THE NEW ENVIRONMENT FOR RURAL HMOS

Transcription

1 THE NEW ENVIRONMENT FOR RURAL HMOS by Jon B. Christianson, Maureen Shadle, Mary M. Hunter, Susan Hartwell, and Jeanne McGee Prologue: The common perception of rural America is that its pastoral environment is difficult to penetrate with new ideas and slow to change. Such perceptions, coupled with the sparse populations that live there and the relative absence of capital, have discour aged the developers of health maintenance organizations (HMOs) from creating alternative delivery plans in rural settings. In the following article, Jon Christianson of the University of Arizona and his colleagues question this bit of conventional wisdom. Indeed, the authors declare that many rural areas are experiencing dynamic changes in their health care delivery systems. In June 1984, 118 HMOs were serving rural areas in thirty-four states. That total represented a 50 percent increase from the seventynine HMOs in operation in rural areas in twenty-nine states in This substantial growth was a surprise to the authors leading them to examine what factors were responsible. Christianson, who holds a doctorate in economics from the University of Wisconsin, has been studying marketplace competition in health care for nearly a decade. For the last six years, Christianson has been on the faculty of the University of Arizona studying, among a number of issues, the states competitive bidding program for Medicaid. Maureen Shadle and Jeanne McGee are colleagues at Inter- Study, a Minneapolis-based health policy research organization. Shadle, project director for this study, is senior research analyst and McGee, who holds a Ph.D. in sociology from the University of Indiana, is a senior vice-president. Mary Hunter, who holds a master's degree in business from the Wharton School at the University of Pennsylvania, is a Minneapolis-based health care consultant. Susan Hartwell, who holds a master's degree in human service studies from Cornell University, is a service representative at Primary Care Network, a Minnesota HMO that serves markets outside of the Twin Cities.

2 106 HEALT H AFFAIRS Spring The passag e o f the Health Maintenanc e Organizatio n (HMO ) Ac t of created high expectations concerning the ability of HMOs to remedy many of the perceived problems of the existing health care delivery system, including the limited acces s to medical care experienced b y larg e numbers o f rural Americans. 1 Originally, 2 0 percent o f the Offic e o f Healt h Maintenanc e Organization' s (OHMO ) budge t fo r HMO developmen t wa s set asid e fo r nonmetropolita n areas. However, no mor e tha n 1 4 percent o f HMO gran t dollar s i n an y fisca l yea r wer e actually use d fo r thi s purpose. Between and 1979, forty-two rura l projects receive d gran t funding ; twenty-tw o o f thos e project s faile d t o develop further. 2 Rura l areas thus acquired a reputation a s unattractive locations for HMO development. The informatio n presente d i n Exhibit 1 suggests that thi s perceptio n may no w b e outdated. I n 1981, seventy-nine HMO s (3 3 percent o f al l HMOs) reporte d servin g a tota l o f 23 8 rura l countie s i n twenty-nin e states. By June 1984, 11 8 HMOs (o r 39 percent) were servin g 408 rura l counties in thirty-four states. Even among metropolitan counties, it is the smallest ones, wit h population s o f les s than 50,000, tha t sho w bi g in - creases in both number and percentage of counties served. In an attempt t o understand wh y the availabilit y o f HMOs i s increasing i n rura l areas, w e undertoo k intensiv e cas e studie s o f seve n rura l HMOs. Our cas e studies were guided by a relatively structured researc h design similar to approaches advocated i n the recent literature on qualitative research and employed by Stephen Shortell and colleagues in their evaluation of hospital-physician joint ventures. 3 Framework For Rural HMO Developmen t Through various efforts including literatur e reviews, discussions with HMO administrators, an d a surve y o f HMO s servin g rura l areas w e identified a relatively larg e number o f problems tha t coul d explai n th e lack of availability o f HMOs in rural areas in the past. The fou r genera l categories discussed below encompass most of these barriers. Acquiring financing. HMOs, particularly of the group practice or staf f model, can require substantia l development capita l and severa l years of operational experienc e befor e the y reac h a financial break-eve n point. Locating financing fo r this period i s obviously an issue for all HMOs, but our literatur e revie w suggest s tha t i t ha s bee n a critica l on e fo r rura l HMOs. Multiple factors contribute to this situation. Group practices ar e This article was based on research sponsored by the Office of Health Maintenance Organization (OHMO) through a contract with InterStudy. The support and encouragement provided for this project by Robert Hamel of OHMO was greatly appreciated. Since the article summarizes findings of a larger study, many details concerning hypotheses, analytic methods, and interpretation of findings are omitted. Any comments or requests for further information should be addressed to Maureen Shadle, project director.

