured, and they ranged in size from 8,000 to 200,000 employeesnationwide. BHCAG contracted with a benefit consulting firm todevelop a request for proposals (RFP) that could be circulated topotential provider groups and third-party administrators. The re-sultant RFP was a very detailed document in which potential bid-

I 35


rs were asked to identify the specific sources ofhigh-cost, complex technological services such asMRIs, lithotripsy, and hemodialysis, and to speci-fy the volume of services provided by thosesources during the past year. Bidders were also re-quested to submit the credentials of the physiciansproviding these services along with any availableinformation on clinical outcomes. The benefitpackage included in the RFP was based on the ex-isting programs offered by the sponsoring firmsand, consequently, included an extensive array ofservices consistent with the traditionally generoushealth care benefits of these firms.

Six health care provider groups bid on theBHCAG request for proposals. These firms in-cluded two large HMOs (MedCenters Health Planand Group Health, Inc.) that joined together to re-spond to the RFP, Medica health plan (a large in-dependent practice association HMO), Blue Crossand Blue Shield of Minnesota, and Preferred One(a hospital-owned PPO). The MedCenters/GroupHealth coalition (HealthPartners) was the suc-cessful bidder, although at least one other bidderoffered the services at a lower price. However,Health Partners and its providers were judged to beIn a position to provide services that were moreappropriate and accessible while still assuring thatcomplex technological procedures were appropri-ately limited to settings with highly qualified per-sonnel, extensive experience, and sufficientvolume to assure quality of care.

The network offered by Health Partners in-cludes approximately 1,000 primary care provid-ers practicing in over 175 primary care clinic sites.The network’s primary care physicians are sup-ported by more than 4,000 specialty physicians,nearly 30 community hospitals and specialty hos-pital “centers of excel lence.” When enrollees join

the pian, they must select a primary care clinicwithin the network. However, each member of thefamily may choose his or her own clinic and canchoose a physician from among those practicingat the selected clinic. The clinics have differentpolicies concerning access to specialty care. Mostc1inics require that a patient receive a referral by aprimary care physician before making an appoint-ment with a specialist. If enrollees seek care out-

side the network, they have a lower level of benefitcoverage and must submit claims forms.

| ImplementationAt the firm level, the Health Partners product is be-ing offered as the basic self-insured plan. Mostemployers offer employees an option to enroll inthis plan or select one of the other plans offered bytheir employer. Currently, employers are offeringtheir employees two or three competing plans, butit is generally assumed that these plans will be re-placed over time with new health plans that con-tract with BHCAG. In most cases, the employer’scontribution to an employee’s health plan is 1im-ited to the contribution that would be made to theBHCAG plan.

During the first year of operation, 10 of the 14coalition members offered the BHCAG healthplan. During 1993, an additional eight employersjoined the coalition, bringing total membership to22 companies. While interest was reported to behigh, many of the firms had employee and unioncontractual agreements and commitments to otherproviders that made it difficult to shift to theBHCAG plan. During 1992, the first year of’ theHealth Partners’ contract, 55,000 employees en-rolled, with about 70 percent of these being pre-vious MedCenters or Group Health members. It isanticipated that enrollment will grow to 85,000enrollees during 1994, representing about 30 to 35percent of the eligible employees.

Although the BHCAG plan is modeled on theHMO concept, it is not currently operating on ac:apitation basis. Instead, providers are paid on afee-for-service basis. Each employer has a con-tract with the providers. To do otherwise wouldviolate the self-insured provisions of ERISA andwould place the program under the supervision ofthe state insurance commissioner. This reportedlywould limit the flexibility of employers in struc-turing their benefit package. Currently, there is nostandard fee schedule. Rather, fees are negotiated,often resulting in discounts of 20 to 30 percent.BHCAG is experimenting with severity measuresto adjust fees, and is attempting to implement theJohns Hopkins Ambulatory Care Groups method-

Chapter 5: Development and Role of Purchasing Coalitions | 37

ology for selected cases. Health Partners is paid aset fee (currently 8 percent of total expenditures)to administer the program. BHCAG employersestimate that the plan reduced their expectedhealth care costs for the enrolled employees byabout 10 percent during 1993 compared with sim-ilar coverage available through competing man-aged care products.

I Programs for Quality ImprovementThe BHCAG quality improvement program con-sists of both clinical guidelines focused on cost-effective modalities and clinical outcomesassessment programs. Eight of the provider medi-cal groups have volunteered to serve as pilot sitesfor testing and implementing guidelines. Sixteenguideline topics were selected for the initial pro-gram, and installation of these guidelines is nowunder way at pilot medical groups. In total, 18medical groups will participate in guideline devel-opment and implementation, representing about88 percent of the volume of care delivered by par-ticipating providers in the BHCAG health plan.

Six clinical indicators have been identified tobegin benchmarking quality across the BHCAGprovider network (table 5-1 ). Joint purchaser/pro-vider assessment of new and emerging technolo-gies is another key quality control initiative.During 1993, a joint purchaser/provider groupwas established to review the effectiveness of cer-tain medical technologies. The scientific assess-ment of these technologies will be linked tobenefit coverage decisions. Topics reviewed todate include: cochlear (ear) implants, bone mar-row rescue with chemotherapy for breast cancer,laser surgery to correct vision problems, pancreastransplants, chest compression devices for cysticfibrosis, immune globulin for neurological condi-tions, lung transplantation, and PSA for prostatecancer screening. Development of a prototype au-tomated medical record will be completed by theend of 1993.

I ImplicationsThe employers who formed BHCAG feel thatknowledgeable purchasing groups are the key to

restructuring the health care system. To that end,they believe BHCAG has the responsibility to de-velop quality assurance programs, structurehealth plans, and use its purchasing power tocreate competing cost-effective provider systems.From this perspective, they view BHCAG as acommunity service as well as a service to themember firms. Consequently, although BHCAGplans to invest at least $30 million in the develop-ment of practice guidelines, these guidelines willbe available at no cost to non-BHCAG health careproviders as well as competing health insuranceplans.

According to BHCAG’S annual report, the av-erage cost to employees for single coverage is$1,200, while family coverage is $2.500. BHCAGestimates its savings in the first year to be about 11percent. However, it cautions that approximately2 percent of these savings are due to higher copay-ments and another 2 percent are due to state taxesand fees not payable under self-funded plans.

BHCAG has expressed strong support for com-petitive markets for health services. However, itsactions helped precipitate the merger of two largeHMOs—MedCenters and Group Health. In re-sponse to criticisms that it has not promoted acompetitive health plan market in the Twin Cities,BHCAG has stated its intention to offer otherproducts in the near future to member firms incompetition with the current health plan--Health-Partners. These products presumably will offermore choices with respect to benefit coverage,providers, and coinsurance and deductible provi-sions.

THE STATE OF MINNESOTA GROUPINSURANCE PROGRAM

The State of Minnesota Group Insurance Pro-gram (or State of Minnesota Employees HealthBenefit Program) covers 57,000 employees.Along with dependents and retirees, the number oflives covered by the program is approximately144,000, making it the largest employer-basedhealth insurance group in the state. State em-ployees work in every county in Minnesota and,


I

Breast cancer

% of women with mammogram ordered

% of women aged 50-74 with mammogram

% of newly diagnosed cases stage I or less

% of diagnosed cases stage II or less

Total hlp replacement

Measure of functional status before and aftersurgery

ChildbirthVaginal birth after C-section rate

C-section rate% of deliveries that occur at <37 weeks

Heart disease

30-day mortality following Coronary Artery BypassGraft

Childhood infectious disease% of children aged 27 months who have had lmmu-nizations recommended by the Minnesota Depart-ment of Health

Asthma in children

Rate of hospital admssion for asthmatic childrenaged 0-18 years

SOURCE Business Health Care Action Group, 7993 Annua/Repor,Mmnetonka, MN, 1993

therefore, the state’s health benefits program mustserve the needs of people in urban and rural areas,and in areas with and without HMOs.

