Printer Friendly

Factors related to the provision of hospital discounts for HMO inpatients.

Using 1986 AHA hospital survey data, we analyzed hospital-HMO contract provisions, hospital operating characteristics, and market conditions for a national sample of 801 hospitals with HMO contracts to determine the factors related to provision of a discount and the magnitude of the discount if present. Seventy-eight percent of the hospitals reported that at least one of their HMO contracts provided a discount for inpatient services. Risk-sharing provisions, the number of hospitals within a five-mile radius, the proportion of the populations enrolled in HMOs, and the number of HMOs operating in the metropolitan statistical area (MSA) were directly related to provision of discounts. Public hospitals were less likely than other facilities to provide discounts. For the magnitude of the discounts, risk-sharing provisions and the number of hospitals within a five-mile radius were again related, as was the number of HMOs operating in the MSA - but his time the number-of-HMOs variable had an inverse relationship. The results suggest that increased HMO market activity does result in price competition for hospital services but that hospital discounting strategies are extremely complex and may not follow conventional market theories. Hospitals appear to be using contracts both to stabilize their relationship with HMOs and increase market share, and they are increasingly giving discounts to achieve those ends.

Much has been written about the role of health maintenance organizations (HMOs) in the development of competitive markets for health care services (Christianson and McClure 1979; Enthoven 1978; Goldberg and Greenberg 1980). HMOs provided a stated range of services for a defined enrolled population for a fixed per capita premium and, as such, they have a direct financial incentive to reduce input costs (Luft 1981). This is often achieved by gaining price concessions from providers such as hospitals, allegedly through creating competitive markets for those services.

The emphasis on competition as a cost-containment strategy in the public and private sectors consequently has caused a rapid growth of HMOs both in number and size. In 1980, there were 236 plans providing services to 9.1 million members. By the end of 1984, 366 HMOs were serving 16.7 million people, and in 1987, 662 plans were in operation with a total enrollment of over 28.5 million. This represented a membership increase of 20.8 percent during the last half of 1986 and the first half of 1987 (Gruber, Shadle, and Polich 1988). More recently, the rate of growth has slowed somewhat but enrollment has continued to increase steadily, with 32.6 million individuals enrolled in HMOs by the end of 1988 (Gold and Hodges 1989).

In an earlier study on hospital-HMO relationships, one of the authors found that HMOs often select a hospital for strategic reasons, concentrate patients at that hospital, and then work to obtain price concessions (Kralewski et al. 1983). A more recent study (Feldman et al. 1990) found that, while HMOs are obtaining substantial discounts from hospitals for inpatient services, the dynamics of these relationships are not understood well at all. Physicians in staff-network HMOs appear to be willing to change hospital allegiances in order to protect the HMO's economic base (Kralewski et al. 1991). On the other hand, independent practitioner association (IPA) HMOs generally do not rely on patient concentrations as a means of obtaining price concessions.

The effect of HMOs in promoting a competitive market for hospital services is also unclear. In a study of the Minneapolis-St. Paul metropolitan area, which is thought to be highly competitive, Feldman and Dowd (1986) found that the price elasticity of demand for administration was - 1.124 for a group of payers that included HMOs. They concluded that hospitals charging higher prices to HMOs would lose some patient volume but not much. A more recent study of HMOs in four cities (Feldman, Chan, Kralewski, et al. 1990) found that the price sensitivity of demand varies by HMO model type. The elasticity of demand for admissions by staff-network plans was - 3.044. In other words, a hospital would gain 3 percent more admissions from a staff-network plan if price was decreased by 1 percent. This is contrast to a less price-sensitive response by IPA plans.

The research to date has dealt primarily with HMOs and the development of their contracts with hospitals. Little research has focused on discounts from the hospitals' perspective. Specific reasons why some hospitals choose to contract with HMOs while others do not have not been investigated.