3 RURAL HMOS 10 7 Exhibit 1 Number And Percent Of Counties With HMO Providers : County size Rural <10K 10K-24K 25K-49K 50K+ Total Number of counties And 1984 Percent of counties Percent change in counties Metro <50K K-99K K-149K K-249K K-499K K Total Note: Counties are classified accordin g to population reported in the 1980 U.S. Census. Counties served by HMOs are identified i n the National HMO Census (1981 and 1984) conducted b y InterStudy, Excelsior, Minnesota; counties in the Ne w England state s of Connecticut, Massachusetts, Maine, Ne w Hampshire, Rhod e Island, an d Vermont are omitted. Percent changes are calculated using 1981 as the base. less common in rural areas ; where present, the y typicall y consist of relatively few physicians, and many hav e limite d financia l resource s to devote to HMO development. Also, the financial condition of rural hospitals varies greatly, making them an undependable sourc e of HMO develop - ment capital. Urban-base d investor s o f all types hav e generall y foun d rural HMO development to be an unattractive venture, possibly because of the limited enrollment potential of rural HMOs and the apparent high failure rate of rural HMO development efforts. During th e years immediatel y followin g passag e o f the HMO Act, federal gran t an d loan fund s wer e availabl e t o assist rura l HMO s in meeting preoperationa l expense s an d covering initia l losses. However, this funding wa s not used to the extent anticipated; consequently it was later drastically reduced and finally eliminated. Therefore, we identified the acquisition of start-up financing a s an important issue for rural HMOs, and particularly for those HMOs established since Overcoming the opposition of rural providers. Until quite recently, HMOs hav e bee n oppose d by most physicians. 4 Whil e thi s oppositio n has gradually softened in many urban areas, we expected that traditiona l physician antipath y toward s prepai d medicin e woul d continu e t o be relatively stron g in rural area s for several reasons. First, the physicians who practice in rural areas may locate there because they prefer to practice independently ; certainl y man y rura l physician s hav e develope d reputations as "rugged individualists," and may be philosophically opposed +71

4 108 HEALT H AFFAIRS Spring to HMO practice. Moreover, providers who have practiced medicine fo r some tim e i n a rural settin g ma y have had littl e contact with o r knowl - edge o f prepaid medicin e an d therefor e ma y hold misconception s con - cerning th e requirement s o f HM O participation. Finally, w e surmise d that physicians in rural areas may feel competitive pressures for patient s less strongly than their urban colleagues. While the number o f specialists in rural areas has been increasing over the past decade (as has the acces s to medical care for many rural residents), most rural areas continue to be less intensivel y serve d b y physician s tha n urba n areas. 5 Fo r thes e rea - sons, a second focu s o f our analysi s was to learn ho w rura l HMOs ha d secured physicia n suppor t fo r thei r development an d growt h o r had a t least minimized the impact of physician opposition. Achieving financially viabl e enrollmen t levels. Buildin g an d main - taining sufficien t membershi p t o become a financially viabl e HMO wa s expected t o be a more complex task in rural than in urban area s from a t least tw o perspectives. Mos t obvious, th e populatio n i s limited i n rura l areas. Clearly, th e rura l HM O mus t appea l t o a comparativel y broa d cross-section of the population in order to achieve any sizable enrollment level. Secondly, maintaining enrollment level s in rural HMOs was expected to be complicated by the HMO's relatively high visibility as a community organization. Any individual's perception o f poor medical care or inadequate coverag e in a rural HMO would likel y reach, through forma l an d informal channels, a large r proportio n o f th e communit y (an d thu s o f potential HMO enrollees) than would be the case in urban areas. Whether accurate o r not, thes e negative perceptions could result in a poor imag e that woul d b e difficul t fo r th e HM O t o counterac t throug h its subse - quent marketin g efforts. Thu s a third focu s o f our stud y examine d th e ways in which rural HMO marketing efforts differe d fro m those of urban HMOs and th e specifi c methods developed b y rural HMOs for creatin g and maintaining positive relationships with the rural community. Containing costs and rate increases. Our review identified th e limited supply o f physician s i n a rura l HMO' s servic e are a a s a majo r facto r hindering the abilit y o f the HMO t o contain its costs. In rural areas, the historical competitor s fo r th e HM O hav e bee n Blu e Cross/Blu e Shiel d and indemnit y insurers. Rura l physician s typicall y provid e car e t o pa - tients covered under various options, including a rural HMO when available. Thus if the HMO is not price competitive, and the patient therefor e chooses a traditional insuranc e plan, physician s d o no t necessaril y los e their patients. Because the rural HMO appear s to have a limited abilit y to generate ne w patient s fo r rura l physicians, there ma y be littl e incen - tive for thos e physicians to accept financial ris k sharing or strong utiliza - tion revie w procedure s a s cost-containmen t device s an d condition s o f participation in the HMO.

5 RURAL HMOS 10 9 The situation with respect to rural hospitals was expected t o be muc h the same. The monopoly status of many rural hospitals frequently place s them i n a relativel y stron g positio n t o negotiat e wit h th e HM O ove r reimbursement an d othe r matters. Furthermore, th e HM O require s a viable rural hospital for its enrollees who require basic inpatient services. Thus i t may not be i n the HMO's interest t o seek discounted payment s and ris k sharin g fo r inpatien t car e i f suc h arrangement s contribut e t o financial instability for the hospital. While rural HMOs appear to be more limited in their ability to contain costs throug h financia l incentive s o r discounte d payment s tha n urba n HMOs, they nevertheless experience some cost-containment pressure. In order t o attrac t enrollees, rura l HMO s mus t b e pric e competitiv e wit h insurance plans. They also must keep premiums at levels that are accessible t o rura l resident s wh o typicall y hav e lowe r income s an d smalle r health benefit contribution s from employer s than many potential urba n HMO enrollees. Therefore, th e fourt h objectiv e o f ou r stud y wa s t o analyze how rural HMOs contain costs. Selection Of Rural HMOs For Analysis Rural HMO s wer e selecte d fo r intensiv e cas e stud y analysi s usin g a three-stage procedure. First, InterStudy HMO Census data were analyzed to determine whic h HMO s serve d rura l areas. These HMO s wer e the n grouped according to whether they were based in rural or urban areas. A major objectiv e in selecting the case study sample was to obtain approximately equa l representatio n fro m thes e tw o categories, unde r th e as - sumption that they may represent fundamentally differen t approache s to the provisio n o f HMO service s in rural areas. I t was als o intended tha t the analysis sample provide diversity along dimensions such as geographic location, size, sponsorship, operationa l date, an d model. The composi - tion of the sample was also influenced b y an assessment o f the coopera - tion that each HMO would provide to the research investigators, since it was essential that participating HMOs cooperate full y wit h the data collection process. Finally, th e numbe r o f HMO s chose n was, o f course, constrained by the research budget. Of the seven HMOs chosen for the study, four are based in rural areas, and three are urban-based HMO s that expanded int o rural areas. Three of the fou r rural-base d HMO s wer e formed durin g th e 1970s, and on e became operationa l i n January Of th e three urban-base d HMOs, one ha s serve d rura l area s sinc e it s formation i n 1975, while th e othe r two firs t expande d t o rural area s in and Within th e limita - tions of this study, considerable geographic dispersion was achieved. The HMOs wer e locate d i n northern California, souther n Colorado, south - western North Dakota, northern Minnesota, Wisconsin (two), and upstate