| EligibilityEmployees and their dependents are eligible to en-roll. However, spouses of eligible employees whocan participate in their own plan, but who have re-ceived cash or credit not to participate, are ineligi-ble for state health benefits. There is a 28-day waitfor eligibility from the point of hiring. Upon be-coming eligible, the employee has an open enroll-ment choice of plans. Moving to an area where theemployee’s plan is not available is the only situa-tion in which the employee is allowed to changeplans between scheduled annual open enrollmentperiods. If the employee moves to a county wherethe present plan is offered but where the employ-er’s premium contribution is different, the con-tribution is frozen at the previous level for the restof the year.

| History -The original health plan offered to state em-ployees before the advent of HMOs, and the onlyplan available to employees statewide, was a fee-for-service, Blue Cross and Blue Shield of Minne-sota product. During most of the 1980s, this planenrolled half or more of the State of Minnesotagroup. By state law, any HMO that wished to beoffered to state employees was allowed into thestate employees’ program. As a result, the state of-fered a large number of HMOs--at times as manyas 10. Under the same law, the state’s contributiontoward the cost of health insurance was tied to thefee-for-service premium--100 percent contribu-tion for employee coverage and 90 percent for de-pendent coverage. Employees did not receive apremium rebate if they picked an HMO that costless than the fee-for-service plan, but they had topay the difference if they picked a more expensiveplan. HMO rates tended to cluster near the fee-for-service rate. The rates submitted by the HMOs andBlue Cross/Blue Shield were not examined criti-call y by the state. In other words, the State of Min-nesota was a fairly typical large employer offeringmultiple health plans.

During 1985 the state consolidated its HMO of-ferings and changed the basis for determining itspremium contribution. The number of HMOs par-ticipating in the state employee’s program fellfrom a high of 10 to 6 at the beginning of 1990.This reduction occurred for a variety of reasons,including: the 1985 repeal of the law requiring anopen-door policy toward HMOs; HMO attritionand mergers; rejection of applications to join theplan submitted by HMOs that did not meet thestate’s criteria and objectives; departure of anHMO that could not maintain reasonable pre-mium rates; and, an insurer or HMOs no longerbeing allowed to offer more than one option to em-ployees or to add plans at its own initiative. Offer-ing fewer HMOs resulted in larger market sharesfor the remaining plans and provided a chance forthem to gain even more enrollees by offering an at-tractive, well-managed plan. Offering fewerHMOs also diminished the prospects for biasedselection and eliminated the possibility that health


plans could add options to undercut a competitor’sposition and/or to ● *shore up” an existing plan.

The most significant reform during this periodwas changing the formula for determining the em-ployer contribution. The state, through collectivebargaining with 10 unions that represent state em-ployees, replaced the formula based on the fee-for-service plan with one based on the low-costcarrier (health plan) serving a given county. Underthe new formula, which was adopted in October1985 for the 1986 contract year, the state contin-ued to pay 100 percent of the premium for em-ployee coverage and 90 percent for dependentcoverage, but the contribution was based on thelow-cost carrier in each county rather than the fee-for-service plan premium. To be eligible to be thelow-cost carrier, health plans were required toserve the entire county.

In the first few years after the low-cost carrierformula was introduced (from 1986 through1988), the fee-for-service plan continued to havethe lowest rate and remained the basis for the em-ployer contribution. (From 1985 to 1988, the statecontinued to contract with an outside vendor forits computerized payroll systems, and this vendorhad difficulty implementing the low-cost formu-la. For that reason, HMOs may not have fullyreacted to the low-cost carrier program until thecomputer system was changed in 1989.) Overtime, however, the HMOs were able to offer lowerrates despite offering better coverage. In 1989,seven different HMOs were low-cost carriers in atleast some part of the state. (Plans submit one sta-tewide premium, but not all plans are offered in allcount ies, so d ifferent plans are the low-cost carrierfor different counties.)

Introduction of the low-cost carrier formula ledto striking changes in the pattern of health planpremiums, as shown in tables 5-2, 5-3, and 5-4.HMO premium rates in 1988 (the last year beforethe formula began to have an impact) still tendedto cluster around the fee-for-service rate. Pre-miums for the four largest HMOs (Group Health,Share, MedCenters, and Physicians Health Plan)averaged 112 percent of the low-cost fee-for-ser-vice plan sponsored by BCBSM. Table 5-5 showsthe trend in average total premiums (the portion

paid by both the employer and employee) forsingle and family contracts for all state employees(not just Twin Cities’ employees) from 1980 to1994. These average premiums reflect not onlychanges in premiums from year to year, but alsochanges in the health plans’ market shares. Thetransition period to the low-cost program(1985-1988) was chaotic, characterized by wildswings in premiums. During much of this period,the premium growth rate was above the nationalaverage. Since 1989, however, the percentage in-crease in total expenditures on health insurance bythe state and its employees has fallen steadily toless than 3 percent in the period 1993 to 1994 andhas been below the national average (62).

The low-cost carrier formula created a substan-tial incentive for plans to submit the lowest pos-sible rate regardless of the fee-for-service rate.Holding the premiums of other plans constant,each health plan knows that a $1.00 increase in itspremium will reduce the employee’s out-of-pock-et premium differential between itself and highercost plans by $1.00 and increase the premium dif-ferential between itself and lower cost plans by$1.00. The effect of both types of changes result-ing from a premium increase will be to decreasethe health plan’s market share. The new formulaenhances regional competition among HMOseven if a plan is not the low-cost carrier in the TwinCities, it may be low-cost in another area.

Since 1989 Group Health, a staff model HMO,has been the low-cost carrier in every county inwhich it was offered with family premiums some-times $1,500 per year below those of competingplans. In contrast, independent practice associa-tion-model HMOs with open networks, and plansthat allow out-of-network coverage, have had thehighest premium rates. This pattern was not evi-dent until the low-cost carrier formula took effect.Over the period 1988 to 1994 it is interesting tonote the experience of the two largest plans:BCBSM, which evolved into the “State HealthPlan” option, and Group Health. The BCBSMplan typically was the most expensive or secondmost expensive plan during that period, whileGroup Health was consistently the lowest cost

— — . —

% Yo Yo Yo Yo Yo 9’0 Yo1986 1987 Incresse 1988 Incresse 1989 Increase 1890 Increase 1991 Increase 1992 increase 1993 Incresse 1994 Increase

Group Healtha $2118 $2173 26% $2315 65%First Plan HMOb 2008 2059 25 2228 8 2

Shareb 2142 2010 - 6 2 2394 191

Central MN GroupHealthb 2361 1990 -157 2293 152

MedCenters HealthPlanb 2142 2207 30 2363 70

HMO Goldc — — — 24(X) —Physlclans Health

pland 2309 2076 -101 2481 195Mayo Health Plane — — — 2397 —Blue Cross/Blue

Shlelc# 1998 1908 -45 2128 115

State Health Plan9 — — — — —

a staff-~del health maintenance organization (HWb netwxk-nl~el HMOc nefwork-rnodel HMO with out-of-network c0Vera9ed Independent practice assoclat~on (I PA) with out-of-network c0vera9ee IPA-model HMO

‘ fee-for-serwce plan9 preferred provider organization with out-of-network coverage

KEY — md!cates that a plan was not offered

$266227592848

15 o%238190

$305031823273

14 6%153149

$349238223665

145% $3715 64%201 4257 114120 3891 62

$391944504274

55% $42054 5 4735

9 8 4290

7 3%6 404

3031 322 3343 103 3834 147 — — — — — —

28343241

191350

3655—

290—

4087—

118 4537 110— — —

4603—

15 5127— —

114—

33983655

370525

3909—

150—

4405—

127 4790 87— — —

5422—

132 5497— —

14—

3461—

62.6—

-1=m

—4037

——

—4263

— — —5 6 4477 50

—4710

— —5 2 4759

—10

SOURCE: J Klein, State Employee Insurance Group Program Manager, Department of Employee Relahons, State of Minnesota, Mmneapok MN. personal commumcatlon, 1994


State Health Plan

5Z0 of total

premium

Group Health Inc

Y. of total

premium

Medlca Choice

% of total

premium

MedCenters

% of total

premium

Medlca Primary

70 of total

premium

Coordinated Health Care

94 of total

premiumHMO Gold

% of total

premium

Total

70 of total

3,259

42.53’XO

$01,894

24 72Y0

$0481,082

14.1270

$2.22586

7.65Y0

$0290

3.78Y0

$ 3 1 8114

1 49’%0

$4.28437

5.70%

$4287,662

10070

2,114

23 49Y0

$40.06

3,320

36.89Y0

$01,391

15 46?40

$19.581,022

11,3670

$2.92722

8.02%

$218

431

4,79%

$2486

9,000

100%’0

1,050

12,0170

$4676

3,90044,59Y0

$02,006

22.94Y0

$22.88990

11.327.

