P. 1
Moderation Journal of Operations Management (a. Mehta)

Moderation Journal of Operations Management (a. Mehta)

|Views: 13|Likes:
Published by Mirza SaRa

More info:

Published by: Mirza SaRa on Apr 12, 2012
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





Journal of Operations Management 20 (2002) 63–75

Technical note

The effect of location, strategy, and operations technology on hospital performance
Susan Meyer Goldstein a,∗ , Peter T. Ward b , G. Keong Leong c , Timothy W. Butler d

Department of Operations and Management Science, University of Minnesota, 321 19th Ave. S, Minneapolis, MN 55455, USA b The Ohio State University, Columbus, OH 43210, USA c University of Nevada, Las Vegas, NV 89154, USA d Wayne State University, Detroit, MI 48202, USA Received 16 June 1998; accepted 13 August 2001

Abstract Hospitals in the US are faced with challenges in how to compete and remain viable in an increasingly competitive environment. Using data from a primary survey of hospitals and from various secondary sources, we investigate the incremental effects on hospital performance of location, strategy, and technology. We find that hospital location is significantly related to performance, but that a hospital’s choice of strategy can moderate the effect of location. Additionally, we find hospitals that invest more extensively in clinical technologies tend to be better performers regardless of location. Hospital size, measured as number of beds, captures the effects of location and technology investment in accounting for a major portion of hospital performance. While we cannot argue that larger is always better for hospitals, mergers, partnerships, and other forms of consolidation currently observed in the marketplace indicate that managers in the hospital industry understand the advantage of size. © 2002 Elsevier Science B.V. All rights reserved.
Keywords: Hospital management; Operations strategy; Hierarchical regression

1. Introduction As the hospital industry in the US continues to consolidate, decisions are made daily to close, merge, acquire, and sell (Japsen, 1996). Some of the stress experienced by hospitals is well understood. For example, the tremendous pressure on hospital operations from third party payors to shorten patient hospital stays and to perform procedures on an outpatient basis resulted in excess capacity throughout the industry.
∗ Corresponding author. Tel.: +1-612-626-0271; fax: +1-612-624-8804. E-mail address: smeyer@csom.umn.edu (S.M. Goldstein).

But there are other factors in this industry that are not clearly understood. What causes some hospitals to struggle and which management responses are most appropriate in particular settings? Organizations in all industries, including hospitals, develop strategies to respond to environmental factors and competitive challenges. Those strategies drive operational decisions regarding investments in new or updated technologies. The downstream effects of strategic choices and operational decisions on organizational performance are difficult to measure, but are a topic of great interest. We use empirical methods to assess the interplay of environmental factors with organizational strate-

0272-6963/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 2 7 2 - 6 9 6 3 ( 0 1 ) 0 0 0 8 1 - X

and technology There has been little research that develops models of the relationships between hospital location. Hospital urban/rural location.64 S. and technology investments are analyzed together to assess their influence on hospital performance. strategy. where high levels of regulation circumscribe the structural variation in hospital organizations. there are specific requirements for licensing of physicians. 1. We develop a model based on hospitals’ responses to being in an urban or rural location through their choices of strategies and operational decisions on technology investments.. Hambrick. 1. 1983). Goldstein et al. 1995). strategy. this research suggests that industries behave differently with respect to how external environment and internal business decisions influence business performance. and government-owned facilities. In addition. nurses. Following a brief review of related research. gies and operational decisions regarding technology investments in the hospital industry. and the link between these factors and hospital performance. The hospital industry is unique in that it includes for-profit. 1985. non-profit. Bourgeois. strategy. sometimes competing for the same patients. / Journal of Operations Management 20 (2002) 63–75 Fig. and technology.g. urban and rural location. In general.. Previous researchers have used models of environment and strategy to predict performance in specific industrial settings (e. Other environmental factors such as munificence and hostility have little effect on hospital performance (Veliyath and Shortell. 2. 1. 1993). Location is an important factor in high contact service industries such as health care when services cannot be delivered remotely. 1968). we present the research hypotheses suggested in Fig.M. Only 5% of revenue in the health care industry is self-funded by patients (Standard and Poor. Operational definitions of the studied variables are provided next. For example. Government funding (Medicare and Medicaid) accounts for 54% of revenue and private insurers provide 34% (7% is from other sources). thus affecting hospitals’ choices with respect to role specialization (Pugh et al. as shown in Fig. 1996). unemployment rate and per capita income of the local population are significant predictors of hospital closure (Longo et al. and technicians that preclude non-compliant task assignments. Similar arguments can be made about organizational structure. We investigate hospital strategies to determine if they moderate any apparent advantage or disadvantage directly attributable to urban or rural location. and the analysis and results are presented and discussed. Hierarchical model of location. Empirical evidence suggests that some environmental dimensions normally considered in analyzing organizations are not germane in analyzing hospitals while others explain a great deal. For example. location and strategy interaction. . The idiosyncratic nature of the environment in the hospital industry suggests the need to develop models that are specific to this industry. strategy.