With the emphasis being placed on cost containment by virtually every purchaser of health care services, hospitals are no longer able to engage in cost shifting from one payer group to another or simply to raise prices to recoup revenues lost by declining inpatient admissions and lengths of stay. Moreover, since hospitals are a major cost center in the health care field, they have become primary targets for attempts to control health care expenditures. Since hospitals are experiencing significant decreases in demand due to these pressures, it seems reasonable to expect them to respond favorably to overtures from HMOs that would increase their patient days by inviting them to sell empty beds at marginal costs.

The degree to which HMO are able to take advantage of these conditions to create competitive markets for hospitals services is part of an important policy issue. This issue centers on the potential use of price competition in the health care system as a means to improve the effectiveness and efficiency of the system. It is argued that HMOs, by concentrating patients at particular hospitals and exerting pressure on those hospitals to reduce prices and costs, hold the promise of improved cost control without government intervention.

Establishment of a national policy promoting competitive markets for health care services and promotion of the HMO as an agent to help develop these markets have made it fundamental to reach a better understanding than we have thus far of the various factors that influence a hospital's response to HMO overtures. What are the conditions that lead hospitals to negotiate discounts for HMO patients? Does competition among hospitals within a community affect HMO negotiations? These research questions are important to this issue.

Building on our previous work, this study examines the nature of contracts between HMOs and hospitals, from the hospitals' perspective, in an attempt to identify the factors causing hospitals to negotiate discounts with HMOs and the conditions determining the magnitude of those discounts. The study was conducted using a resource dependence model. This model is based on the premise that organizational survival is dependent on the acquisition of appropriate and sufficient resources from the environment (Pferrer and Salancik 1978; Thompson 1967). In the case of hospitals, these would include physicians, capital resources, and patients.

This model also proposes that since organizations do not control the resources they need, they attempt to stabilize existing supply lines and develop predictable new ones as much as possible, particularly if the resources are scarce or the available sources of supply are unreliable. Entering into exchange relationships with other organizations is one method of securing needed resources or ensuring resource stability - but at the price of losing some power or control to these other organizations.

A hospital's choice of whether or not to contract with an HMO is expected, therefore, to be based on (1) the hospital's expectation that direct benefits will result from resource exchange with the HMO, (2) the degree to which the hospital needs additional resources, and (3) the hospital's tolerance for interdependence with external organizations. The benefits may include increased patient volume, a reduced number of bad debts, reduction in administrative costs, and the development of referral networks. Need for additional resources might be represented by low occupancy rates and high costs, while scarcity of resources (i.e., competitiveness of the market) may relate to increasing numbers of hospitals and HMOs operating in the same market.

Hospitals lacking the ability to obtain the necessary resources from the external environment in an explicitly autonomous manner would hypothetically, then, be most likely to enter into HMO contracts. Those hospitals with alternative means of attracting and maintaining resources - and less likely to provide discounts (or other concessions) to HMOs - might be expected to be (1) teaching hospitals, which often function as regional referral centers or providers of services unavailable in community hospitals; (2) members of a multi-hospital system, able to obtain resources as well as bargaining power because of their part in the system; and (3) hospitals that assure resource supply lines by themselves owning an HMO. These hospitals can be expected to be both less dependent on local hospital market conditions and less likely to provide discounts. A hospital's tolerance for interdependence with external organizations may be increased by contracts that provide risk-sharing provisions or guarantees, or both.