$19.08

800

9.1 5Y0

$2.78

8,746

1 OOYO

1,04011 ,97’?40

$39,72

3,936

45.31 v.

$01,997

22.99?4.

$24.10946

10.89?40

$18.70

767

8.8370

$0.54

8,686

100’XO

1,079121970$ 4 0 1 6

4,178

47 205Z0

$01,879

21 .23~o

$ 2 8 9 8

893

10.09~o

$ 2 6 6 8

823

9.30’XO

$0.28

8,852

10070

1,201

13 65%

$ 4 1 8 8

4 1 8 0

47 51%

$01,653

18 79~o

$ 4 2 4 8

917

10 42%

$2200848

9 64%

$ 5 8 0

8,799

1 00%

SOURCE J Kleln State Employee Insurance Group Program Manager Department of Employee Relattons, State of Minnesota Mmneapols, MNpersonal communlcallon 1994

plan. Group Health’s market share rose from 18.7percent for employee contracts ( 18.7 percent forfamily contracts) to 36.3 percent for employeecontracts (29.5 percent for family contracts) from1988 to 1994, while BCBSM’s market share fellover the same period from 56.0 percent for bothsingle and family contracts to 43.9 percent (50.4percent for family contracts).

Feldman and Dowd evaluated the State of Min-nesota’s low-cost carrier formula by simulatingthe expenditures that. would have occured if theplan switching had not taken place and enrollment

by plan had remained constant. They found thathealth plan switching saved the state employees$3.8 million dollars in 1993 alone (24).3

| Evolution of the Fee-For-Service PlanAfter the low-cost carrier system was installed,the BCBSM plan incurred a $9 million deficit andannounced it would no longer offer the high-cov-erage option. Several options were considered, in-cluding limiting open enrollment to once everythree years to discourage “hit and run” utilization,

3~e ~u~{)m ~autlon ~at their estimate d{~s n(>[ consider how premiums may have changed w a result of the reforms to the stale emPlf)Yee

pr{)gram and, if plan competitwn increased, maybe too small. In additl(m, they note that the low-cost” plans may have eny)yed fawmable selec-

uon compared with the high-c(xt plans, m which case thetr estimate would be t(w large because It w(mld include favorable selec(l(m as well asefficiency.


l+ealth Plan 19aa 1989 1990 1991 1992 1993

State Health Plan 4,452

70 of total 41.8870

premium . $9.96Group Health Inc 3,064

70 of total 28.25”A

premium $25.52Medlca Cho[ce 1,284

70 of total 11 .84%

premium $3940MedCenters 665

940 of total 6. 12Y0

premium $29.58Medtca Primary 384

70 of total 3.54Y0

premium $32.12Coordinated Health Care 190

?0 of total 1 ,75%

premium $27.66HMO Gold 717

Y. of total 6 581%0

premium $32.66Total 10,846

Y. of total 100’%0

2,75723.60%$79.824,83741 ,41?40$13.181,1229.60Yo$74.621,42712.22Y0

$27.54813

6.96%

$28.72

726

6.21 ‘A

$614611,682

1 OOYO

1,526

13.81%

$97.285,576

50.46%

$15,101,669

15,1 oOA

$86.621,284

11 .62°A

$65.56996

9.01%

$33.64

11,051

100’%0

1,618

1 4,47%

$81.586,798

51 .84V0

$17.321,538

1 3,75%

$93.441,267

11 .33Y0

$66.94963

8.61%

$31.80

11,184

1 OOYO

1,733

1 5.37%

$81.906,020

53,39°A

$18.421,384

12.27Yo

$107.981,170

10.38’XO

$86.92968

8.59Y0

$33.08

11,275

10070

1,957

17.1 o%

$85.306,221

54.37Y0

$19.441,113

9.73Y0

$144.701,21010.5770

$76.42942

8.23Y0

$49,02

11,443

1 OOYO—SOURCE J Klein State Employee Insurance Group Program Manager, Department of Employee Relauons, Stateof Minnesota, Mmneapds, MN,personal commumcallon 1994

dramatic increases in the plan’s deductibles, gate-keeper systems and conversion of the fee-for-ser-vice plan to a preferred provider organization(PPO). The state felt that three-year “lock-ins”would decrease price competition among plansand the unions were not amenable to a “gatekeep-er” system at that time. Large deductibles, on theother hand, had the potential to affect risk selec-tion among the health plans. Ultimately, the statechose to replace the fee-for-service plan with apreferred provider organization. The specifica-tions for the fee-for-service plan were revised ac-cording] y and the plan was put out for competitivebids. Because the plan had to be a PPO and coverproviders statewide, BCESM was the likelychoice and, in fact, was awarded the contract.

In 1994, further changes were made to the PPOand IPA plans (the State Health Plan and MedicaChoice Select, respectively). Medica Choice Se-

lect was experiencing rapid premium increasesrelative to the other plans and responded with theinstallation of a gatekeeper system. The StateHealth Plan, fearing adverse selection, alsoinstalled a gatekeeper system. The effect of thosechanges was to hold the premiums virtuallyconstant for the State Health Plan and the productsoffered by Medica. The family coverage pre-miums for Medica Choice (renamed MedicarePremier) and the State Health plan increased 1.0and 1.4 percent, respectively.

| Managing Open EnrollmentIn February of each year, the state assembles in-formation for the health plans, which it mails tothe plans in March. This information includes fullspecifications for proposals from the health plans.In late April or early May, the plans submit their

-—— —


Average premium Average premiumYear for singie ooverags for famiiy covemge

1980 $ 37.93’ $ 9 6 . 4 51981 45.62 110.231982 54.21 130,791983 60.32 140,511984 68.32 156.671985 7276 170191986 73,31 171,711987 71.06 165,901988 78.37 186,031989 11146 262.991990 126.67 304.341991 138,82 333.041992 14758 352.811993 156.21 372.331994 160.36 38347

Percent change in Percent change inaveqe premium average premium

Year for singie coverage for famiiy coverage

1980-81

1981-82

1982-83

1983-84

1984-85

1985-86

1986-87

1987-88

1988-89

1989-90

1990-91

1991 -92

1992 -93

1993-94

2 0 2 6

1 8 8 3

1 1 2 9

1 3 2 6

6 5 0

0.75-3.07

1

10.28 12.54’”42.2213659.59

631

5.852.66

1428

18667 4 4

11.50

8.63.0.89

-3.39

}

12.13 12.75”’41,37

1572

9 4 3

5.94

5.53

299● Includes both the employer- and employee-paid portion of the pre-

mium““ Average over four years

SOURCE J Klein State Employee Insurance Group Program Man-ager, Department of Employee Relatlons, Stale of Minnesota, Mmne-

apohs, MN, personal communication, 1994

,-

proposals in two parts during meetings with thestate. The first part lists participating providersalong with the capacity of each provider. Thehealth plans inform the state whether their net-works are expanding or contracting. Plans have toindicate not only their participating providers, butalso which providers are accepting new patients.

Complaints registered by employees assist thestate in determining if there are capacity problemswith particular provider groups. In the past, therehave been problems with providers not acceptingnew patients from particular health plans. A meet-ing is then held where the state entertains sugges-tions from the plans for plan design changes.