6 110 HEALT H AFFAIRS Spring New York. The uppe r Midwes t wa s overrepresented i n th e sample, i n part due to the high degree of HMO activity in that area. No rural HMOs were selecte d fro m th e South, where HM O developmen t ha s generall y lagged behin d th e res t o f th e nation. Exhibi t 2 show s th e penetratio n levels throughout th e country. Overall rural penetration i s just under 2 percent, compared t o the enrollment in the South which is slightly more than 1 percent. Overall U.S. HMO penetration wa s 9.7 percent i n June Exhibit 2 Estimates Of Rural HMO Enrollmen t And Penetration By Region, Jun e Northeast New England Middle Atlantic Midwest East North Central West North Central South South Atlantic East South Central West South Central West Mountain Pacific Totals Estimated rural HMO enrollment 98,481 49,701 48, , ,688 53,740 56,336 41,152 11,237 3,947 83,148 13,026 70, ,393 Rural population in counties served by HMOs 6,243,428 4,974,899 1,268,529 14,205,456 10,134,843 4,070,613 4,197,819 2,167,952 1,129, ,210 3,011,754 1,519,628 1,492,126 27,658, Estimated rural penetration (% ) 1.6% 1.0% 3.8% 1.6% 1.8% 1.3% 1.3% 1.9% 1.0%.4% 2.8%.9% 4.7% 1.7% Note: Rura l HM O enrollmen t figure s ar e fro m InterStudy' s Nationa l HM O Censu s survey ; the y ar e base d o n estimates supplie d b y HMO s o f the percentag e o f thei r member s tha t receiv e primar y car e service s from th e HMO i n areas with n o central cit y larger than 25,000. Twenty-seven o f the 11 8 HMOs tha t reported servin g rural counties di d not supply these rural membership estimates. In addition t o representing differen t region s o f the United States, th e study HMOs also represent various degrees of "ruralness." The Colorad o and North Dakota HMOs ar e located in areas with very low populatio n densities an d ar e fa r remove d fro m majo r metropolita n areas. I n con - trast, the rural areas served b y the Californi a an d Ne w York HMOs ar e in proximity t o relatively larg e metropolitan area s (Sacramento an d Al - bany, respectively). The HMO s i n our stud y hav e a variety o f sponsors, includin g grou p practices, community organizations, insurers, a medical care foundation, a hospita l cooperative, an d a hospital. Thre e o f th e fou r rural-base d HMOs ar e relatively smal l and ha d enrollment s o f under 10,00 0 a t th e time interview s wer e conducted. Th e fourt h i s an extraordinaril y larg e rural HMO with an enrollment o f over 60,000. The urban-based HMO s in our study all have total enrollments of greater than 50,000, with vary-

7 RURAL HMOS 11 1 ing proportions of total HMO membership drawn from rural areas. Two of the four rural-based HMOs and tw o of the three urban-base d plans in the study are individual practice associations (IPAs). Two of the HMOs ar e networks ; tha t is, the y ar e grou p practic e base d bu t hav e elements o f a n IP A sinc e the y contrac t wit h primar y car e physician s outside o f their groups to provide service s to HMO enrollees. The stud y sample also includes one staff model HMO. While th e seve n HMO s var y i n thei r characteristic s (summarize d i n Exhibit 3), the y d o no t represen t a rando m sampl e o f rura l HMOs. Therefore, ou r finding s shoul d b e viewe d primaril y a s interpretive an d hypothesis-generating. I n this respect, their usefulness i s similar to those attributed by Shortell et al. in their case studies. 6 Exhibit 3 HMO Characteristic s Location Years operational Sponsorship Model Profit status Enrollment (June 1984 ) Rural-based Greater Marshfiel d Community Healt h Plan North centra l Wisconsin 13 Clinic/ Hospital/ Insurer Network Nonprofit 63,993 HMO Health Plan s HMO of Wisconsin Southern Colorado Southern Wisconsin 9 1 Community Rural hospital cooperative IPA IPA Nonprofit Nonprofit 7,825 8,300 West River HM O Southwest North Dakot a 5 Clinic Network Nonprofit 6,248 Urban-based Capital Area Community Health Plan Upstate New York 6 Community Staff Nonprofit 51,552 (-10,000) Foundation Health Plan HMO Minnesot a Northern California Northern Minnesota 6 9 Medical care foundation Insurer IPA IPA For profi t Nonprofit 83,196 (-2,000) 72,000 (-25,000) Notet Fo r th e urban-base d HMOs, th e enrollmen t figure s i n parenthese s represen t approximat e enrollmen t wit h participating rura l providers. HM O model s ar e define d a s follows : Staff an HM O tha t deliver s service s throug h a physician group that i s controlled by the HMO; Group an HMO that contracts with one independent group practice to provide services ; Network an HM O tha t contract s wit h tw o o r mor e independen t grou p practice s (an d fe w sol o practices) t o provid e services ; IPA a n HM O tha t contract s wit h a n associatio n o f physician s fro m variou s practic e settings (predominantly fro m sol o practices) to provide services. Findings: The New Environment For Rural HMOs In this section, we summarize the findings o f our study with respect to each of the four issue s identified i n the study framework. Fo r each issue,