Accordingly, the more important an environmental resource is to a hospital's survival, and the greater its scarcity, the more likely it is that the hospital will attempt to obtain the resource by sacrificing some institutional autonomy and control - by entering into a contract with an HMO that provides a discount to the HMO. To explore this premise, we focused on the following working hypotheses:

1. Hospitals with HMO contracts that have risk-sharing provisions will tend to offer discounts for those patients. 2. Hospitals with HMO contracts that specify volume or revenue guarantees will tend to offer discounts for those patients. 3. Hospitals with an equity position in an HMO will be less likely to offer a discount to that HMO. 4. Hospitals with lower occupancy rates will tend to offer discounts to HMOs. 5. Hospitals with higher expenses per hospital day will tend to offer discounts to HMOs. 6. Hospitals that are members of a multihospital system will be less likely to offer discounts to HMOs. 7. Hospitals that have a mission of teaching and tertiary care will be less likely to offer discounts to HMOs. 8. Hospitals operating in environments served by a higher number of hospitals will tend to offer discounts for HMO patients. 9. Hospitals located in geographic areas with high numbers of operating HMOs and a large proportion of the population enrolled in HMOs will tend to offer discounts for HMO patients. 10. The magnitude of the discounts offered by hospitals in HMO contracts will be higher as (a) risk-sharing provisions are included, (b) volume or revenue guarantees are negotiated, (c) the occupancy rates becomes lower, (d) expenses per day increase, and (e) hospital competition in the service area increases. The magnitude will be lower in contracts with (a) hospitals that have an equity position in the HMO, (b) hospitals that are members of multihospital systems, and (c) teaching hospitals.

Source of Data

This study is based on data supplied by the American Hospital Association (AHA), from its 1986 Survey of Hospital Affiliations with Health Maintenance Organizations; on the hospital market density data described by Luft, Robinson, Garnick, et al. (1986); and on 1985 HMO market conditions by metropolitan statistical areas (MSAs) (Christianson et al. 1991).

The AHA survey was sent to 1,760 AHA-registered community hospitals located in the United States that had indicated in a 1985 survey that they provided care to members of an HMO. The AHA received 1,046 returned surveys for a return rate of 59.4 percent. Of these respondents, 868 or 83 percent had a formal relationship with an HMO and 801 hospitals in this group provided data on the reimbursement arrangement in at least one HMO contract. The descriptive analysis is based on these 801 hospitals.

The AHA survey questionnaire allocated space for information for up to six hospital-HMO contracts but gave no directions on how they should be ordered. Also, many of the hospitals did not answer the questions regarding the number of admissions, inpatient days, or revenues that were accounted for by each HMO, making it impossible to rank the HMOs by contributed volume of hospital activities for the entire sample of hospitals. Thus, the first regression analysis uses information from the hospital-HMO contract listed first, based on the assumption that the first HMO listed was the most important. In 720 of the 801 cases, reimbursement information was provided for the first-listed hospital-HMO contract. Thirty-five of these cases were dropped because of missing data in the independent variables (32 cases) or data that were present but obviously in error (3 cases); this left a sample of 685 hospitals for the first analysis, to use in determining the factors related to the provision of discounts. One hundred thirty-four of these hospital-HMO contracts were reported to have provided a discount that consisted of a percentage off full rates and to have listed the percentage amounts, which were used to analyze the factors related to the magnitude of the discount.

A subsequent analysis was done using the largest reported hospital-HMO contract with determination of size based on the proportion of total admissions attributed to each HMO. Cases were used even if the hospital did not report on all contracts, so long as some information on admissions, revenues, or inpatient days could be attributed to an HMO thus enabling an index of the HMO's volume as a proportion of the hospital's total activities to be defined. Of the 801 cases, 401 gave no information on HMO volume for any HMO; this left 400 cases. Of these cases, 18 were excluded because they lacked information on reimbursement for the largest reported HMO (2 cases), had other missing independent variables (13 cases), or had obvious errors in reported information (3 cases), leaving a sample of 382 for which some HMO volume information was known. For cases in which the proportion of total admissions attributed to an HMO was not available (60 cases), the variable was imputed from the proportion of total revenues that was attributed to the HMO (54 cases) or the proportion of total inpatient days that was attributed to it (6 cases) using weighted coefficients obtained from regression analyses. Revenue information was used first because it correlated more highly (.6942) with HMO admissions than did inpatient days (.6108). Thus, this analysis to determine the factors related to the provision of a discount to HMOs is limited to the subset of hospital-HMO contracts, and includes a volume variable. (It should be noted that the first hospital-HMO contract was the only one or the largest reported one in 80 percent of the cases where volume information was reported; this supports the assumption that the first-listed hospital-HMO contract is an important one.) Of this subset, 81 cases gave a discount - a percentage off full rates - and reported the percentage amount; these were the cases used to determine the factors related to the magnitude of the discount.