Rate requests submitted by health plans mustbe supported by demographic information on thestate group and special categories of early retirees,the disabled, and COBRA continuations as well askey assumptions and methods used to project uti-lization and price trends. In evaluating the rate in-formation, the state has found it necessary todevelop a format to frame the discussion. Thestate’s actuaries provide corridors for expectedtrends in various factors in a cost “grid” format ofinflation rates, cost per service, and number of ser-vices. When this format was first instituted, defi-ciencies in HMO data systems made it difficult forthem to comply. Now, however, the state de-scribes the process as approaching a “partner-ship.” During the process of rate negotiation, eachplan remains unaware of the other plans’ sub-mitted prices. The state uses an independent actu-ary to help evaluate the proposed premiums, andpremiums are finalized each year by June 30.

| Collective Bargaining ConsiderationsIt remains important for employee and union per-ceptions that a single plan be available on a state-wide basis with uniform benefit levels andpremium rates---criteria that no HMO has beenable to satisfy. Blue Cross/Blue Shield played thisrole before 1990, with its statewide fee-for-serviceplan. However, cost increases forced the state todrop the BCBSM plan in 1990. In order to offer astatewide plan, the state and the unions negotiatedreforms that substituted a preferred provider or-ganization (PPO) administered by Blue Cross/Blue Shield. Aggressive management of the newPPO held its premium increases to 5.6 percent in1990- 1991,5.0 percent in 1991-1992, and 5.2 per-cent in 1992-1993. These percentages compare fa-vorably with premium increases posted by otherplans, including HMOs. However, many state em-


ployees have expressed dissatisfaction with thePPO’S limited provider network.

The collective bargaining agreement specifiesa certain percentage increase in total compensa-tion for each year of the biennium. Health plansknow that increases in benefit costs will cut intothe increase in compensation, and that the state’snegotiators will adopt a tougher stance when theincrease in annual compensation is small. Everysecond year, when union contracts are being nego-tiated, meetings with unions begin in December,and contracts are settled by July 1 of the followingyear. The printing deadline for the fall open enroll-ment information distributed to employees is mid-August. Thus, health plans must project theircosts as much as 18 months into the future. OnSeptember 15, information is mailed to state em-ployees. Open enrollment takes place during Oc-tober. There is no official grace period duringwhich enrollees can change their mind with re-spect to their choice of health plan, but they are al-lowed to change during November and Decemberif they offer a convincing rationale.

| Current Strategy for SelectingHealth Plans

The state has targets for participation by differenttypes of plans. In general, the state has no wish tolimit the number of “integrated” plans (e.g.,Group Health, Share, First Plan, and Central Min-nesota Group Health) in the state program. Nordoes the state object to two different types of plans(e.g., an IPA and a network HMO) offered by thesame corporation. However, the State is not inter-ested in “look-alike” plans offered by the samecorporation, including plans that vary only in theamount of coverage offered.

| SummaryThe performance of the State of Minnesota GroupInsurance Program can be attributed to severalfactors. An important factor appears to be limita-tion of the employer’s premium contribution tothe lowest cost plan. Another important factormay be the size of the plan and its goals. At144,000 covered lives, the program is large

enough to elicit swift response on the part ofhealth plans to demands for better products andlower prices. However, the program is not so largethat its decisions have become politicized andsubject to regulatory capture. The program’s ad-ministrators expressed concerns about becomingtoo large or influential in the market (34). Ratherthan trying to reform the entire system, they havedirected their efforts at providing the best possibleproducts and prices to state employees.

THE MINNESOTA EMPLOYEESINSURANCE PROGRAM

I Motivation for FormationTo understand the need for the Minnesota Em-ployees Insurance Program, and the key role of in-surance pools versus simple underwriting reform,it is useful to first discuss the nature of the prob-lems faced by individuals and small groups inmarkets for health insurance. Consumers in thismarket may face higher administrative expenses(insurer overhead) than consumers in large groupsand they have higher search costs. Most impor-tantly, however, it has proven difficult for con-sumers in the individual and small group marketto prevent having their risk redefined by their in-surer after the occurrence of a serious illness or in-jury with lingering effects ( 15).

Insurers are often blamed for medical under-writing which results in high-risk consumers pay-ing higher premiums but, in fact, the need forrisk-rating arises not from insurer greed, but fromconsumer behavior. If an insurer offers a health in-surance product that charges a single ● *commune-ty-rated” premium, enrollees are likely to beindividuals with expected health expendituresabove the premium. In order to keep premiumsfrom rising, low-risk individuals and businessesmust subsidize the costs of higher risk individualsand businesses. In the past, that sort of altruismhas not been the norm. The response of insurershas been to risk-rate applicants, so that individualsin the same risk class pay the same premium, re-sulting in high-risk consumers paying higher pre-miums.

—


This problem cannot be solved by requiring ev-eryone to purchase insurance. Suppose that con-sumers in the same risk class pay the samepremium in the first contract period. During thatfirst period, some people experience illness or in-jury that changes their risk going into the secondcontract period. When premiums are quoted forthe second period, the insurer could continue tocharge everyone the same premium, but only if thepeople who remained healthy during the first timeperiod agreed not to switch to another insurer of-fering them a lower rate due to their low-riskstatus. In order to prevent the good risks from be-ing picked off by competitors, insurers have had toprotect the good risks within their own pools byputting them into a separate policy with a lowerpremium. Of course, that leaves the high-risk indi-viduals paying a higher premium.

One approach to the problem of risk redefini-tion is simply to prohibit insurers from raisingpremiums when an individual or group of individ-uals become ill. However, at least one health econ-omist has noted that forcing insurers to charge thesame price to individuals in different risk cate-gories may exacerbate, rather than alleviate, dis-crimination against high-risk individuals (56).Under these circumstances, insurers may find co-vert, nonprice ways to discriminate against high-risk individuals.

Employees of large firms, even those offeringmultiple health plans, do not face the problem ofhaving their risk redefined, as long as they remainemployed. Large employers offer employees amultiperiod contract that protects them againstrisk redefinition in exchange for the employees’willingness to remain in the “pool. ” The employeecontracts are with the pool, however, not with asingle insurer. That arrangement allows em-ployees to change insurers if they become dis-satisfied with the service they are receiving, evenif they are in poor health. Increases in an em-ployee’s premium are based on the experience ofthe pool, not the individual’s experience. Insur-ance pools may have other advantages, includinglower administrative costs, better consumer in-formation on health plan choices and greater pur-

chasing “clout” in the health insurance and healthcare services markets.

| ImplementationIn Minnesota, the state legislature responded tothe problems in the individual and small groupmarket by creating the Minnesota Employees In-surance Program (MEIP) as part of the 1992 Min-nesota Care legislation. Private businesses withtwo or more employees and 95 percent of theiremployees working in Minnesota are eligible toenroll all their employees in MEIP (but eligibilityis not limited to small employers). Minnesota em-ployers with less than 95 percent of their em-ployees working in Minnesota may enroll theirMinnesota employees in MEIP. Employees mustwork at least 20 hours per week for the business,and 75 percent of the business’ employees mustenroll in MEIP, not including those employeeswho have insurance through another source.

MEIP offers a choice of up to four health plans,depending on the county in which the employeelives. All plans have restricted access to special-ists, with two offering some coverage of self-re-ferral services not approved by the designatedprimary care (gatekeeper) clinic. Employers mustpay at least 50 percent of the premium for singlecoverage but cannot pay more than 100 percent ofthe cost of the lowest priced plan in each marketarea. There is no requirement that employerscontribute to the cost of dependent coverage. Em-ployers must sign up for two years, but open en-rollment among health plans is held every year,and employees have unrestricted access to healthplans in their market area during open enrollment.Premiums are guaranteed for 12 months and poli-cies are sold through private insurance agents ordirectly through a MEIP agent. The MEIP pool isdesigned to be self-financed, with no public subsi-dy of premiums or administrative costs. MEIP be-gan taking applications in July 1993 and is justbeginning to enroll firms in the pool.