once made. Specifically. (1993) evaluate the performance effects of hospital strategies in response to the environmental change in the health care industry introduced by Medicare’s prospective payment system (PPS) in the mid-1980s. Similarly. Heskett. we investigate how hospital management uses strategy to respond to the environmental factor of urban or rural location. 1985. having an urban or rural location is an important environmental factor for hospitals. but they do not address the links between environment and strategy or between strategy and operational decisions (Albrecht and Zemke.’s environmental factor of the introduction of a PPS is the same for all of the studied hospitals. Hospital urban and rural location Location and proximity to markets are important factors for service organizations generally and hospitals in particular. reducing certain types of risk to investors (Campbell. Hospital location is important because the largest segment of a hospital’s market share comes from an area of proximity to the hospital (Robinson and Luft. They find that high and low performing firms use different strategies in the same environment. For example. Little work of this type has been done in the service sector. we consider location as an environmental variable that.. few studies address the linkage between environmental issues. so it is not clear that rural location by itself is an inherent disadvantage. Although a majority of hospital closures in the past occurred in rural hospitals (Cleverly. Lamont et al. we evaluate how hospitals use strategy to respond to their urban or rural location. Henry. Lamont et al. as reported in the literature. 2. 1991). (1993) also find that hospitals can improve their performance by changing their strategy to achieve better fit with the environment. 1985). Lamont et al. These results are important because they show that hospitals can use their strategies to respond to environmental conditions. McLaughlin et al. as a source of competitive advantage. Nath and Sudharshan (1994) address location relative to other hospitals as part of a hospital’s business strategy rather than as an environmental factor. While hospital strategies have been studied extensively. lack of investment in medical technologies severely limits the services that are offered. supporting their hypothesis that high performers develop strategies that are more responsive to environmental demands. these variables have frequently been used in studies of manufacturing operations. Their results indicate that hospitals with ‘proper’ fit between environment and strategy (the hypothesized ‘best’ strategy is Miles and Snow’s (1978) differentiator strategy) have the best performance. This research investigates the dichotomy of urban and rural location in the context of strategy development and technology investments. such as location. In the study reported here. It is also plausible that while market size may be adequate. but do not consider the long-term nature of location decisions.M. They use location relative to other hospitals as a proxy for convenience. 1994). Goldstein et al. In contrast. 1991. 2. In the study reported here. The literature generally regards rural location as a disadvantage for hospitals but provides limited empirical evidence that this is true. In contrast to the study reported here in which the environmental factor of hospital location is dichotomous (urban or rural).1. (1995) show that environmental factors can be used to predict the operations strategy used by successful manufacturing firms. Rural hospitals sometimes have no competition in their immediate region.2. 1986. / Journal of Operations Management 20 (2002) 63–75 65 operational decisions. Many authors provide insight into the development of strategy in service organizations. 1991). Hospital strategy In this study. rural hospitals have increasingly become targets for purchase by hospital chains because they are often inexpensive and have little competition in their immediate region. Roth and van der Velde. Swamidass and Newell (1987) use environmental uncertainty to predict elements of operations strategy which in turn are used to predict business performance. 1997). and performance. Hospitals in rural locations have struggled in recent years and their survival may depend on developing strategies that are appropriate for their location (Hudson. Ward et al. Nath and Sudharshan identify five strate- . 1995. including occupancy rate. are used to assess fit. cannot be changed. Both financial and operations performance measures. In one such study. and strategic decisions. The size of potential markets in rural areas may be an impediment because some hospitals are located near limited populations.S.

The hospital strategy data used in the study reported here are gathered using items similar to those used by Goes and Meyer (1990) who base their items on Miles and Snow’s (1978) strategic types and Porter’s (1980) generic strategies. 2. but also clinical (e.. While there are no significant differences in the performance of the five strategy groups. it is clear that they use these investments to support their strategies. Morrisey (1994) finds that the best performing rural hospitals (based on eight financial. and shows a significant correlation between investments in medical technology and net patient revenue. 1994a. Based on findings by Ashmos et al. finance. mergers. and clinical performance measures) have capital asset investments 67% higher than the median investment for rural hospitals. expenses per discharge) performance measures. The continual changes in environment and technology in this industry result in the development of new variations of existing strategies. Investing in technologies to increase clinical excellence means a hospital is focused on providing the medical services that require use of certain technologies. assets per patient day. Goes and Meyer (1990) report a longitudinal study showing that changes in hospital strategy tend to be infrequent. 1978). and contracts with other health care providers make it difficult to study the implementation and performance of various strategies. hospitals which are better performers have more capital to invest. However.g. but rather is associated with other types of performance that are important to hospitals. Similar contrasts show the best performing small urban hospitals and major teaching hospitals have capital asset investments 66 and 53% higher than their groups’ medians. Research on non-profit Catholic hospitals reports these hospitals are falling behind on equipment and technology investments (Prince.g. Efficiency is operationalized using measures of occupancy rate. 1980. While it is unclear from the literature whether urban and rural hospitals have different reasons for investing in technology. mortality rates) and operational (e.g. human resources. net patient revenue may not adequately measure performance in these non-profit organizations.g. This seems to indicate that asset acquisition is not merely driven by financial success. Porter. Miles and Snow. respectively. 1996). Hospitals with few strategy changes have better efficiency than those with many changes. In determining how to measure and classify hospital strategies.b). conversely. 1995). Hartley (1996) reports that rural hospitals purchase computerized tomography (CT) equipment because they believe that access to this medical technology improves either their economies of scale or economies of scope. profitability). / Journal of Operations Management 20 (2002) 63–75 gic groups based on business strategies and marketing. and average length of stay. operational. For example.M..66 S. we identify management responses that urban and rural hospitals use to improve their performance in their given environment. However. and occupancy rate shows the largest negative effect of frequent changes in strategy. This empirical finding suggests that hospitals’ behavior deviates from the adoption of a single set of consistent activities that focus on a single strategy. One of these responses is investment in medical technologies. and others (e. as measured by occupancy rate. There is little difference between the performance of hospitals with frequent and infrequent strategy changes in terms of profitability. While the literature provides several other strategy classification schemes. especially among high performers. Morrisey does not evaluate whether higher investment in assets improves hospital performance or. Technological preeminence means being the first to market with new technologies. than hospitals with incoherent strategies. Finding . hospitals with “coherent” strategies have better performance. 1995) on the use of multiple strategies. capital asset investment is associated with not only financial performance measures (e. as advocated in the management literature (e. There are many reasons that hospitals acquire medical technologies. Coherent strategies are defined as those with the most appropriate combinations (as defined by the authors’ industry experience) of the factors defining the groups.3. clinical excellence. and operations factors. Goldstein et al.g. including maximization of profit. and technological preeminence (Teplensky et al. we seek to identify the most prevalent hospital strategies and how strongly hospitals pursue these strategies rather than to identify a single strategy for each study hospital. Calem and Rizzo. the current trends of hospital closures. it is important to note that hospitals often use multiple strategies simultaneously (Ashmos et al. Hospital technology In this study.