The data set for hospital market density was complete for all 801 hospitals that provided some reimbursement arrangement information for at least one HMO contract. The HMO market conditions data set contained information for MSAs only. Eighty-seven of the 801 hospitals were not located in a MSA; however, 66 of these were located in counties bordering one or more MSAs. We believe that the true values of these missing HMO variables would best estimated by the values of these variables in the adjacent MSAs. Accordingly, the HMO data for the adjacent MSA (or average values if the county bordered more than one) were used for these hospitals. Interaction terms were used in the regression analysis to evaluate this procedure. For the 21 "rural" hospitals, little information was available to indicate what the true values should be for HMO market condition variables except for our theory that they would likely be lower than values for the typical urban hospital. The HMO values were set to zero, and a dummy variable was used to assess the appropriateness of this assumption.

The variables include in the analysis are listed in Table 1. For the purpose of this study we have broadly defined discount to include every pricing arrangement at less than list price.[1] Risk sharing, a separate and identifiable feature of contracts, may or may not be associated with a discount. The dependent variables - the presence or absence of risk-sharing provisions and volume and/or revenue guarantees; hospital equity position in the HMOs; and the proportion of the hospital's admissions related to each HMO - are hospital-HMO contract-specific variables. Most of other, independent variables are hospital-specific. The two variables that measure the number of hospitals located in the market area are used as proxies for the degree of competitiveness, with the assumption that the more hospitals in the immediate geographic area the more competitive the market. The number of HMOs and the HMO saturation in each MSA also are used as proxies for the degrees of market competitiveness, with a direct relationship hypothesized.[2]


Data Analysis

The data analyzed in three stages. First, the hospitals in the survey sample are described, and comparisons are made among respondents, nonrespondents, and community hospitals in general. Second, the factors are analyzed that contribute to the presence of a discount in the hospital - HMO contracts listed first, and in the largest reported contracts; and third, the factors contributing to the magnitude of the discount in the contracts listed first and in the largest reported hospital - HMO has assessed. The analysis of the largest reported hospital - HMO contracts was performed partly because it was not known whether the first-listed hospital - HMO contract was or was not the contract most important to the hospital. Even though the largest reported contract most important contract, it was likely to be a significant one. Thus, by performing the analyses on both the first-listed and the largest reported hospital - HMO contracts, the meaningfulness of the results could be better evaluated. Probit and multiple linear regression are used for the latter two analyses.


Thirty-one percent of AHA-registered community hospitals responding to the 1985 AHA survey reported that they were providing services for HMO patients. This represents a threefold increase over that reported in 1980 (Rahn and Traska 1987). Moreover, the vast majority of these hospitals (83 percent) had formal affiliations with the HMOs they were serving.

Table 2 describes the characteristics of the respondents and nonrespondents to the 1986 Survey of Hospital Affiliation with Health Maintenance Organizations as well as responding to the 1985 group of AHA-registered community hospitals responding to the 1985 AHA survey. Nonrespondent hospitals are somewhat smaller than those responding to the questionnaire and are more likely to be for profit and members of multihospital systems.