MEIP is a blend of two programs started earlierand operated by the same group of administratorsin the Department of Employee Relations. The


first is the State of Minnesota Group InsuranceProgram, which was described earlier in this sec-tion. The MEIP pool resembles this program inthat multiple health plans are offered to consumers(statewide, when available), and consumers facethe marginal cost of choosing more expensiveplans. The second previous program is the PublicEmployees Insurance Program (PEIP), a poolstarted for the employees of small governmentunits in Minnesota. PEIP also required employeesto sign two-year contracts with the pool and thePEIP pool has remained stable through its renewalperiods.

If the MEIP pool is successful, it has the poten-tial, along with the Minnesota Care (subsidizedpremium) pool, to dramatically affect the marketfor health plans and health care services in theTwin Cities. A substantial portion of medicalunderwriting costs are eliminated by purchasinginsurance through MEIP. Many people who pre-viously purchased individual and small group in-surance policies at high prices, and often foundthemselves in high-cost health plans with littlemanagement of care, may have the option to pur-chase lower cost coverage.

EMPLOYERS’ ASSOCIATION HEALTHCARE BUYERS COALITIONI Motivation for FormationThe Minnesota Employers’ Association (EA) is anonprofit association of approximately 1,300businesses that provides its members with varioustraining and management services (17). in late1991, the EA began to develop a health insurancepurchasing strategy in response to a survey of EAmembers that identified health care as a priorityarea for the provision of assistance by the EA.Members were experiencing yearly double-digitincreases in their health insurance premiums thatshowed few signs of abating. Working with a con-sulting firm, the EA developed a joint purchasingstrategy that resulted in the implementation of In-novation, a health insurance program made avail-able to members beginning on January 1, 1993,through a contract with the Prudential InsuranceCompany.

| MembershipIn late 1991, the EA held a series of informationalmeetings for its members in which the outlines ofthe joint purchasing arrangement were explainedand membership in a Health Care Buyers Coali-tion was solicited. By June 1993, a total of 363companies representing about 160,000 em-ployees and dependents had joined the coalitioneffort. These companies, which were mostly smallto medium-sized, contributed between $300 to$600 each to cover the startup costs of the coali-tion. About 325 companies supplied data thatwere made available to insurers wishing to be of-fered to coalition members. Of these 325 firms,between 90 and 100 chose to offer the Innovationinsurance product to their employees during thefirst year of the program, with about 90 percent ofthese located in the Twin Cities area. First year en-rollment in Innovation consisted of 5,000 em-ployee enrollees and their dependents. By thethird year of the program, it is expected there willbe 10,000 enrollees in Innovation.

I Approach to Joint PurchasingIn designing its contracting approach, the coali-tion concluded that managed care itself was notsufficient to control costs. It identified the prob-lem as a lack of competition among managed careplans on the basis of both price and quality (17).A strategy was developed that would pool the pur-chasing power of coalition employers, increasethe amount of information available to providers,payers, and consumers about health care out-comes and qualiity of care, and provide consumerswith a financial incentive to act on the informa-tion. The coalition viewed continuous quality im-provement as the process by which cost increasescould be restrained.

In the spring of 1992, a meeting was held withprovider representatives to assess their interest inbidding for a contract to serve the coalition. A co-alition committee was formed to develop a Re-quest for Proposals. The RFP was issued in thesummer of 1993 and responses were receivedfrom three organizations. (The scope of non-metropolitan provider coverage requested of


bidders ultimately limited the number of orga-nizations able to respond to the RFP.) After hear-ing presentations from each bidder, a coalitionsteering committee selected Prudential. Pruden-tial offered premium increases guaranteed to be 10percent or less a year over a three-year period, aswell as access to data desired by the coalition.

Innovation is designed as a preferred providermodel health plan. The preferred provider net-work in the Twin Cities consists of 865 primarycare physicians, 1600 specialists, 18 urgent carefacilities, 27 hospitals, and a large number of areapharmacies. If an enrollee seeks care within thenetwork, all care is coordinated by a primary carephysician and preventive care is provided withoutany copayment on the part of the enrollee. The pri-mary care physician authorizes all specialty care.Enrollees who seek care from non-network pro-viders must pay a higher share of the costs and alsoface a deductible.

Innovation network providers are involved indeveloping clinical “pathways” that contain crite-ria for physicians to use when making decisionsrelating to certain types of medical treatment. Thecoalition intends to develop data on outcomes,costs. and patient satisfaction for pathway condi-tions so that ● ’continuous quality improvement”can be achieved over time. A “quality council,”consisting of physician, employer, consumer, andPrudential representatives, and chaired by EAprofessional staff, meets regularly to discuss path-way development and other quality-related issues.Also during the first year of Innovation, an arrayof wellness progams and health education mate-rials were made available to enrollees.

A four-tiered premium structure was estab-lished for Innovation, based on previous expendi-tures and/or the demographics of each group.During the three-year contract period, an employ-er can move to a less expensive tier, if that is justi-fied, but cannot move to a more expensive tier.Prudential is responsible for marketing the In-novation health plan, but employers who are pro-spective purchasers frequently indicate an interestin dealing directly with EA staff. Therefore, theEA has a staff person who assists with marketing,

often accompanying the Prudential representativewhen an initial presentation is made to a firm.

1 Outcomes to DatePrudential guaranteed it would not increase ratesby more than 10 percent during any year under thethree-year contract. At the end of the first year itincreased rates by 8 percent. The rates paid byfirms offering Innovation averaged 14 percent be-low the rates paid by the same companies for 1991and 1992.

Initial data on enrollee satisfaction and patientoutcomes are now being examined by the coali-tion’s board, but no analyses of these data havebeen published as yet.

SUMMARYThe restructured State of Minnesota Group Insur-ance Program has increased the level of competi-tion among health plans in the Twin Cities andreduced the costs of care for its enrollees. How-ever, its effects on the reconfiguration of the healthcare delivery system in the Twin Cities are un-clear. The largest coalition formed by private em-ployers the Business Health Care Action Group(BHCAG) appears to be playing a more direct rolein restructuring health care delivery. BHCAG wasformed both as a mechanism to purchase health in-surance in a more effective manner and as a wayto leverage purchasing power to effect change inthe health care delivery system. As stated by onehealth plan CEO, “We believed that this was justthe beginning of a massive concentration of en-rollees and felt that if we were not responsive atthe front end we would be left out .“ In contrast, theemphasis of the State of Minnesota Group Insur-ance Program was on creating and maintainingcompetition among health plans at the enrolleelevel.

In summary, the organization of the demandside of the market was both a response to healthplan strategies and an important factor in shapingthose strategies. First, some of the larger HMOsthat offered IPA or point-of-service programspressed employers to designate them as the sole


HMO offering. The justification for this was thepotential for adverse selection if multiple planswere offered. In return for being designated the ex-clusive HMO, these plans promised to makeavailable multiple options, with one option beinga more restrictive set of providers similar to a staffor network HMO. This accelerated the trend to-ward multi-option plans among all HMOs. Toachieve the levels of access and overall capacityneeded to compete in this arena, some of theHMOs began to explore mergers. The formationof the BHCAG accelerated this process, since

BHCAG contractual requirements includedbroad-based geographical coverage for providersunder the contract. This, coupled with proposalsfor the development of Integrated Service Net-works (ISNS) in Minnesota, with antitrust exemp-tions to facilitate that development, and nationalhealth care reform proposals, caused all of the ma-jor health care provider organizations in the TwinCities to begin to explore ways to link hospitals,physicians, and health insurance plans moreclosely.

Relevanceof the

MinnesotaExperience 6

This background paper has described the evolution of thehealth care system in the Minneapolis/St. Paul metropoli-tan area and has summarized the evidence regarding itsperformance. It has relied on published accounts of the de-

velopment of that system, published empirical analyses of behav-ior and outcomes, and information collected through recentinterviews with community informants. As such, it has many ofthe characteristics of the classic “case study,” including some ofthe well–known limitations of this approach (7). The strength of acase study approach is that it can provide an in-depth understand-ing of how a community’s delivery system evolves over time,identify the key events in that evolution, and describe the rolesplayed by specific actors or organizations (30). However, casestudies have several generic limitations that can restrict their use-fulness. One limitation is that particular readers of a case studymay find it does not provide enough detail on the issues that are ofprimary interest to them. For instance, hospital administratorsmay find that the case study lacks depth in the discussion of hospi-tal motivations and roles, or employers may not find sufficient de-tail to inform them about specific actions taken by Twin Cities’employers to stimulate system change. This limitation is largelyunavoidable. No case study can provide enough information tosatisfy the interests of all potential readers without becoming socomplex that it obscures the essential components of the story.Case studies can, however, set the stage for further, more focusedanalyses of issues that are of particular interest to different stake-holders by providing a useful overview of events and how they areinterrelated.