1996. Swamidass and Newell (1987) show how a strategy aimed at increasing production flexibility can dampen the effects of environmental uncertainty. its influence on performance is evaluated first. Lynch and Ozcan (1994) also find occupancy rate to be a significant predictor of hospital closure. Finally. and previous research shows this measure is a significant indicator of hospital viability. but that can be reacted to through choice of strategy and technology investments. The first hypothesis addresses this issue. Because the literature suggests a hospital’s location (urban or rural) is an important environmental factor. Hospital performance Hospital performance can be difficult to assess because for-profit. 1993). 1992). the literature identifies strategic groups in the hospital industry and begins to link strategies to decision-making and performance. 1994. 1991). 2.g. 1994. Other hospital performance measures that have been used in the literature include clinical measures such as adjusted length of patient stay in the hospital and adjusted mortality rate (Greene. Goldstein et al. we adopt occupancy rate as the primary performance measure for this study. Nath and Sudharshan.5. We treat location as an environmental variable that cannot be changed. non-profit. Additionally. Practitioner-based literature has supported the notion that rural hospitals are at a disadvantage to urban hospitals (Henry. 1995. (1995) report that strategies adopted by successful Singapore manufacturers focus on the competitive priorities that respond to the environmental condition of rapid growth. 1994. In addition. leverage (ratio of long-term debt to assets). Hospital strategy has incremental association with performance after accounting for the effect of urban or rural location. / Journal of Operations Management 20 (2002) 63–75 67 an appropriate performance measure in this industry which includes for-profit and non-profit organizations is a challenge for researchers. efficiency (ratio of total expenses. In the study presented here. Hudson. Morrisey.M. to total number of discharges.4. Performance is evaluated based on occupancy rate. Decisions to invest in medical technologies are presumed to support these hospitals’ strategies. 1994. Burda (1989) reports hospitals that close have an average occupancy rate of 27% versus 47% for hospitals that remain open. Harkey and Vraciu. Hypothesis 1. Ward et al. Morrisey.. For example. Finding performance measures appropriate for all of these types of organizations is challenging.. Shortell et al. and government-owned organizations compete in this industry. Occupancy rate is the average utilization rate of hospital beds. Occupancy rate is an industry-specific measure that has been used frequently in health care research as a indicator of performance (e. Goes and Meyer.S. and financial measures such as operating costs and operating margin (Shortell et al. Hypotheses We test the performance effects in hospitals of urban and rural location. Gaynor. 1990. DesHarnais et al. adjusted for medical case mix and local wages. 1996. Similarly. Jaklevic. 1995). in this case urban or rural location. the role of technology investments in strategic decision-making needs to be evaluated. More empirical evidence is needed to determine the environmental or organizational factors that prompt these strategies. are addressed in the second and third hypotheses. and a financial measure. 1995. 2. 1995. Nath and Sudharshan (1994) show that having a coherent strategy is correlated with higher occupancy. There is no evidence from the literature that for-profit and non-profit hospitals or urban and rural hospitals have different competitive reasons for acquiring medical technologies.. and Goes and Meyer (1990) show higher occupancy rates are associated with consistent strategies. Hypothesis 2. We validate our findings using an additional operational measure. adjusted for numbers of inpatients and outpatients). Urban location is positively associated with hospital performance.. Strategies are frequently used to counteract the effects of environmental conditions such as those associated with location. we measure the number of medical technologies used by each of the study hospitals. Ketchen et al. and technology decisions. In short. For these reasons. . strategy. The main effect of strategy and the extent to which strategy can be used to moderate environmental effects. assessment of strategies should be tied to hospital performance.