Of the 801 hospitals that provide any information on reimbursement arrangements, 21 percent have a contract with only one HMO. The number of contracts varies from 1 to 16 (for one hospital) with a mean of 3.4. However, 62 percent of the hospitals with contracts have three or fewer contracts. Information on other hospital characteristics is shown in Table 3. The average size is about 300 beds with an occupancy of approximately 65 percent and an average daily expenditure per inpatient of $618. Most of the hospitals in the survey are private not-for-profit hospitals; 13 percent are public and 6 percent are for-profit facilities. Thirty-three percent are members of multihospital systems and 18 percent are members of the Council of Teaching Hospitals. On average, five hospitals are located within five miles of each of the study hospitals and 12 are between 5 and 15 miles from the study hospitals. The average number of HMOs operating in each MSA in which the hospitals are located is seven, with about 12 percent of the population enrolled in HMOs.


The correlation matrix of the variables is given in Table 4. The variables that have correlation coefficients greater than .5 are all positive: bed size and membership in the Council of Teaching Hospitals; the two variables indicating the number of other hospitals in the geographic area; the number of HMOs operating in each MSA; and the number of hospitals located within a 15-mile radius but over 5 miles away. None of the variables has correlations high enough to suggest a problem with multicollinearity.


Most of the hospitals (78 percent) report that at least one of their HMO contracts includes a discount provision, and 64 percent include a discount provision in all of the HMO contracts for which they report reimbursement information. Since the types of discount provisions vary, it is not possible to determine either the true price paid for the hospital services provided or, with the exception of the percent discount off full rates, the actual amount of discount.

Hospitals have not negotiated a high number of risk-sharing provisions or volume/revenue guarantees in their contracts. Only 27 percent of the hospitals have any fewer (6 percent) have any contracts containing volume or revenue guarantees. Only 11 percent of the hospitals have an equity position in any of the HMOs which they have a contract.


Data were analyzed using probit analysis and ordinary least squares (OLS) regression. First, the relationships between contract provisions hospital characteristics, and competitive environment explanatory variables and the presence of a discount for HMO services were analyzed by probit analysis for the first listed hospital-HMO contracts. As shown in Table 5, three factors were found to be statistically significant. Risk-sharing provisions, the number of hospitals within a five-mile radius, and the proportion of the population enrolled in HMOs were all positively related to the provision of discounts in hospital-HMO contracts. When these variables were analyzed using the largest contract reported by the hospitals (N = 382), the only significant change was that being a public hospital had a negative effect on the provision of discounts (Table 5). As with the previous analysis the presence of risk-sharing provisions was related to the presence of a discount. Although the degree of HMO presence was still an important explanatory variable, the number of HMOs, rather HMO enrollment, became the significant indicator.


To examine the effects of our explanatory variables on the magnitude of the discounts, OLS regression was used to analyze those hospitals that gave a percent discount off full rates and that also listed the magnitude of the discount. Of the hospitals with this arrangement, 134 provided the amount of the discount for the first-listed hospital-HMO contract. The range of the discount was from 1 percent to 50 percent, with a mean of 9.45 and a median of 8 (see Table 6). The analysis of these data using OLS regression (Table 7) found a significant positive relationship between risk-sharing agreements and the number of hospitals within five miles, and an inverse relationship between the number of HMOs operating in the MSA and the amount of discount. Next, these relationships between contract provisions, hospital characteristics, and competitive environment explanatory variables and the magnitude of the discount for HMO services were analyzed for the largest reported hospital-HMO contract. No variables were significant at the .05 level. but the importance of risk-sharing provisions followed the trend of the previous regression analyses (Table 7).


The interaction terms (D1 Number HMOs and D1HMO Saturation) used to evaluate the appropriateness of imputing HMO market conditions for hospitals located adjacent to urban areas were not significant in any regression. This means that such imputations are appropriate for this study. The dummy variable for rural hospitals (DV) was insignificant, which means that the number of HMOs and HMO saturation can be imputed as zero for rural hospitals.