I 49


Perhaps a more important limitation relates tothe generalizability of case study findings. Thedangers of generalizing from the findings of asingle case study have been discussed at length inthe evaluation literature (e.g. Wilson, 1979 (68)).Presumably, generalization is less risky when thecomparison environment is similar in its charac-teristics to the case study setting. However, it isnot always clear which characteristics are relevantin determining degree of environmental similarityor dissimilarity. For instance, is the health caremarket in Chicago similar to the Twin Cities be-cause both contain multiple HMOs, or is it dissim-ilar because the population of Chicago is muchlarger and the employer community in Chicago ismore fragmented? Obviously, determining theimplications of the Twin Cities’ experience for thenation as a whole is even more complicated thanassessing its relevance to a single other metropoli-tan area.

IMPLICATIONS FOR MANAGEDCOMPETITION REFORMS

While the need to be cognizant of and sensitive tothe dangers of generalizing from the Twin Cities’experience is obvious, there may be elements ofthat experience that do have implications forhealth care reform on a national scale. The remain-der of this section briefly highlights and discussesseveral tentative conclusions suggested by theevolution and performance of the health care de-livery system in the Twin Cities.

Development of managed competition is likelyto be associated With reconfiguration of commu-nity hospitals, such as the creation of multihospi -tal systems.

During the 1970s and 1980s there was a reduc-tion in the number of hospital beds nationally.During this period, hospital capacity in the TwinCities declined even more dramatically and hascontinued to decline in the 1990s. As recently asJanuary 1994, a large hospital system in the TwinCities announced the closure of one of its hospi-tals for financial reasons. The increased enroll-ment in HMOs during the 1980s has been cited byhospital administrators in the Twin Cities as one

of the reasons for reductions in community hospi-tal capacity, both because inpatient use has de-clined with increasing HMO enrollment andbecause hospitals have been forced to containcosts by reducing capacity in order to offer price–competitive contracts to HMOs.

While the specific contribution of increasedHMO enrollment to reductions in the number ofhospital beds in the Twin Cities is difficult to de-termine in any rigorous way, the chain of eventsappears reasonably clear. Managed competition asstructured in the Twin Cities first reduced demandfor inpatient hospital services and then createdprice competition among hospitals for the patientsof managed care plans. Hospitals responded byconsolidating their operations into a relativelysmall number of systems in order to negotiatemore effectively with HMOs and to facilitate theclosure or conversion of individual facilities tore-duce acute care capacity. Interview respondentsassociated with hospitals strongly believed thatthe reduction in acute care beds would continue,possibly resulting in a decline of over 50 percentin the next decade. Several respondents noted thathospital utilization would probably fall to about200 days per 1,000 population within the next five

years. From a strategic standpoint, survival in thisenvironment has increasingly been viewed byhospitals as dependent on the establishment ofstrong linkages with managed care organizationsor group practices, through merger or long–termcontracts. Consequently, while the reconfigura-tion of the hospital system in the Twin Citieslargely focused on the formation and merger ofhospital systems in the 1970s and early 1980s, therestructuring that is now occurring has shifted tothe vertical integration of hospitals, physicians,and insurance plans.

Managed care organizations will respond com-petitively to even moderately–sized purchasingcoalitions, for example, by merging to providegreater geographic access.

In the Twin Cities, the Business Health CareAction Group (BHCAG), the State of Minnesota(group Insurance Program, and the EmployersAssociation’s Buyers Coalition all appear to have

— —


influenced the delivery of health services to theirenrollees. The BHCAG precipitated the merger oftwo large HMOs and stimulated collaborationamong several provider groups, including theMayo Clinic, in the development of practiceguidelines. The Buyers Coalition has negotiated along–term contractual relationship with a majorinsurer, instituting a total quality management ap-proach and limits on premium increases. TheState of Minnesota Group Insurance Program hasmanaged a multiple health plan benefit offeringwith a fixed dollar contribution tied to the lowestpriced health plan, and recently has benefitedfrom declining increases in health plan premiums.

Organization of the demand side of the healthcare market under managed competition is likelyto encourage the consolidation of providers andmanaged care plans, suggesting that specific pub-lic and provider sector strategies maybe neededto maintain a competitive market structure.

Organizing the demand side of the health caremarket often entails offering consumers discretechoices among standardized health care coverageswith the consumer bearing the additional costassociated with the more expensive option. A‘“sponsor” or purchasing alliance aggregates pur-chasing power and manages the processes of en-rolling individuals into health plans andcontracting with health plans. This organizationof demand is intended to create pressure on healthplans to control their premium increases. Recent-ly, it appears to have been successful in the case ofthe State of Minnesota Group Insurance Program.

The Twin Cities’ experience suggests that pro-viders will respond to greater organization on thedemand side with greater aggregation of supply.When the demand side of the market is organized,health plans have the potential to secure largernumbers of enrollees under each contract. Theircontrol (actual or potential) over larger numbersof patients gives them greater leverage in contract-ing with providers. Providers quite naturally re-spond by consolidating to counterbalance thenegotiating power of health plans and/or by affili-ating with plans through mergers or long–termcontracts. In the absence of antitrust actions (or,

when state anti-trust policy facilitates consolida-tion), and with the encouragement of buyers’ co-alitions that value broad geographic coveragefrom contracting provider networks, the consoli-dation of the supply side of the market can occurrelatively quickly, as it has recently in the TwinCities.

The consolidation of the supply side of thehealth care market could benefit consumers inseveral ways. It creates the potential for the reduc-tion of excess capacity and the achievement of ef-ficiencies in service delivery. At least some of thegains from these effciencies may be captured bypayers and consumers in the form of the lower pre-miums, if buyers coalitions are able to use theirbargaining power effectively in negotiations withhealth plans. These coalitions may also be able touse their bargaining power to achieve improve-ments in the quality of care, and to effect changesin the way care is delivered. Ultimately, whileconsumers may have more restricted choicesamong health plans and fewer options in theirbenefit coverages, these drawbacks may be offsetby the gains they experience due to improvedquality and/or lower prices resulting from the ef-forts of the buyers’ coalitions. This is thought tobe particularly true for small firms, where it maynot be feasible to offer employees multiple healthplans under any circumstances.

The buyers* coalitions in the Twin Cities areaware that consolidation of the supply side of thehealth care delivery system poses risks in thelonger term. Specifically, unless entry of newhealth plans into the market remains feasible, co-alitions may become ● *locked into” their existingplan offerings and find their leverage in negoti-ations with these plans diminishes over time. Inthe Twin Cities, the existence of multiple buyers’coalitions helps to reduce the likelihood this willhappen. Also, as described previously, at the pres-ent time the affiliations among providers andhealth plans are not exclusive. This permits someflexibility in the market that could allow develop-ment of new, competing plans through realign-ment of provider groups. However, the generalissue remains an important one with far–reaching


implications. Feldman, using an econometricmodel to predict impacts, estimated that the bene-fits consumers receive from the purchase of healthinsurance could decline by 4 to 5 percent in theTwin Cities as a result of the Group Health/Med-Centers merger (20): While Feldman acknowl-edges several limitations in his approach, hisanalysis nevertheless highlights the difficultpolicy issues raised by consolidation of the supplyside of the health care system (20).