but it is important to account for this variable because previous research indicates that it may influence performance. rehabilitation. indicating the answers to the perceptual questions used in this questionnaire are reliable. Questionnaires were mailed to two individuals at each hospital. psychiatric. 1981). The first response from each hospital is used in analysis. Higher than average investment in technology is often associated with good performance in the hospital industry and therefore. This study focuses on the strategies used by general acute care hospitals. We conservatively test the significance of partial effects of each component as it is added to the model so that only its incremental explanatory power (after accounting for the previous components in the model) is measured. 1994). Morrisey. 1994a. hospitals generally compete for a local market. focusing on one region may be most appropriate when determining the strategies that hospitals choose to be competitive in the marketplace. and the second responses are used to test the reliability of survey responses. hospitals that invest more heavily in technology than other hospitals with similar location and size may see improved performance (Prince. We test the research hypotheses by estimating the model shown in Fig. and . / Journal of Operations Management 20 (2002) 63–75 Hypothesis 3. The fourth hypothesis addresses the partial effect of technology (after accounting for location and strategy). Hospitals’ technology investments are incrementally associated with performance. the questionnaire captures the strategies used among a group of competing hospitals. Hierarchical regression analysis allows us to determine whether the partial effects of strategy and/or technology investments as decision variables influence performance once the effect of urban and rural location has been accounted for. It seems intuitive that the number of medical technologies a hospital uses would be highly correlated with its size. 1 as a hierarchical model. In addition. while 30 hospitals returned two completed questionnaires. 1995) and restricting sample size. This methodology is used because we build a model that includes environmental (urban or rural location) and decision (strategy and technology investments) variables. Two hospitals for which published performance data are not available were removed from analysis. Sixty-seven hospitals returned at least one survey.2. which the sample used here has.1. There are several advantages of using a single geographic region for this study.b. and it is likely to be associated with the decision variables of strategy and technology investments. Hypothesis 4. 1994). so specialty hospitals (i. but there is evidence that technology investments are associated with other factors (Kimberly and Evanisko. These hypotheses are tested using hierarchical regression analysis which is useful for determining the incremental contribution of each variable that is evaluated.68 S. The interest a hospital has in particular technology investments is in response to both environmental conditions and the strategy that the hospital has chosen to pursue. 3. Goldstein et al. Urban or rural location cannot be changed once hospitals are established. Because this study investigates urban and rural hospitals. Results of second rater tests are reported in Section 3. The association of urban or rural location with performance is moderated by hospital strategy. By studying a single region.e. it is critical to have adequate representation of both. The limitations of using a single state for this study include the possibility of excluding strategies used in other regions (Solovy. Therefore. after accounting for the effects of urban or rural location and strategy.9%. for a response rate of 41. Technology investments result from a set of operational decisions that organizations make to support their strategies.e. Sample and operational definitions Data on strategy decisions and technology investments are obtained from a questionnaire that was mailed to 160 general acute care Michigan hospitals. as well as their interactions (Hayes.M. children’s and veterans’ hospitals) are excluded from analysis. i. Solovy (1995) reports that hospital strategies vary by region in the US. organizations try to adopt technologies that will improve their performance and capitalize on their competitive strengths. No significant difference is found between the data obtained from primary and secondary respondents. testing the statistical significance of the partial effect of each component shown in the model. mainly CEOs and Vice Presidents of Operations.

47. This bias is measured by splitting hospitals into two groups (fewer than 150 beds and greater than 150 beds) and performing a χ 2 -test (χ 2 = 3. is marginally significant. These two factors are correlated (R = 0. This study includes 43 urban hospitals and 22 rural hospitals. 1.. Government-owned hospitals are non-profit hospitals.. 1991).1. Bias based on hospital ownership (i.S.42). Goldstein et al. The communities of responding rural hospitals have a significantly higher percentage of individuals over 65 years old (the age at which Medicare coverage begins) than the communities of urban hospitals.26. which account for 53% of the variance in the strategy items. The item loadings on this factor range from 0. In addition.1. The strategy items and their correlation matrix are shown in Table 1.f. P = 0. Hospitals often use more than one strategy at the same time (Ashmos et al. Strategy items are factor analyzed (using unweighted least squares) to determine factor loadings of the items. Non-respondent bias is assessed by comparing hospitals that completed at least one questionnaire with hospitals that did not respond. 3. Responding hospitals are somewhat larger on average than non-respondents.2. bias is assessed based on occupancy rate.01). an indicator of hospital size. An oblique rotation is used allowing the factors to be correlated because this rotation more accurately represents the complexity of constructs in the real world (Ford et al. The second factor includes five items related to service quality and dependability.40). d. the percentage of Medicare and Medicaid patients treated in the urban and rural hospitals in this study are not significantly different (HCIA. non-profit or government-owned.M.1. No variables loaded higher on a subsequent factor than on the first two factors so only the first two factors. These statistics may indicate that patients go outside their local community for medical care. / Journal of Operations Management 20 (2002) 63–75 69 the final sample size used in this analysis is 65 hospitals. indicating many hospitals pursue both strategies simultaneously.1. P < 0. that Medicare and Medicaid patients seek fewer medical services when they reside in rural communities. 1976). 1995).485 to 0.07).399 to 0.794. The first factor includes six items related to marketing and market-driven strategy. Principal component analysis is used to reduce the item scores associated with each factor to single scores . there are no for-profit hospitals among the respondents) is not significant (χ 2 = 0. P = 0. or that these groups’ access to medical care is somehow limited. There are several significant differences in the populations of the regions of the urban and rural hospitals studied here. 1996). and is termed the marketing strategy factor. The items associated with this factor have loadings ranging from 0. Operational definitions 3. Strategy Hospital strategy data are obtained from the questionnaire already described and include items similar to those used by Goes and Meyer (1990). while larger hospitals responded at a rate that is higher than expected. and is referred to as the operations strategy factor. 3. The data used to define these differences are obtained from census data.f. Respondents were asked to rate the importance of each item in meeting their hospitals’ strategy on a seven-point Likert-type scale.833. based on the zip codes of the hospitals in this study (CACI.72. as reported in Table 2. However. while survey data are used for measures of strategy and technology.80. = 1. Finally. These analyses of non-respondent bias show that. P = 0. but a distinction is often made between them because they may attract different customer types. d. Non-respondent bias based on number of hospital beds. 1986. the median and per capita incomes of these rural locations are significantly lower than the urban locations on average. HCIA defines each hospital as urban or rural based on US Bureau of Census’ designation of urban and rural communities. = 1.e. Harman. as reported in Table 2. the performance measure used in this study. Second source data are obtained for environment and performance variables. the respondents are not different from non-respondents in terms of ownership or performance. as designated by HCIA (1995). are used in this analysis. A Student’s t-test comparing this performance measure for respondents and non-respondents is not significant (t = 0. Following is a description of the methods used to gather data to evaluate the relationships shown in Fig. Urban/rural status Each hospital included in this study is defined as having either an urban or rural location.