8 112 HEALT H AFFAIRS Spring we describ e th e experience s o f th e HMO s a s the y relat e t o tha t issue, discuss an y recen t change s i n th e rura l environmen t tha t ma y be rele - vant, an d examin e th e implication s fo r futur e HM O developmen t i n rural areas. Acquiring financing. I t is clear from ou r analysi s of this issue that th e recent demis e o f federa l fundin g fo r HM O developmen t coul d hav e a critical influenc e o n th e wa y i n which HMO car e becomes availabl e i n rural areas in the future. Specifically, i t will become much more difficult, although no t impossible, fo r "indigenous' ' rura l HMO s t o form, an d much more likely that the exposure o f rural areas to prepaid health car e will be the result of the expansion of urban-based plans. Of th e fou r rural-base d HMOs, one (Greate r Marshfiel d Communit y Health Plan ) was formed prio r t o the passag e o f the federa l HM O Act, and consequentl y receive d n o federal funds. Becaus e i t is based on on e of the fe w larg e group practice s i n rural areas, its start-up experienc e i s likely to be o f very limited relevance to the development o f other rural - based HMOs. The tw o rural-based HMO s that began i n the lat e 1970 s received federal suppor t i n the form both o f feasibility stud y grants an d of loan s t o cove r initia l operatin g losses. Curren t manager s expresse d doubt tha t thei r HMO s coul d hav e succeeded, o r indeed, woul d hav e been attempted, withou t suc h financia l assistance. Bot h HMO s ar e lo - cated in sparsely populated areas characterized by agricultural economies and limite d pe r capit a incomes. Therefore, neithe r communit y woul d likely be regarded a s a good opportunity fo r private HMO investors, nor would eithe r on e ten d t o be viewe d a s an attractiv e expansio n site fo r urban-based HMOs. With the elimination of OHMO funding fo r rural HMO development, the formatio n o f tru e rural-base d HMO s wil l clearl y b e a muc h mor e difficult undertakin g financiall y tha n i t was during the 1970s. However, the experience o f HMO o f Wisconsin, which was capitalized by a group of rural hospitals, suggests that in some environments rural HMO development can be successfully accomplishe d with private funding. Two key factors facilitate d th e acquisitio n o f funds fo r HM O o f Wisconsin. First, there wa s a perception o n the par t o f rural hospital s that nearb y urba n HMOs wer e beginning t o attract enrollee s fro m rura l area s and the n t o use urba n hospital s fo r thos e rura l enrollees. A rural-base d HM O wa s identified a s on e metho d o f protectin g th e patien t bas e o f th e rura l hospitals sinc e enrollee s could b e directe d t o these hospitals. Second, a mechanism existe d t o facilitat e a cooperative fundin g effor t sinc e rura l hospitals i n th e are a alread y participate d i n a hospita l cooperativ e t o share services and engage in joint purchasing activities. The financing requirements for expansion of urban-based HMOs into rural area s appear t o depend o n th e structur e o f the HMO an d o n th e nature of the expansion attempted. The Capital Area Community Healt h

9 RURAL HMOS 11 3 Plan (CACHP ) mad e substantia l us e o f a variet y o f fundin g sources, including federal grant and loan monies, in its expansion efforts. CACH P is a staf f mode l HM O wit h salarie d physicians. Therefore, expansio n required th e purchase, construction, o r leas e o f rural facilities, an d th e addition o f salarie d staff. Th e initia l cos t associate d wit h thi s typ e o f expansion clearly is greater than that needed for the expansion of an IPA or network model HMO. The othe r tw o urban-based HMO s i n our stud y ar e IP A models tha t expanded b y securin g the participatio n o f physicians alread y practicin g in the targeted rural areas. In both cases, a rural expansion represented a marginal additio n t o the organization' s existin g responsibilities, an d th e initial financin g o f thos e expansion s di d no t requir e a majo r outla y o f funds. It i s ou r conclusio n tha t th e expansio n o f urban-base d IPAs, o r o f group practice s that employ an IPA or "group network" model for thei r rural sites, will become the most common means for expanding prepai d care i n rura l areas. Thi s metho d minimize s th e financia l requirement s associated with the development and initial operation of a rural site, and it can be accomplished i n a relatively short period o f time. The develop - ment o f rura l staf f o r grou p practic e mode l HMO s fro m scratc h ha s become less probable, in our view, since the withdrawal of federal fund - ing to cover the cost of feasibility studies, planning, and initial operating losses. The creation o f rural-based IPA s does appear t o be a feasible alterna - tive t o urban-base d expansion, provide d tha t sufficien t financia l com - mitment can be secured from participatin g providers and that the HM O can b e lodge d withi n a n existin g organization. Th e degre e o f financia l support offere d b y rural provider s fo r thi s effor t i s likely to depend o n the extent to which they experience or anticipate competition fo r their patients from urban-based HMOs. Overcoming the opposition of rural providers. Our case studies indicate tha t fiv e o f th e seve n rura l HMO s di d fac e initia l oppositio n o f various kinds from loca l physicians. This opposition was manifested pri - marily in public statements of concern about quality of care and "social - ized medicine," rather tha n i n overt action s to inhibit HMO formation. The involvement o f the federal government i n providing feasibility an d other start-u p monie s an d th e us e o f Nationa l Healt h Servic e Corp s (NHSC) physician s by som e o f the HMO s sometime s were regarded a s evidence that the rural HMO represented a precursor t o a public secto r delivery syste m for rura l areas. However, i n the judgment o f HMO ad - ministrators, much of the philosophical opposition expressed publicly by rural physicians actually reflected a lack of knowledge about HMOs an d a very understandable fear of the HMO's potential impact on local physician incomes and practice styles.