Feldman argues that aggressive enforcement ofantitrust laws is needed to ensure that mergersamong health care organizations benefit consum-ers and are not anti-competitive (20). However,some interest groups in the Twin Cities havecalled for very different initiatives in response torecent supply–side consolidation activity. Seniorcitizen advocates have characterized the consoli-dation as evidence that the ‘*managed competi-tion” approach to health care reform in the TwinCities is not tenable, and that stronger regulationof providers under a “single–payer” approach isneeded. Clearly, “managed competition” reformefforts will need to develop an explic it strategy forcreating and maintaining a competitive structureon the supply side of the market through publicpolicy (e.g., antitrust), or through the manage-ment policies of purchasing coalitions and largeprivate purchasers. Without such a strategy, the ef-fectiveness of managed competition will be opento question and its political viability in the long-run as an acceptable approach to health care re-form will be threatened.

CONCLUDING COMMENTSPrevious studies of the Minneapolis/St. Paulhealth care market (e.g. see Anderson, et al., 1985(6)) have noted several characteristics that mayhave facilitated early support for the. HMO.modelof health care delivery. These include participa-tion of a substantial proportion of physicians ingroup practices, a relatively homogeneous cul-

ture, civic leadership provided by a small numberof large corporations, and entrepreneurial effortsof several HMO supporters. To the degree thatthese factors were important in the early stages ofHMO formation, they probably also contributedto the development of a mature HMO market morequickly in the Twin Cities than in most other met-ropolitan areas. This in turn contributed to theconsolidation of the hospital market, puttingbuilding blocks in place for the very rapid ag-gregation of providers and health plans that is nowoccurring. One clear stimulus to these recentevents—the actions of buyers’ coalitions—wouldseem replicable in other areas, as these coalitionsnow are commonplace in most large cities andquite active in many. However, the ability of pro-viders to respond quickly to attempts to organizedemand is likely to vary across communities, de-pending on existing configurations of local healthcare delivery systems. While it seems clear thatthe actions of purchasing coalitions can contributeto the restructuring of relationships among healthcare providers, they may take longer to have animpact in communities where providers lack sup-portive organizational structures and have littleexperience in managed care.

A second important caveat relates to the inevi-table difficulties encountered in attempting toidentify and describe the important features of ahealth care market when that market is in a periodof rapid transformation. While the description inthis report of market evolution in the Twin Citiescovers the period through the beginning of 1994,continuing change appears likely, at least in thenear term. There is the possibility that the observa-tions contained in the report have been distortedby the swiftness of the change that is now takingp lto attempts to learn from most other health caremarkets during the current period of restructuringthat has been stimulated by state and nationalhealth care reform initiatives.

— .—.—— .————

Appendix A:Acknowledgments A

OTA wishes to thank the individuals and organizations listed below for their assistance with this back-ground paper. These individuals and organizations do not necessarily approve, disapprove, or endorsethis background paper OTA assumes full responsibility for the background paper and the accuracy of itscontent.

Arthur CapIan Catherine M. Kunkle Pamela B. PeeleUniversity of Minnesota National Business Coalition Virginia Polytechnic Institute

Minneapolis, M N Washington, DC Blacksburg, VA

Steven R. Eastaugh Glenn Melnick Steve WetzellGeorge Washington University Rand Corporation Business Health Care Action GroupWashington, D C Santa Monica, CA Minneapolis, MN

Jon R. Gabel Stephen T. Mennemeyer Herbert S. WongKPMG Peat Marwick University of Alabama U.S. Department of HealthWashington, DC Birmingham, AL and Human Services

Rockville. MD

Steven Hill Stephen H. MilesUniversity of Wisconsin . . . . University of Minnesota Jack ZwanzigerMadison, WI Minneapolis, MN University of Rochester Medical

CenterRochester, NY

I 53

References

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11

54

American Hospital Association, HospitaZ Sza-Zistics (Chicago, IL: 1971-1990).American Hospital Association, Guide 10 theHea/th Care Field (Chicago, IL: 1972-1990).Anderson, G., Heyssel, R., and Dickler, R.,“Competition Versus Regulation: Its Effecton Hospitals,” HeaZ/h Afiairs 12(1):70-80,1993.

Anderson, H.J., “Choosing Sides in Twin Ci-ties,” Modern Healthcare 17(6):61 -66, 1987.Anderson, O., Aday L. A., and Kralewski, J.,‘-An Empirical Analysis of the Twin CitiesHMO Market,” University of Chicago, Cen-ter for Health Administration Studies, Chica-go, IL, July 1981.Anderson, O., Herold, T., Butler, B., et al.,HMO Development: Patterns and Prospec-lives (Chicago, IL: Pluribus Press, 1985).Campbell, D., “Degrees of Freedom and theCase Study,” Comparative Political Studies8(2): 178-193, 1975,Christianson, J., Do HA40s Stimulate Bene$-cial Competition?, unpublished monographprepared for Interstudy, Excelsior, MN, Apr.27, 1978.Christianson, J., Lurie, N., Finch, M., et al.,“Use of Community-Based Mental HealthPrograms by HMOS: Evidence from a Medic-aid Demonstration,” American Journal ofPublic Health 82(6):790-796, 1992.Christianson, J., and McClure, W., ‘competit-ion in the Delivery of Medical Care,” NewEngland Journal of A4edicine, 301( 15):812-818, 1979.Citizens League Research, “Minnesota Man-aged Care Review 1992,” Minneapolis, MN,August 1992.

12.

13.

14.

15.

16.

17.

18.

19.

poo

21.

Citizens League Research, “Minnesota Man-aged Care Review 1993,” Minneapolis, MN,September 1993.Dowd, B.E., “HMOS and Twin Cities Admis-sion Rates,” Health Services Research 21 (2):177-188, 1986.Dowd, B. E., and Feldman, R., ● ’Biased Selec-tion in Twin Cities Health Plans,” Advancesin Health Economics: Volume 6, R. Schefflerand L. Rossiter (eds.) (Greenwich, CT: JAIPress, 1985).Dowd, B. E., and Feldman, R., “Insurer Com-petition and Protection from Risk Redefini-tion in the Individual and Small GroupMarket,” Inquiry 29(2):148-157, 1992.Dowd, B. E., Johnson, A., and Madson, R. A.,“Inpatient Length-of-Stay in Twin CitiesHealth Plans,” Medical Care 24(8):694-710,1986.

Ebert, T. E., and Morrow, G.T., “A Partnershipto Reduce Cost by Improving Quality,” A4edi-cal Interface, Apnl:43-45, 1993.Employee Benefit Research Institute,“Sources of Health Insurance and Characten-sitics of the Uninsured: Analysis of the March1993 Current Population Survey,” Washing-ton, DC, January 1994.Families USA Foundation, “Per Capita HealthExpenditures,” Washington, DC, 1993.Feldman, R., “The Welfare Economics of aHealth Plan Merger,” Journal of RegulatoryEconomics 6:67-86, 1994.Feldman, R., Chan, H., Kralewski, J., et al.,“Effects of HMOs on the Creation of Compet-itive Markets for Hospital Services,” Journalof Health Economics 9;207-222, 1990.

References | 55

22.

23.

24.

25.

26,

27.

28$

29

30.

31.

32.