1. Technology Technology is defined here as a hospital’s ability to provide a particular medical service based on either the equipment and/or the skills that are required to provide that service.167 0.215 0.343∗∗ 0.000 0.344∗∗ 0.388∗∗ 1.251∗ 0. Goldstein et al.484∗∗ 0. The technologies selected for inclusion in the questionnaire reflect a broad spectrum of services that might be offered in a general acute care hospital.833 0.634 −0.295 0. P < 0.457∗∗ 0.589 0.176 0.174 0.337∗∗ 0.523 0.033 0.485∗∗ 1. Second responses for the mailed questionnaire are used to assess inter-rater reliability of the strategy Table 2 Factor analysis output for strategy items Item Loadings Factor 1a Marketing factor items Reputation within industry Innovation in health care services Hospital name recognition Extensive advertising Forecasting market growth Beating the competition to market with new services Operations factor items High level of customer service Service quality control Experienced/trained personnel Capability to provide specialty health care Dependable service provider a Factor 2b 0.140 0.288∗ 0.000 0.485 0.3.290∗ 0.185 0.431∗∗ 1. / Journal of Operations Management 20 (2002) 63–75 Table 1 Pearson’s correlations for strategy factor items 1 Marketing strategy factor Reputation within industry Innovation in health care services Hospital name recognition Extensive advertising Forecasting market growth Beating the competition to market with new services Operations strategy factor High level customer service Service quality control Experienced/trained personnel Capability to provide specialty health care Dependable service provider ∗ ∗∗ 2 3 4 5 6 7 8 9 10 11 1. 3.242 −0.000 0.000 0.243 −0. 1978).376∗∗ 0.455∗∗ 1.147 0. There is agreement within one point (on the seven-point scale) in 69% of cases for the marketing strategy items (average for six items) and 78% for the operations strategy items (average for five items). Nunnally.794 0.484∗∗ 1.000 0.000 0.273∗ 0.000 0.377∗∗ 0.383∗∗ 0.070 items using Shortell and Zajac’s method (1990).549∗∗ 0. and each hospital’s 0.000 P < 0.252∗ 1.042 0.352∗∗ .633 Marketing strategy factor.142 0.81 for the marketing strategy items and 0. as recommended by Boyer and Verma (2000).226 0. for each hospital.307∗ −0.320∗ 0.286∗ 0. Alpha coefficients are 0.060 0.427∗∗ 0.000 0.M.354∗∗ 0.000 0.555∗∗ 1.591 0. An example of technology based on equipment is magnetic resonance imaging (MRI) services which require ownership of an MRI machine. Data on technologies are obtained from the questionnaire.448∗∗ 0..558 0.000 0. These two component scores for each hospital are used in the regression analysis performed here.399 0.308∗ 0.329∗∗ 0.01.040 0. b Operations strategy factor.258∗ 0.231 −0.443∗∗ 0.279∗ 0.73 for the operations strategy items.277∗ 0. Cronbach’s coefficient alpha is also used as a measure of scale reliability. An example of technology based on both equipment and skills is heart transplantation services.258∗ 0.241 0.057 0. These alpha levels are acceptable for this type of empirical research (Flynn et al.191 0. . Respondents were asked to indicate whether they are currently using each of the technologies listed.237 0.106 1.360∗∗ 0.556∗∗ 0.125 1.506 0. 1990.193 0.70 S.334∗∗ 0.431∗∗ 0.402∗∗ 0.05.467∗∗ 1.332∗∗ 0.

01). x1 the urban/rural location variable. Data on average daily census and number of staffed beds are reported annually by the AHA (1995). 0. 3.34 R2 0.18 0. Hypothesis 2 not supported: To assess the impact of strategy on occupancy. so scores for both strategy factors are used in the regression equation for each of the studied hospitals.29(x1 x3 ) 0. x2 the marketing strategy. x1 = 1 for rural hospital.05 F 16.21 (F = 16. The average occupancy rate is 62% in urban hospitals and 44% in rural hospitals in this sample.20 0. The purpose of this step is to determine whether the strategy factors make an incremental contribution toward explaining variance in hospital performance.15 5. occupancy rate is the primary performance measure used in this study. F measures the significance of the change in R2 at this step in the hierarchical regression analysis. 1.40 Adj R2 0.46 −0. ei is the error-term. 4. / Journal of Operations Management 20 (2002) 63–75 71 score is calculated as the total number of the queried technologies currently used.13 0. x3 is the operations strategy.58. Findings are summarized in Table 3.24 P < 0. a the constant.21 0.07(x2 )a .21 0. P < 0. Performance Consistent with other studies. a x is marketing strategy factor. The hypothesis is tested by evaluating the significance of the F-change ( F) statistic. The resulting R2 for Eq. x1 = 0 for urban hospital. thus testing the hypotheses.22 0.58∗∗∗ 0. meaning that 21% of the variation in occupancy rate among hospitals in Michigan is accounted for by urban or rural location. Two additional measures of performance—financial leverage and efficiency—are also evaluated to validate the results obtained when using occupancy rate as the performance measure. the marketing and operations strategy factors are added to the hierarchical regression equation containing the location variable. hierarchical analysis allows the contribution of each component of the model to be evaluated. Urban/rural location is a significant predictor of performance as measured by occupancy rate. (1) is 0.29 0. Findings We use hierarchical regression analysis to determine the significance of the incremental contribution of each of the variables shown in Fig. P < 0. This result is shown in Table 3. 0. While the final hierarchical regression model provides the same explanatory power as simultaneous regression.35 0. Some of the variables in Eq.01 0. 2 b x is operations strategy factor.04(x3 )b −0.4.05. ∗∗∗ P < 0. This result supports the first hypothesis that urban or rural location has a significant effect on hospital performance.01.S.10.1. The regression equation used to test the first hypothesis is yi = a + b1 x1 + ei (1) where yi is the performance variable for hospital i. There is evidence from the literature that hospitals pursue multiple strategies simultaneously.56(x1 x2 ). There is a minimal change in R2 when the strategy variables are added to Table 3 Hierarchical regression results with occupancy rate as dependent variable (n = 65) Step 1 2 3 4 ∗ ∗∗ Variables Urban/rural location Strategy Urban/rural × strategy (interaction) Number of technologies R2 0. Goldstein et al.M.11∗ β −0. b1 the weighted value of x1 variable. (2) are defined in Eq.95∗∗ 5. Support for Hypothesis 1: The first step in the hierarchical regression analysis performed here is to test the urban/rural location dichotomy for significance. Each hospital’s rank (relative to all Michigan hospitals) on these two measures is obtained from HCIA (1995). (1): yi = a + b1 x1 + b2 x2 + b3 x3 + ei (2) where bz is the weighted value of xz variable. Occupancy rate is calculated from these data (average daily census/number of staffed beds). 3 .