10 114 HEALT H AFFAIRS Spring The HMOs in our study adopted various strategies to gain the cooper - ation, or at least to reduce the hostility, of rural physicians. One genera l strategy involved emphasizing the potential benefits to the community of HMO development. West River HMO pointed out that Blue Cross/Blue Shield rate s for th e community wer e based o n stat e averages ; therefor e community residents did not benefit from the cost-effective styl e of medicine practice d i n tha t community. A n HM O woul d allo w communit y residents t o pa y les s for thei r health car e while freein g the m fro m con - cerns about out-of-pocket expenses. HMO Health Plans emphasized the potential of the HMO a s a vehicle for attractin g federa l communit y healt h cente r fund s t o th e are a an d dispersing them among local providers. This funding woul d increase the access of many community residents to care while reducing the bad debts incurred b y physicians and hospitals. Other HMO s emphasized th e potential fo r improve d acces s o f are a resident s t o medica l car e an d fo r alleviation of perceived physician shortages. A secon d genera l strategy involved variou s attempts to diffuse physi - cian an d hospita l concern s abou t th e impac t o f a n HM O o n thei r in - comes. Al l o f th e HMO s attempte d t o refe r withi n thei r ow n servic e areas as much a s possible, thus retaining income for local specialists and hospitals. At the same time, most of the HMOs initially did not attemp t to disrup t th e traditiona l out-of-are a referra l pattern s o f participatin g physicians. Even though this policy clearly was costly, it was important in securing physician participation since it demonstrated that the HMO was not interferin g i n the practic e o f medicine. I n the cas e o f HMO Healt h Plans (where loca l physician oppositio n wa s perhaps th e mos t intense), HMO physician s wer e initiall y allowe d t o participat e o n a discounte d fee-for-service basi s so as to minimize their financial risk. HMO Minnesot a an d HM O o f Wisconsin received th e greatest initia l cooperation fro m loca l physicians. In both cases, there was the threat o f expansion int o th e communit y b y a n HM O base d elsewhere. Thi s ex - pansion wa s considere d undesirabl e b y rura l provider s becaus e the y believed tha t enrollee s i n thes e HMO s woul d b e referre d ou t o f thei r communities t o urban-base d specialist s an d hospital s a t a greate r rat e than wa s alread y occurring. Wit h thi s consideratio n i n mind, th e rura l physicians chos e t o contrac t a s a group wit h a n existin g HM O (HM O Minnesota) o r t o participate i n the formation o f a new HMO (HM O o f Wisconsin). I n eac h instance, participatio n wa s structure d t o provid e considerable local physician involvement in decision making and contro l over referral patterns. One unexpecte d findin g o f our stud y wa s the degre e t o which rura l physicians no w perceiv e urban-base d HMO s t o b e potentia l competi - tors. Apparently, rural physicians are sensitive to the increased competi - tion fo r patient s amon g physician s in urban areas, and the y do not fee l

11 RURAL HMOS 11 5 insulated fro m th e effect s o f that competition. Th e rural-base d HMO s that experienced initia l opposition from physician s now report consider - able physician support, in part because they offer a rural-based alternativ e for patient s intereste d i n enrollin g i n a prepai d pla n an d als o brin g a rural perspective to plan management. The urban-based HMO s servin g rural area s i n ou r stud y hav e trie d t o b e sensitiv e t o thes e physicia n concerns. Their expansion strategie s typically combine decentralized de - cision making in the delivery of medical services with centralized admin - istration functions suc h as claims and enrollment processing, management information systems, and marketing. Achieving financially viable enrollment levels. In our framework, w e identified tw o genera l issue s relating t o th e enrollmen t o f member s b y rural HMOs : th e potentia l effec t o f a negativ e communit y imag e o n enrollment an d th e constraint s o n enrollment s impose d b y populatio n size and composition. Virtually al l of the HMOs reaffirmed ou r expectation tha t developin g a positiv e communit y imag e i n rural areas, and maintainin g tha t imag e over time, wa s crucia l t o buildin g enrollment. Durin g th e proces s o f HMO development, tow n meeting s wer e ofte n hel d durin g whic h th e concept wa s explaine d an d potentia l communit y benefit s wer e identi - fied. Influential community residents were recruited t o speak out in support of the HMO or serve on the HMO's board. As the HMO progresse d t o it s operational stage, staf f an d marketin g representatives typicall y were hired fro m th e local population to solidif y the HMO' s imag e a s a n organizatio n wit h stron g loca l ties. Insuranc e agents from th e community sometimes were solicited to assist in marketing th e HMO. Thi s approac h gaine d acces s t o th e agents ' networ k o f clients and funneled som e commission revenue t o prominent loca l business people. Once operational, many o f the HMOs continued t o devote considera - ble attentio n t o the maintenanc e o f the suppor t o f community leaders. To avoid antagonizing influential community leaders, the federally qualified HMO s almos t neve r utilize d th e mandat e whe n approachin g em - ployers. Instead, federal qualification wa s employed in a symbolic way to reinforce th e legitimac y o f the HMO a s an organization worth y o f com - munity trust. In additio n t o thes e genera l approaches, severa l HMO s develope d specific strategie s that prove d effectiv e i n their ow n environments. Fo r example, th e Greate r Marshfiel d Communit y Healt h Pla n an d HM O Minnesota both regard their strong affiliation s wit h the Blue Cross/Blu e Shield organization s i n thei r respectiv e state s a s helpful i n establishin g their legitimacy in rural areas. The Greater Marshfield Communit y Healt h Plan and the West River HMO found tha t sponsorship by clinics already established in their communities was extremely valuable in gaining com-