Feldman, R., and Dowd, B. E., “Simulation ofa Health Insurance Market with AdverseSelection,” Operations Research 30(6):1027-1042, 1982.Feldman, R., and Dowd, B. E., “Is There aCompetitive M~ket for Hospital Services?”Journal of Health Economics 5(3):277-292,1986.Feldman, R., and Dowd, B. E., “The Effective-ness of Managed Competition: Results froma Natural Experiment,” Health Policy Re-form: Competition and Controls, R.B. Helms(cd.) (Washington, DC: The American Enter-prise Institute Press, 1993).Feldman, R., Dowd, B. E., Johnson, A., et al.,“The Competitive Impact of HMOS on Hos-pital Finances,” Journal of Health Policy,Politics and LuN’ 10(4):675-697, 1986.Feldman, R., Gifford, G. G., and Dowd, B. E.,“Effect of HMOS on Premiums in Employ-ment-Based Health Plans,” Health ServicesResearch 27(6):779-81 1, 1993.Feldman, R., Jensen, G., and Dowd, B. E.,‘*What are Employers Doing to Create a Com-petitive Market for Health Care in the TwinCities?” Contemporaq Policy Issues 3(2):69-88, 1984-85.Feldman, R., Kralewski, J., and Dowd, B. E.,“’HMOS: The Beginning or the End?” HealthServices Research 24(2): 191-211, 1989.Foster Higgins, ‘b 1992 Health Care BenefitsSurvey: Medical Plans,” Medical Benefits,Washington, DC, Mar. 30, 1993.Greene, D., and David, J. L., ● ’A Research De-sign for Generalizing from Multiple CaseStudies,” Evaluation and Program Planning7(1 ):73-85, 1984.Iglehart, J., “The Twin Cities Medical MarketPlace,” New England Journal of iWedicine31 1:343-348, 1984.Jackson-Beeck, M., and Kleinman, J. H., “Ev-idence for Self-selection Among HealthMaintenance Organization Enrollees,” Jour-nal of the American Medical Association250(20):2826-2829, 1983.

33.

34.

35.

36.

3’7.

38.

39.

40.

41,

42,

43.

Johnson, A.N., Dowd, B. E., Morris, N., et al.,● *Differences in Inpatient Resource Use byType of Health Plan,” lnqui~~ 26(3):388-398,1989.Klein, J., and Cooley, R., ● *Managed Competi-tion in Minnesota,” Managed Care Quarterly1(4):58-67, 1993.Kosterlitz, J., “’Monopoly Medicine?” Na-tional Journal 28: 1746-1749, 1993.Kralewski, J., Countryman, D., and Shatin,D., ● ’Patterns of Interorganizational Relation-ships Between Hospitals and H MOS,” Inquiry19(4):357-362, 1982.Kralewski, J., deVries, A., Dowd, B., et al.,“The Development of Integrated Service Net-works (ISNS) in Minnesota,” prepared for theMinnesotaCare Legislative Oversight Com-mittee, Minneapolis, MN, Feb. 21, 1994.Kralewski, J., Dowd, B. E., Feldman, R., et al.,“The Physician Rebellion,” New EnglandJournal of Medicine 316(6):339-342, 1987.Leichter, H. , “State Model: Minnesota,”Health Aflairs 12(2):48-58, 1993.Luft, H. S., Maerki, S. C., and Trauner, J. B.,“The Competitive Effects of Health Mainte-nance Organizations: Another Look at the Ev-idence from Hawaii, Rochester, andMinneapolis/St. Paul,” Journal of HealthPolitics, Policy and La~’ 10(4):625-657,1986.Lurie, N., Christianson, J., Finch, M., et al.,“The Effects of Cavitation on Health andFunctional Status of the Medicaid Elderly,”American College of Physicians 120(6):506-511, 1994.Lurie, N., Finch, M., and Dowd, B., “Who arethe Uninsured in Minnesota?: A Report to theMinnesota Health Care Access Commis-sion,” Institute for Health Serivces Research,University of Minnesota, Minneapolis, MN,October, 1990.Lurie, N., Moscovice, I., Finch, M., et al.,“Does Cavitation Affect the Health of theChronically Mentally 111: Results from a Ran-domized Trial,” Journal of the American


A4edical Association 267(24):3300-3304,1992.

44. Metropolitan Healthcare Council, “Report onTwin Cities Hospitals from the Council ofHospital Corporation’s Inpatient UtilizationData Base,” Minneapolis, MN, 1993.

45. Milliman and Robertson, Inc., “MedicalCosts by Geographical Area,” research byMilliman and Robertson, Inc., Brookfield,WI, 1992

46. Minneapolis Star and Tribune, “HMOS KeepTwin Cities Health-Coverage Costs Down,”Minneapolis Star and Tribune, Mar. 2, 1993.

47. Minneapolis Star and Tribune, “Minnesota’sBig Blues Aim to Adapt,” Minneapolis Starand Tribune, June 21, 1993.

48. Minneapolis Star and Tribune, “HealthPar-tners Will Merge with Ramsey HealthCare,”Minneapolis Star and Tribune, Sept. 17,1993.

49. Minnesota Department of Health, “1992HMO Operations in Minnesota, Volume II:Patterns in Health Care Utilization,” HeaZthA4aintenance Organization, Minnesota De-partment of Health, HMO Unit, Minneapolis,MN, 1993.

50. Minnesota Department of Health, “HealthCare Reform in Minnesota”, Media Back-grounder, Prepared by the Minnesota Depart-ment of Health and the Minnesota HealthCare Commission, Minneapolis, MN, Sep-tember 1993.

51. Minnesota Department of Health, “IntegratedService Networks and Regulated All-PayerSystem: Design and Implementation Plan,”discussion draft, Nov. 10, 1993.

52. Minnesota Department of Health, Health Sys-tems Development Division, Health Eco-nomics Program, “Providing Medical Care inRural Minnesota,” Minneapolis, MN, March1991.

53. Minnesota Department of Health, MinnesotaHealth Care Commission, “Universal Cover-age Report,” Minneapolis, MN, 1994.

54. Minnesota Hospital Association, ● ’Twin Ci-ties Hospital Occupancy Rate,” Health in-

formation Resources of llinnesota, St. Paul,MN, 1992.

55. Moscovice, I., Lurie, N., Christianson, J., etal., ‘*Access and Use of Health Services byChronically Mentally 111 Medicaid Beneficia-ries,” Health Care Financing Revie~ 14(4):75-87, 1993.

56. Pauly, M., ‘*Is Cream Skimming a Problem forthe Competitive Medical Market,” Journal ofHealth Economics 3(1):88-95, 1984.

57. Reece, Richard, “Information for Physiciansin Control,” Reece Report 2:4, 1987.

58. Shadle, M., and Christianson, J., “The Impactof HMO Development on Mental Health andChemical Dependency Services,” Hospitaland Communi/y Psychiat~, 40(11):11 45-1151, 1989.

59. St. Paul Pioneer Press Dispatch, “SpeedyHealthEast Rise Startles Some,” St. Paul Pio-neer Press Dispatch, Aug. 23, 1987.

60. St. Paul Pioneer Press Dispatch, “HealthCare’s Future,*’ St. Paul Pioneer Press Dis-patch, p.l E, 3E, Nov. 1, 1993.

61. St. Paul Pioneer Press Dispatch, “Medica,HealthSpan Will Merge,” St. Paul PioneerPress Dispatch, p. 1A, 6A, Dec. 8, 1993.

62. Sullivan, C., Miller, M., Feldman, R., et al.,“Employer-Sponsored Health Insurance in1991,” Health Aflairs 11(4):172-185, 1992.

63. University of Minnesota, Health Sciences,Health Sciences New’s, Special Edition, Jan.11, 1994.

64. U.S. Congress, OffIce of Technology Assess-ment, Health Care in Rural America,OTA-H-434 (Washington, DC: U.S. Gover-nment Printing Office, September 1990).

65. U.S. Department of Commerce, Bureau of theCensus, Statistical Abstract of the UnitedStates: /993, 113th edition (Washington, DC:U.S. Government Printing OffIce, 1993).

66. U.S. Department of Health and Human Ser-vices, Health Resources and Services Ad-ministration, Bureau of Health Professions,Seventh Report to the President and Congresson the Status of Health Personnel in the

References 157

United States (Washington, DC: U.S. Gov- Case Study Evaluations,” Evaluation Quar-emment Printing OffIce, March 1990). lerly 3(3):446-459, 1979.

67. Ware, J., Rogers, W., Davies, A., et al, “Com- 69. Wisner, C., “Health Plan Effects on the Healthparison of Health Outcomes at a Health Main- Status of Medicare Beneficiaries in the Twintenance Organization with Those of Fee- Cities,” unpublished Ph.D. dissertation. Uni-For-Service Care,” Luncet 8488: 1017-1072, versity of Minnesota, Minneapolis, MN,1986. 1992.

68. Wilson. S., “Explorations of the Usefulness of

managed care and competitive health care markets: the twin ... · care, in part in response to the...

Documents