and three-way interactions among the independent variables and none is significant. 5. this means that strategy can alter the significant influence of location on performance.35 and the change in R2 is significant at 0. To demonstrate that the model applies when other measures of hospital performance are used. The change in R2 from the previous equation is significant ( F = 5. The hierarchical regression statistics for this model are given in Table 3. Technology investments can be thought of as a set of decisions used to support the strategy or strategies that a hospital has chosen to pursue.13 ( F = 5. These data support the third hypothesis that the effect of urban or rural hospital location can be moderated by focusing on particular strategies. Bromby. Specifically. The findings indicate a disadvantage for hospitals in rural locations. strategies. In the telecommunications industry. the negative coefficient of the marketing strategy–location interaction suggests that a marketing-oriented strategy is ineffective for rural hospitals. Goldstein et al. The resulting regression equation is yi = a + b1 x1 + b2 x2 + b3 x3 + b4 x1 x2 +b5 x1 x3 + b6 x4 + ei (4) where bz is the weighted value of xz variable. income from urban areas has been used to support the provision of services in rural areas (Cobb. Using efficiency as the performance measure results in a model with R 2 = 0. Discussion and conclusions This study investigates the impact of urban/rural location. Strategy is cast in the role of a “pure moderator” (Sharma et al. Independent rural hospitals may find that consolidations among themselves or with urban organizations provide the redistribution of capital that they require for operation. 1981) because there is no significant main effect of strategy on performance while there is a significant strategy–location interaction effect and a significant urban/rural location effect..M. Both of these models are statistically significant. adding the technology variable to the hierarchical regression equation results in R 2 = 0. and their interaction. the cost of supplying certain types of services to rural areas sometimes exceeds the financial benefits.40. Although none of the performance measures used is a perfect indicator of hospital performance. P < 0.31. . x1 x3 is the interaction between urban/rural location and operations strategy. strategy. P < 0. x4 is the technology variable. 1998). generally with significant coefficients for the same regressors. The x1 x2 and x1 x3 terms account for the interactions between the urban/rural location variable and the two strategy variables. The regression equation used to address the third hypothesis: y i = a + b 1 x1 + b 2 x2 + b 3 x3 + b 4 x1 x2 +b5 x1 x3 + ei (3) sample of hospitals. and technology investments on hospital performance. Support for Hypothesis 3: The third hypothesis is tested by adding the interactions between location and the strategy variables to the hierarchical model. In practical terms. the regression model containing the same independent variables is evaluated using hospital efficiency and financial leverage as performance measures. and the F is not significant as shown in Table 3. These data show that the extent to which hospitals focus on each of the measured strategies (marketing and operations) does not significantly explain the performance of the study hospitals when the effect of location is already accounted for. the analysis demonstrates that the findings are robust across a number of performance measures.22. We also tested all remaining two. Support for Hypothesis 4: The fourth hypothesis is tested by adding the technology variable to the regression equation containing location. Using leverage as the performance measure results in a model with R 2 = 0.11. / Journal of Operations Management 20 (2002) 63–75 the regression equation. This result provides evidence that there is no ‘best’ strategy for hospitals to pursue because neither of the strategy factors contributes significantly to explaining hospital performance. x1 x2 the interaction between urban/rural location and marketing strategy. particularly for capital-intensive services such as those provided by hospitals.05).95.72 S. The total R2 following this step in the hierarchical model development is 0.10). 1999. For this where bz is the weighted value of xz variable. Not surprisingly.

this study suggests that strategy moderates the inherent disadvantage of rural location.18 0.10 0. the sign is negative because urban hospitals are coded as 0 and rural hospitals as 1) and. not surprisingly. 1981. 1996.00 −0. Hospitals with the most technologies tend to be better performers. By evaluating the correlation between size and the other studied variables. AdditionTable 4 Correlations among independent variables and dependent variable 1 IVs Size Urban/rural Marketing strategy Operations strategy Number of technologies DV Occupancy rate ∗ ally. Data on number of beds.06 0. the literature is inconclusive on the relationship between hospital size and performance.01). Goldstein et al. Rogers. 1996.50. only the size variable and the moderating effect of strategy on location are significant.50∗∗∗ 0. The data are positively skewed. ∗∗∗ P < 0.. Results indicate that size explains a major portion of variance in occupancy rate (R 2 = 0.21∗ −0. P < 0. This study shows that using a marketing-oriented strategy in a rural location is not effective.S. and other forms of consolidation currently observed in the market- 2 3 4 5 6 1. 1994. However. supporting Morrisey’s (1994) findings that small hospitals with higher than average technology investments prosper while those with the least technologies sometimes struggle to survive.15 0.10 1.46∗∗∗ 1. Kimberly and Evanisko (1981) provide evidence that size is a confounding variable because it is a significant predictor of heavy technology investment.9.00 −0. 1997). Hospital size may be associated to some of the variables studied here as previous research indicates size may be a useful predictor of hospital closure and the extent of technology investment. Kimberly and Evanisko. an indicator of hospital size. P < 0. Health Care Strategic Management. ∗∗ .01. this time recalibrating by inserting the hospital size variable in the hierarchical regression model first.04 1.01. it appears that the marketing strategy is effective for urban hospitals. 1997).01). P < 0. mergers.. partnerships.00 0. The hierarchical regression analysis described was repeated.20 −0. / Journal of Operations Management 20 (2002) 63–75 73 More important than the location issue alone.79∗∗∗ 1. There is no significant interaction between size and the strategy variables. This log transformation is routinely used in the literature when evaluating the effect of organizational size (e.60∗∗∗ 0.M. Size is correlated with location (R = −0. and the urban/rural location and technology variables are no longer significant. others argue that at-risk small hospitals have already closed and future downsizing will occur in larger hospitals in an effort to squeeze excess capacity out of the system (Cleverly. and a log transformation is performed to normalize the data.62. Longo et al. indicating collinear relationships with size. or a combination of the two strategies results in better performance.00 0. we can identify collinear relationships as shown in Table 4.21∗ −0. Boyer et al. P < 0. However. Snow.10. 1991. but that other factors are also important.00 P < 0.05. are obtained for the study hospitals from a published source (AHA.g. 1995). In this analysis.60. The identification of an interaction between hospital location and strategy provides evidence that hospitals can moderate location effects on performance through their choice of strategy. F = 102.00 0.31∗∗ 1. focused on building capabilities. 1996). While some authors find small hospitals are unstable and more likely to close (Lynch and Ozcan. Investments in technology have an incremental effect on performance even after accounting for the effects of location and strategy.47∗∗∗ −0. While we cannot argue that larger is always better for hospitals. the technology variable (R = 0. Either an operations-oriented strategy.