12 116 HEALT H AFFAIRS Spring munity acceptance, a s the prestig e o f the clinic s transferred t o the affili - ated HMOs. While a positive communit y imag e enhance d th e potentia l fo r HM O enrollment growth, the relatively smal l populations o f most o f the rura l HMO site s in our stud y impose d natura l limitation s on enrollmen t po - tential. In order to expand their enrollments, all of the HMOs develope d strategies aime d a t securin g a relativel y larg e marke t penetration. Be - cause of the limited number o f sizable private employers in their servic e areas, mos t o f th e rura l HMO s sough t t o enrol l self-pa y individuals, Medicare beneficiaries, and/o r Medicai d recipient s a s well as employe d groups. The HMOs' main concern with respect to individual enrollmen t was that only people with severe medical conditions would find i t worthwhile t o pa y the entir e HMO premiu m (whic h tende d t o be highe r fo r individual coverag e tha n premium s o f competin g insuranc e plan s tha t offered fewe r benefits ) i n orde r t o receiv e th e comprehensiv e benefit s offered b y the HMO. To guard against this possibility, three of the HMOs employed medical underwriting for individual enrollees, with as many as 50 percen t o f applicant s denie d membershi p becaus e o f thei r presen t physical condition o r medical history. While thi s approach riske d creat - ing a negativ e communit y imag e fo r th e HMO, i t wa s regarde d a s a necessary step from a financial perspective. All o f th e rura l HMO s offere d Medicar e supplementa l policie s an d saw the m a s valuabl e i n attractin g enrollees. However, the y wer e fa r more cautiou s abou t enterin g int o contract s (eithe r risk - o r cost-based ) with th e Healt h Car e Financin g Administratio n (HCFA ) t o provid e al l required medical services to Medicare eligibles. The negative experienc e of th e Greate r Marshfiel d Communit y Healt h Plan, whic h incurre d a substantial debt while participating in a HCFA HMO capitation demon - stration, appears to have raised concerns among other rural HMOs. The rura l HMO s experience d mixe d result s i n servin g indigen t pa - tients unde r contrac t wit h stat e governments. HM O Healt h Plan s an d HMO Minnesot a reporte d satisfactor y experience s wit h prepai d con - tracts fo r indigents, whil e th e Greate r Marshfiel d Communit y Healt h Plan withdre w fro m Wisconsin' s Medicai d capitatio n progra m whe n i t could not reach agreement with the state on an acceptable rate. All of the HMOs expressed reservations about becoming overly dependent on th e enrollment o f indigents ; the y feare d tha t futur e stat e effort s t o contro l program cost s could constrai n capitatio n rate s t o th e poin t wher e suc h contracts would become a financial drain on the HMO. A further strateg y for expanding enrollment within the HMO's natural service are a involve d th e developmen t o f multipl e benefit s packages, including a t leas t on e requirin g significan t deductible s and/o r copay - ments. This wa s deemed necessar y b y th e HMO s i n order t o compet e effectively fo r enrollees in groups with limited health insurance coverage.

13 RURAL HMOS 11 7 Also, th e lowe r premium s o f HMOs wit h cost-sharin g provision s ofte n succeed in attracting younger, healthier individual enrollees. Those HMO s tha t wer e federall y qualifie d foun d th e benefi t restric - tions accompanying federal qualification t o be a hindrance in competing for enrollees. Specifically, require d HM O benefit s ofte n excee d thos e offered b y competing insurance plans. Combined wit h the requiremen t of basing premiums on a community rate, benefit level s sometimes caused rural HM O premium s t o be unattractivel y hig h compare d t o thei r in - demnity competitors. The rura l HM O marketin g strategie s describe d abov e ar e als o em - ployed by many urban HMOs, although perhap s not to the degree em - phasized b y rura l HM O managers. Tw o caution s wit h respec t t o enrollment wer e identifie d tha t d o see m specifi c t o rura l areas. First, many rural communities ar e served b y only one physician, or by two to three physician s i n a joint practic e arrangement. I f these physician s decline to participate in the HMO, then the entire community i s effectivel y lost t o th e HM O a s a potential sourc e o f enrollee s unles s th e HM O i s willing to place a physician into the community. Thus the maintenanc e of effective provide r relations can be a very important facto r i n securin g HMO enrollment. I f participating physicians are pleased with the HMO, they can ac t as influential marketin g representatives with their patients. If they are dissatisfied, the n the HMO will simply not sound attractive to community resident s who must rel y on these providers fo r th e majorit y of their care. A secon d cautio n wit h respec t t o physicia n participatio n an d enroll - ment wa s voice d b y manager s o f tw o plan s that, i n thei r earl y years, utilized NHSC physician s to provide car e to HMO enrollees. There i s a relatively hig h turnover rat e amon g thes e physician s due t o the limite d nature of their contract, and such turnover can result in enrollee dissatisfaction. Also, other physicians may resent the presence of NHSC provid - ers and ma y withdraw fro m th e HMO, takin g their patients with them. Both occurrences can have a negative impact on enrollment. Yet another approach to increasing the enrollment of rural HMOs is to expand th e HMO's servic e area. We found tha t al l of the HMOs in ou r case studies had implemented, o r were considering, programs o f expansion. Thre e o f th e rural-base d HMO s ha d expande d int o othe r rura l areas, while a fourth wa s studying the opportunities fo r expansion o f its service area. In addition, rural-based HMO Health Plans had expande d into an urban are a (Pueblo, Colorado) and was drawing most of its new members from thi s location. The urban-base d HMO s had al l expanded the rural portions o f their servic e areas recently an d wer e planning fur - ther rural expansions in the future. The aggressiv e expansio n activitie s o f rural HMOs constitute d a second unexpecte d findin g o f this study. Whereas in the past, rural HMO s