Hospitals are increasingly developing formal relationships with HMOs and negotiating some form of discount. From a perspective of resource dependency, hospitals appear to be using contracts to stabilize their relationships with HMOs, and it appears that many are being forced to give discounts in order to achieve that end. The tendency of the hospitals to provide discounts to HMOs is consistent with the findings of our previous research (Kralewski et al. 1991). Given the overcapacity of hospital beds, HMOs are likely to have access to hospital services even without a contract. Consequently, discounts are becoming are a major focus of contract negotiations. Whether discount provisions actually achieve lower prices or not remains an unanswered question, although one would expect that price comparisons drive the negotiations. The failure of hospitals to obtain volume guarantees or similar concessions in return for discounts is puzzling. Either they are in such a weak bargaining position that they cannot press the HMOs for guarantees, or they reason that the contracts by themselves are sufficient to make the HMOs resource dependent and thus to stabilize HMO patient flow to their beds.

Although more hospitals are developing formal contracts with HMOs and a higher proportion of those contracts provide some type of discount, the factors causing this to occur are not at all clear. As expected, risk-sharing provisions in contracts tend to be associated with the presence of a discount as well as with a discount's increasing magnitude. Hospitals appear to be more willing to give a discount if they are protected in some way from the possibility of extreme adverse financial consequences. This provides considerable support for the resource dependency theory. Hospitals apparently view contracts with HMOs as important resource-stabilizing mechanisms and are willing to provide discounts to achieve those ends if they can be protected from biased selection of catastrophic illnesses or other financial risk factors. As such, risk-sharing provisions in hospital contracts add to the "stability" aspects of the resource dependency model.

Hospital competition is an important consideration. Hospitals appear to be particularly vulnerable in environments with many hospitals in a small geographic area (i.e., within five miles). The presence of other hospitals is thought to result in a more competitive environment, and the results of this study support that contention. Contracts in these settings contain a discount more often than in other communities. Moreover, the magnitude of the discount increases as hospital competi- tion in the five-mile area increases. This variable was analyzed further to determine if the findings resulted from single - versus multiple-hospital hospital settings. This was not the case, but there did seem to be a threshold for "competitiveness": two hospitals within five miles. More than two provided little additional explanatory value.

Competition resulting from the extent of HMO presence in a market area is also found to provide discounts and the magnitude of the willingness of hospitals to provide discounts and the magnitude of those discounts. However, these relationships again underscore the complexity of HMO markets. While higher levels of HMO enrollment generally were found, in our study, to increase both the probability that a hospital would be giving a discount and the size of that discount, the number of HMOs in the MSA had a negative effect on the magnitude of the discount. It appears that while increases in HMO enrollment in a community tend to create price competition, as reflected by higher discounts for hospital services, increases in the number of HMOs do not. Yet hospitals appear to be more willing to provide at least some type of discount when there are more HMOs in the community. This may be related to hospital strategies to maintain a presence in the HMO market as the number of HMOs increases and to position themselves for any significant future growth in HMO enrollment. Therefore, as the number of HMOs increases, hospitals are forced to provide discounts in order to maintain contracts and patient flow but are not forced to provide extensive discounts until enrollment increases significantly. In other words, a high HMO enrollment creates a more competitive environment than does a high number of HMOs with low enrollment.

In one set of our analyses hospital ownership was an important factor related to discounts in that public hospitals were less likely to provide discounts. Whether this is related to charter constraints in these institutions or to their lack of dependency on HMOs for resources is unknown. Conceivably their tax support would argue for the latter to be the case and, if so, this finding lends considerable support to the resource dependency model.

While our data support the contention that in certain markets hospitals develop contracts with HMOs and provide discounts in order to maintain their patient resource base, the decisions to do so appear to be based on rather narrow sets of considerations. Unlike the situation in other markets, factors such as concentration of volume of business (in this case, volume of HMO patients), excess capacity (occupancy rates), and production costs (expenditures per day) do not appear to be influential in the hospital-HMO competitive arena. Moreover, networking variables such as HMO equity position, membership in a multihospital system, and membership in the Council of Teaching Hospitals were not found to be important. Either these factors are not very effective in making hospitals less resource dependent or HMO conditions tend to override these other factors.