The AHA Guide to the Health Care Field. McDaniel. Reinventing rural hospitals. and performance: a longitudinal study of California hospitals. 1989. 1995. TX. Modern Healthcare 26 (22). 1999. Service America! Doing Business in the New Economy. Chicago. strategies can be developed that complement a hospital’s urban or rural location.C. 1985. 1998. 1995. / Journal of Operations Management 20 (2002) 63–75 CACI Marketing Systems.J. Specifically.J. 23. R. J. McMahon. Sakakibara.. K. Leong. 1990..R. Hartley. MA. 1994. 1991. 15 (3) 12. Campbell. Bromby. and economic performance in volatile environments. Why hospitals close. Healthcare Financial Management 45 (9).. In: Proceedings of the Best Paper Academy of Management Meetings. D. 1990. environmental change.. Mobile technology in rural hospitals: the case of the CT.. this analysis suggests that strategy can be used to moderate the inherent disadvantage of rural location regardless of size.E. Ford. 687–707. Multiple raters in survey-based operations management research: a review and tutorial. Tait.. Vraciu. Modern Healthcare 26 (10). 1996. Greene.K.A.. Hambrick. 1995. Journal of Operations Management 15 (4). K.. Health Care Strategic Management. Larger urban hospitals increasingly on closed list. Michigan association reports on hospital performance.. Harvard Business School Press. J.. Regardless. Academy of Management Journal 26 (4). W. Academy of Management Journal 28. In: Proceedings of the Best Paper Academy of Management Meetings.). 194–200. Managing in the Service Economy. the structure of hospital costs.F. T. and the cost of empty hospital beds. Rural priorities. While our study is limited to Michigan hospitals. E. Gaynor. 24–35. The hospital industry in states or regions with larger rural areas may result in different configurations than those in more urban areas as these hospitals find differing advantages from size and the strategies that they pursue. OH. MD. perceived uncertainty. Boyer. 1995. Health Care Strategic Management 15 (3). Southern Economic Journal 61. Green acres: rural hospitals are lucrative targets for investor-owned hospital chains.A. P. Jaklevic. Small hospitals can also share mobile technologies (Hartley... 16–18. Telstra caught in rural services storm. the findings present clear implications for hospital managers. S. OH. Albrecht. W.. Telephony 236 (23). Uncertain demand. Modern Healthcare 26 (52). High profit strategies in mature capital goods industries: a contingency approach. Another record year for dealmaking.L. Health Care Management Review 17 (4). Health Services Research 26 (4). Homewood. University of Chicago Press.H. 40–44.S.K. L. Baltimore. Quality of health care and financial performance: is there a link. P. D.K.. Cleverly. 77–78. G.T.. 37.74 S.O. Ashmos. R.. 1997.L.. B.. R. Cincinnati. Bourgeois. . 1182–1198. Schroeder. Krajewski. Hospitals and Health Networks 69 (4). R.D. The Sourcebook of Zip Code Demographics.. 22–23. Unlocking the potential of advanced manufacturing technologies. HCIA. 2000. 1991.G. Arlington. B. Chicago. 85–89. Cincinnati. R. The nature of internal complexity in unusual organizations: a study of environmentally sensitive and insensitive hospitals. J. Calem. 291–314. Boyer.. AHA.K. The application of exploratory factor analysis in applied psychology: a critical review and analysis. Modern Healthcare 19 (12). Wroblewski. 80. M. 1996. Harkey. Hauge.. K.. Zemke. Harcourt Brace College Publishers. Modern Factor Analysis. Heskett. MacCallum. Regardless of size.. 1994. Japsen. 128–140. pp. IL. Directory of US Hospitals.. J. IL. Boston. Meyer.K. 1992.. L. Goldstein et al. Production and Operations Management 9 (2). K. 1996. Flynn. HCIA (Health Care Investment Analysts Inc. S. Flynn. Cobb. M. Competition and specialization in the hospital industry: an application of Hotelling’s location model. Florida 2nd state to issue ‘report cards’.C.. Personnel Psychology 39 (2). 23.. D.. Statistics. 1996) and participate in outreach programs in which specialists bring previously unavailable services to their facilities. D. A. M. Journal of Operations Management 9 (2). Measuring outcomes of hospital care using multiple risk-adjusted indexes.B. place indicate that managers in the hospital industry understand the advantage of size (Japsen. Patterns of strategic change. Henry. R. health care utilization and medical expenditures. 55–64. L.B.. 548–573.. R.. Dow Jones–Irwin. Hospital size. 1996). Hayes. R.. H.R. R.. 1997. 1996.P. J. D. is positively related to performance and consolidations among hospitals and other health care organizations will likely continue unabated for this reason.C.. 1996.. Bates. J. Journal of Health Economics 14 (3). Hudson. 1991. 1983. W.. Ward.. Goes. Rizzo.M. References AHA (American Hospital Association). Balancing hospital productivity. 250– 284. 1995.A. 1986. 425–445. Verma. W. 213–234. 1986..F. 291–317. CACI.. Telecommunications International Edition 32 (8). Health Services Research 31 (2).. Harman. 1997. Empirical research methods in operations management. Building the info highway to anywhere. Strategic goals. 331– 347. perhaps a surrogate measure for location and technology investment. a marketing strategy may be effective for urban hospital and pursuing an operations-oriented strategy appears to generally pay off for rural hospitals. 1985. VA.. USA. Burda. S.. Fort Worth. Duchon.. 1976. Trustee 47 (8). DesHarnais....