14 118 HEALT H AFFAIRS Spring typically struggle d t o gai n provide r an d communit y acceptanc e an d achieve financia l viability, the y no w ar e emphasizin g growt h throug h contractual arrangement s wit h provider s i n other rural, an d sometime s urban, communities. Ther e appea r t o b e thre e reason s fo r th e ne w emphasis o n geographi c growth, althoug h no t ever y reaso n applie s t o each HMO studied. First, many of the HMOs believe that they have reached, or will shortly reach, the upper limits of feasible enrollment in their immediate areas. If they are to continue to grow as organizations, they must do so by expanding their servic e areas through contract s with other providers. For thre e of the rural-based HMOs, this additional growth is viewed as desirable to enhance th e financia l stabilit y o f th e organizations. I n al l cases, enroll - ment growt h appear s t o be a n importan t interna l organizationa l objec - tive and a measure of successful performance. Second, th e climat e fo r servic e are a expansio n i s becoming mor e fa - vorable. If in the past rural HMOs were regarded by rural providers with distrust, skepticism, o r eve n ope n antagonism ; no w th e rura l HM O i s increasingly viewed a s a mechanism by which providers can retain thei r patients and maintain some control over the financial and medical aspects of their practice. Rural-based HMO s i n particular ar e see n a s desirabl e alternatives to the intrusions of nearby urban HMOs and hospitals. Therefore, ther e no w appea r t o be enhance d opportunitie s fo r expansio n o n the par t o f rura l HMOs. Suc h opportunitie s ma y als o b e availabl e t o urban-based HMO s that offe r decentralize d approache s t o the manage - ment of affiliated providers. Third, competition for providers between rural-based and urban-base d organizations has led some rural-based HMOs to accelerate their expan - sion plans. The y hav e hastene d t o secur e affiliation s wit h othe r rura l physicians befor e thos e physician s becom e partie s t o agreement s wit h urban-based group practice HMOs or statewide IPAs. While servic e are a expansio n i s a majo r objectiv e o f th e HMO s w e studied, thi s expansio n i s not bein g undertake n withou t som e misgiv - ings. In particular, the rural-based HMOs expressed concern that expansion may dilute efforts t o serve the needs of their immediate communitie s and may stretch their financial and managerial resources to the breakin g point. The manager s o f these HMO s worr y abou t thei r abilit y t o introduce greate r cost-consciousnes s int o th e behavio r o f provider s i n rela - tively distant communitie s an d t o create a sense o f shared commitmen t to organizationa l objectives. Nevertheless, despit e thes e concerns, the y view the imperatives for expansion as too strong to resist at this time. Constraining costs and rate increases. We expected that rural HMOs would be concerned about constraining costs for two reasons. First, rural consumers, who were expected t o bear a larger share of premium costs, typically hav e lowe r income s tha n thei r urba n counterparts. Therefor e

15 RURAL HMOS 11 9 we anticipated tha t thes e consumers would be relatively pric e sensitive. Second, th e comprehensiv e benefit s o f rura l HMO s wer e expecte d t o make it relatively more difficult fo r these HMOs to compete on the basis of price with traditional insurance plans, especially sinc e such insuranc e plans are more likely to feature high deductibles and coinsurance in rural areas. In our rural HMO sit e visits, we found considerabl e suppor t fo r thes e expectations. More significantly, all of the rural HMOs in our study cited as a third sourc e o f cos t restraint th e competitio n fro m othe r operatin g and developing HMOs. HMO managers identified th e growing competition fo r patient s a s a major chang e i n their environment, a change tha t allowed them to take actions in the interest of cost control that would not previously have been possible. In particular, suc h competition increase d the willingness of rural providers in IPA-model HMOs to accept a larger share o f the HMO's financial ris k and t o permit relatively stron g utiliza - tion review measures. In three o f the IPA-model plans in our study, th e physician groups accept capitated payments that place participating physicians a t variou s degree s o f financia l risk. Som e o f thes e contractua l arrangements are quite complex. Contractual relationships with rural hospitals now appear to be evolving in a similar manner. Excep t where hospita l rates are constrained b y state regulation s o r b y th e stron g monopol y positio n o f a communit y hospital, rural hospitals now share financial ris k with community HMO s through primar y sponsorshi p (HM O o f Wisconsin), capitated payment s (HMO of Wisconsin, Greater Marshfield Community Health Plan, HMO Minnesota), and/or negotiate d per diem payments (HMO Health Plans, Foundation Health Plan). Where risk sharing does not occur at present, it is now unde r negotiation. Som e urba n referra l hospital s ar e als o reim - bursed at per diem rates by rural HMOs; however, reimbursement base d on discounted charges appears to be more typical. Rural hospitals hope, through risk-sharing arrangements, to secure and retain more HMO patients either in competition with other rural hospitals or as protection against the siphoning of f o f patients by urban medi - cal centers. Urba n referra l hospital s ar e becomin g mor e aggressiv e i n bidding for rural HMO patients, not only to increase their present census but als o to secure an advantage over other urban competitors in attract - ing future rura l patients. In effect, thes e referral hospitals are betting o n the future enrollment growth of rural HMOs. Perhaps because rural providers are assuming greater financial risk for HMO performance, the y appea r mor e willin g to participate i n compre - hensive utilizatio n review. Most o f the rural HMOs studied rel y heavil y on physician education, assisted by detailed profiles of physician practice patterns and hospita l expenses, to alter the practic e patterns o f individ - ual physicians. I n man y cases, participatin g physician s ha d previousl y