The lack of significance of HMO volume on either the presence of a discount or the magnitude of discounts is surprising. A problem of simultaneity may occur when the proportion of the hopital's activities contributed by an HMO (Volume HMO) is used as an independent variable to predict the provision of a discount (Discount). If Discount and Volume HMO are endogenous variables, the assumptions of the ordinary least squares are violated and as a consequence the estimated coefficients will be biased. If the error term increases when Discount is regressed on Volume HMO, this would directly increase Discount. But if Volume HMO is a function of Discount, increasing Discount increases Volume HMO. Thus the error term in the regression equation and Volume HMO are positively correlated. An increase in the error term (directly implying an increase in Discount) results in an increase in Volume HMO (also implying an increase in Discount).

When estimating the influence of Volume HMO on Discount, however, the regression technique attributes both of these increases in Discount (instead of just the latter) to the accompanying increase in Volume HMO. Thus the estimated regression coefficient is biased upward. Consequently, the lack of significance cannot be attributed to a bias, since a simultaneity problem could only have resulted in a false positive significance error. Thus, we must conclude that in these hospitals the proportion of hospital inpatient business controlled by an HMO is not related either to the provision of a discount to that HMO or to the magnitude of the discount.

It should be noted that our respondent and nonrespondent hospitals were different in several characteristics. Nonrespondents were smaller, more likely to be members of a multihospital system, and for-profit hospitals. There likely is some interaction among these variables in that for-profit hospitals are more likely to be members of larger hospital systems. Whether for-profit hospitals are less willing to participate in surveys because of cost or proprietary concerns or because they had fewer HMO contracts and therefore felt that the survey did not apply to them is unknown. Our magnitude of discounts data may also be somewhat biased since we were unable to determine the types of services covered by those discounts. The contract providing large discounts may represent a rather narrow set of services. Moreover, it must be emphasized that the analysis of magnitude of discount is based on only 134 hospitals and therefore must be regarded with caution.

Several issues related to the resource dependency model merit special attention in future research. First, the relative effectiveness of different types of HMO contracts in stabilizing patient flow has not been established. More work needs to be devoted to understanding the nuances of hospital-HMO contracts and the effects of those differences on patient concentration. Second, while our previous work found that some types of HMOs concentrate patients in hospitals that provide higher discounts, this clearly does not hold true for all HMOs. In part this is related to the structure of HMOs (i.e., group versus IPA models, etc.) and the maturity of the HMO market. The relative contribution of contracts to the stabilization of hospital resources under differing structural and market conditions is therefore an important area for future inquiry. Third, study of the relationship of discounts to the actual price paid is important. It could be the case that a hospital that did not provide a discount had a lower price than another, even after that other provided a discount. Actual price data are difficult to obtain but are very important in understanding hospital-HMO contracting.

Our data, along with our previous studies, indicate that HMOs are playing an increasingly important role in the health care delivery system and are creating markets that serve as important agenda items for hospitals. It appears that the majority of the hospitals are responding to these markets by attempting to stabilize patient volume through formal organizational contracts with the HMOs rather than attempting to improve volume through direct price concessions. Still it is clear that hospitals are finding it increasingly necessary to provide substantial discounts in their HMO contracts, particularly in markets with a high geographic concentration of hospitals and HMO penetration. The lack of influence of other hospital factors on the presence of discounts or the magnitude of those discounts again underscores that fact that much remains to be learned about how hospital-HMO markets function.


We extend our gratitude to the American Hospital Association for the use of their data; to Harold S. Luft, Ph.D., Professor of Health Economics, School of Medicine, University of California, San Francisco, for providing the hospital competition data; and to Douglas R. Wholey, Ph. D., Professor of Social and Decision Sciences, College of Humanities and Social Sciences, Carnegie Mellon University, Pittsburg, for providing the data on HMO market conditions.