J.C.M..M. Unclear visions for the future. Snow. The Free Press. J. E. Operating performance and financial constraints of Catholic community hospitals. Hickson. D. operations strategy. A. average length of stay. Health Care Management Review 19 (4). 1995.X. M. Sohn.T. and performance: empirical study of Singapore manufacturers. Organizational innovation: the influence of individual. What’s next for US hospitals. C. Administrative Science Quarterly 13 (1). Strategic Management Journal 15 (1).S.J.J. C. 25–37. 293– 391.. Nath.R. Kimberly. 1995. Hinings. Foster. 1995. 689–713.. Shortell.. Sum... J. Predicting the unpredictable. 1990. N. 377–401.E... O. Solovy.. Psychometric Theory. Pannesi.. G. 1985. Boerstler. Dimensions of organization structure.S. Hospital closure: an efficiency analysis. Nunnally. 1968. 63–76.. 1994b. Porter. Modern Healthcare 24 (46). S. Shortell. and performance: a longitudinal study of changing strategies in the hospital industry.M. Turner.T. Snow. S. Ward. Prince. Veliyath. Competitive Strategy. 359–381. McGraw-Hill. Y... Management Science 33 (4)..R. J.R. S. 1978.. J. S. Longo. Standard and Poor.. Journal of Health Economics 4 (4). Academy of Management Journal 33 (4). strategic planning system characteristics and performance. Lamont. Shortell...M. Hoffman.J. R.J. Rogers. Journal of Operations Management 10 (4)..T. Gur-Arie.F. M. R. 38–48. Miles.. D.. Academy of Management Journal 36 (6).S. C. Leong. O’Connor. Carman... McGraw-Hill. and performance: a path analytic model. 1978. S. 1993. Pugh.R. Health Care Management Review 19 (4). Sharma. 1980.. J.M.R. 1991. Hospital and Health Services Administration 39 (2). Top 100 hospitals set new success standards.B.. 1996..T. P. Manufacturing strategy. Pauly. C. Zajac. M. Duray. International Journal of Operations and Production Management 11 (3). 1993. 74–76. Hospital adoption of medical technology: an empirical test of alternative models. Snow. Perceptual and archival measures of Miles and Snow’s strategy types: a comprehensive assessment of reliability and validity. Identification and analysis of moderator variables. Health Services Research 28 (5). . McLaughlin. 1994. 1–11.P. D. Journal of Marketing Research 18.. Sudharshan.. Evanisko.. J. 43–51.. 437–465.S.. J. Kathuria.. P.. Durand.. organizational.R. J. D.M. 509–524. 1994.. Business environment.... 1996. Marlin. and Process.. 291–300. Hughes. R.. The different operations strategy planning process for service operations. 1987.M.V. Managed Healthcare 6 (1). R. T. 1996. Newell.. Journal of Heath Care Finance 22 (3). T.. Ozcan. 1993. 1981. Structure.R.. Assessing Catholic community hospitals versus nonprofit community hospitals. discontinuous environments. The etiology and determinants of hospital closure.R. 76–80.. Shortell. 18–20. Standard and Poor’s Industry Surveys. Porter’s generic strategies. 1994a.. 34–48.. Kimberly. 623–640. Measuring strategy coherence through patterns of strategic choices. 65–105.J. 1991. R.. New York.D. Journal of Management Studies 30 (3). H. A. C. J. 817–832. and contextual factors on hospital adoption of technological and administrative innovations. J. Prince.. Assessing the impact of continuous quality improvement/total quality management: concept versus implementation.. C. Health Services Research 30 (2). 75 Robinson.. 1995. 99–115.. 1995.L..K. B. J.L. 26–29.. M. D.. E. O’Brien. Morrisey.C..C. / Journal of Operations Management 20 (2002) 63–75 Ketchen. New York. W..W.. The impact of hospital market structure on patient volume. Roth. Strategic orientation. Teplensky. M.. Organizational Strategy.. Modern Healthcare 26 (19). Goldstein et al. Schwartz. Operations and marketing: a competitive service strategy. M.E.A. D. Health Services Research 30 (3). Swamidass. Academy of Management Journal 24 (4). 1994. 205–220. Journal of Operations Management 13 (2). C. Organizational configurations and performance: a comparison of theoretical approaches. and the cost of care. 1278– 1313. Lynch.. New York. A.A. Hospital and Health Networks 69 (1). H. J. S. Luft.. E. van der Velde. 1981. Hillman. environmental uncertainty.V. D. Thomas.

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->