16 120 HEALT H AFFAIRS Spring practiced i n relative isolation from thei r colleagues and were unaware of the extent to which their practice patterns deviated from the norm. In addition to education efforts, virtuall y all of the rural HMOs in our study employed some form of concurrent review of hospitalized patients. However, i t wa s muc h les s commo n fo r rura l HMO s t o requir e tha t physicians obtai n authorizatio n fro m th e pla n prio r t o admittin g a pa - tient to the hospital. The tw o cases where formal preadmission certification was required involve d urban-based HMO s that had expanded int o rural area s by contracting wit h rural physicians. In these instances, preadmission certification fo r nonemergency hospitalizations was utilized i n the HMOs' urban servic e areas as standard procedure. HMOs that were based i n rura l area s appeare d t o rel y mor e heavil y o n educatio n an d moral suasio n t o alte r physicia n hospitalizatio n patterns. I t i s possibl e that this apparent preference reflecte d a greater concern for the potential adverse impac t tha t withdrawa l o f physician s fro m participatio n coul d have on the overall enrollment in the rural-based HMOs. Summary The difficult y o f acquirin g th e requisit e initia l financin g ha s consti - tuted a historical impediment t o the development o f rural HMOs. Th e recent elimination o f federal funds fo r HMO development wil l certainly make it more difficult, althoug h not impossible, for HMOs indigenous to rural areas to form. As a result, it seems probable that the future growt h of prepaid car e in rural area s will occur largely a s a result o f the expan - sion o f existin g rural - o r urban-base d plans, rathe r tha n throug h th e formation o f small rural-based organizations. Other past obstacles to the development o f HMOs in rural areas seem to be diminishing in importance. Continuing increases in the number o f urban HMOs, in HMO enrollment, and in physician supply have created a perception on the part of rural physicians and hospitals that they mus t now compet e mor e aggressivel y with eac h other an d wit h urban hospi - tals and HMOs for patients. This new sense of competitive vulnerabilit y is overwhelming the historical reluctance of rural providers to participate in an HMO. As rural hospitals and physician s search for mechanisms t o retain patients in their own communities, HMO participatio n i s seen by these providers a s one way to reduce, or at least control, the flow of selfreferrals t o urba n area s an d therefor e t o protec t thei r incomes. Rura l providers als o appea r muc h mor e willin g t o accep t bot h financia l ris k sharing and more comprehensive utilization review procedures as conditions of participation in HMOs. Urban-based HMOs that have successfull y expanded int o rural area s have recognized thi s change i n environment ; their decentralize d approache s t o the organizatio n an d managemen t o f rural expansion site s have been tailored to address the concerns of rural

17 RURAL HMOS 12 1 providers. HMOs based in rural communities have taken advantage of changing provider attitudes to overcome the limited enrollment potential in their sparsely populate d catchmen t areas. The y hav e attempte d t o expan d their servic e area s through contractua l relationships with hospitals an d physicians in other rural or urban communities. These efforts ar e being pursued vigorousl y a t present, both to take advantage o f the increase d receptivity o f rura l provider s t o HMO s an d becaus e competitor s ar e believed to be undertaking similar expansion. In our view, these developments suggest that the availability of HMO services in rural areas will continue to increase, and possibly accelerate, in the coming years. NOTES 1. J. L. Dorsey, "Th e Healt h Maintenanc e Organizatio n Ac t o f 1973, " Medical Care 1 3 (1975): 1-9; J. L. Falkson, HMOs and the Politics of Health System Reform (Chicago: American Hospital Association, 1980). 2. Esselsty n Foundation, Repor t of the Conference on Extending Prepayment to Rural Areas (Claverack, N.Y., 29 September-1 October 1978). 3. D. Green e an d J. L. David, "A Researc h Desig n fo r Generalizin g fro m Multipl e Cas e Studies," Evaluation and Program Planning 7 (1984): ; W. A. Fireston e an d R. E. Herriott, "The Formalization o f Qualitative Research : An Adaptation o f 'Soft' Scienc e to the Polic y World," Evaluation Review 7 (1983): ; S. M. Shortell, T. M. Wickizer, and J. R. C. Wheeler, Hospital-Physician Joint Ventures: Results and Lessons From a National Demonstration in Primary Care (Ann Arbor: Health Administration Press, 1984): J. L. Falkson, HMO s and Politics; P. Starr, "Th e Undelivered Healt h Car e System, " The Public Interest 4 2 (1976) : 66-85; P. Starr, "Th e Triump h o f Accomodation : Th e Ris e of Private Health Plans in America, , " Journal of Health Politics, Policy, and Law 7 (1982): W. B. Schwartz, J. P. Newhouse, B. W. Bennett, an d A. P. Williams, "Th e Changin g Geographic Distribution of Board Certified Physicians, " The New England Journal of Medicine 309 (1983): Shortell, et al., Hospital-Physician Joint Ventures. 7. J. K. Iglehart, "HMO s (For-Profi t an d Not-For-Profit ) o n th e Move, " The New England Journal of Medicine 310 (1984):