1. Includes percent discount off full rates, full rates below established ceiling, fixed prospective rate per HMO enrollee, single per diem, multiple per diem, retrospective by agreed formula, HMO "lease" of a block of beds, and capitation. 2. To address the shortcoming of the National HMO Census, which attributes HMO enrollment to the MSA in which the HMO is headquartered, InterStudy data regarding the market area served by each HMO were used to prorate enrollment.


Christianson, J. B., S. M. Sanchez, D.R. Wholey, and S. Shadle, "The HMO Industry: Evolution in Population Demographics and Market Structures." Medical Care Review 48, no. 1 (1991): 3-46. Christianson, J. B., and W. McClure. "Competition in the Delivery of Medical Care." New England Journal of Medicine 301, no. 15 (1979): 812-18. Enthoven, A. C. "Competition of Alternative Delivery Systems." In Competition in the Health Care Sector: Past, Present, and Future. Edited by W. Greenberg. Germantown, MD: Aspen Systems Corporation, 1978. Feldman, R., and B. Dowd. "Is There a Competitive Market for Hospital Services?" Journal of Health Economics 5, no.3 (1986): 277-92. Feldman, R., J. Kralewski, J. Shapiro, and H. Chan. "Contracts between Hospitals and HMOs." Health Care Management Review 15, no. 1 (1990): 47-60. Feldman, R., H. Chan, J. Kralewski, B. Dowd, and J. Shapiro. "Effects of HMOs on the Creation of Competitive Markets for Hospital Services." Journal of Health Economics 9, no. 2 (September 1990): 207-22. Gold, M., and D. Hodges. "Health Maintenance Organizations in 1988." Health Affairs 8, no. 4 (1989): 125-38. Goldberg, L. G., and W. Greenberg. "The Competitive Response of Blue Cross to the Health Maintenance Organization." Economic Inquiry 18, no. 1 (1980): 55-68. Gruber, L. R., M. Shadle, and C. Polich. "From Movement to Industry: The Growth of HMOs." Health Affairs 7, no.3 (1988): 197-208. Kralewski, J. E., D. S. Doth, R. G. Rosenberg, and D. G. Burnes. "HMO - Hospital Relationships: An Exploratory Study." Health Care Management Review 8, no.2 (1983): 27-35. Kralewski, J. E., R. Feldman, B. Dowd, and J. Shapiro. "Strategies Employed by HMOs to Achieve Hospital Discounts - A Case Study of Seven HMOs." Health Care Management Review 16, no.1 (1991): 9-16. Luft, H. S. Health Maintenance Organizations: Dimensions of Performance. New York: John Wiley & Sons, Inc. 1981. Luft, H. S., J. C. Robinson, D. W. Garnick, S. C. Maerki, and S. J. McPhee. "The Role of Specialized Services in Competition among Hospitals." Inquiry 23, no. 1 (1986): 83-94. Pfeffer, J., and G. R. Salancik. The External Control of Organizations: A Resource Dependence Perspective. New York: Harper & Row, 1978. Rahn, G. J., and M. R. Traska. "Formal Hospital/HMO Contracts Increase: Survey." Hospitals 61, no. 14 (1987): 47-48, 50. Thompson, J. D. Organizations in Action: Social Science Bases of Administrative Theory. New York: McGraw-Hill, 1967.
COPYRIGHT 1992 Health Research and Educational Trust
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:health maintenance organizations
Author:Kralewski, John E.; Wingert, Terence D.; Feldman, Roger; Rahn, Gary J.; Klassen, Thomas H.
Publication:Health Services Research
Date:Jun 1, 1992
Previous Article:The effects of utilization review on hospital use and expenditures: a covariance analysis.
Next Article:Impact of hospital discharge planning on meeting patient needs after returning home.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters