Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 Role of the Hospital in the Acquisition of Technology GERARD F. ANDERSON AND EARL P. STEINBERG Hospitals are major consumers of both new and established medical technol- ogies. In 1991, for example, hospitals spent $30 billion on capital projects and had to choose which drugs and devices to purchase from among 750,000 differ- ent options (Lumsdon, 1992~. Hospitals need to make decisions regarding the purchase of "big-ticket" items such as automated labs, radiography and fluo- roscopy equipment, patient monitors, magnetic resonance imagers, and comput- ed tomography scanners (Anderson, 1990) as well as "little-ticket" items- such as tissue plasminogen activator versus streptokinase and high- versus low-osmolality contrast media. In addition, hospital managers are confronted with decisions regarding whether their hospitals should perform procedures con- sidered by insurers to be experimental and for which reimbursement is uncertain, such as autologous bone marrow transplantation for women with metastatic breast cancer (Hall and Anderson, 1992~. In addition to their role as consumers of medical technology, hospitals influ- ence the diffusion of technology in other ways. For example, since hospitals are generally the earliest adopters of new technologies, their reactions to those tech- nologies have a major impact on the subsequent acquisition decisions made by other types of providers. In addition, hospitals are the sites for many clinical trials of drugs and devices and for the clinical training of most physicians. In this chapter we discuss how recent public policy decisions have provided hospital managers greater incentives to conduct technology assessments. We then contrast the likely characteristics of technology assessments performed from a societal perspective with those of technology assessments performed from a hospital perspective. Our major conclusions are that technology assessments 61
62 GERARD F. ANDERSON AND EARL P. STEINBERG conducted from the perspective of the hospital could result in the greater and more rapid diffusion of technology, with less emphasis placed on controlling total health care costs or improving the long-run health status of the population, than assessments conducted from a societal perspective. The difference between a hospital's perspective and society's perspective should be kept in mind when regulations and payment systems are designed. THE CHANGING FINANCIAL ENVIRONMENT Until recently, hospital managers had little financial incentive to critically evaluate new or established technologies (Anderson and Steinberg, 1984~. lIos- pitals were paid under a cost- and charge-based reimbursement system that es- sentially gave them a "blank check" to purchase equipment without the need to monitor the medical practices of their physicians (Davis et al., 1990~. For phar- maceuticals in particular, full cost would be paid for virtually any new product that was approved by the Food and Drug Administration and not considered experimental by insurers because public and private insurers were not aggres- sively pursuing utilization review (Hall and Anderson, 1992~. Recent changes, however, have provided hospital managers with a stronger incentive to become involved in both medical practice evaluation and technology assessment. The change from cost- and charge-based reimbursement to prospec- tive payment has changed technology-intensive departments, such as radiology, from being "profit centers" to being "cost centers." In addition, with the inclu- sion of capital expenditures under the Medicare prospective payment system, hospital managers have become more concerned about their capital budgets, since hospitals can no longer pass higher capital costs on to the Medicare program. Instead, they must now finance capital expenditures from internal funds, venture capital, partnerships, retained earnings, philanthropy, leasing, rental, borrowing, or, in the case of not-for-profit hospitals, equity financing. In addition, the in- creased level of enrollment in health maintenance organizations and other man- aged care programs has forced hospitals to scrutinize technology acquisitions more closely. Finally, the judicial system has made hospitals legally responsible for the medical care delivered by the physicians on their staffs (Darling v. Charleston Community Memorial Hospital, 1965~. Although government and private payers have given hospital managers a stronger financial incentive to become prudent purchasers of medical technology and to monitor medical practices more closely, several other legal and organiza- tional forces are counterbalancing the financial incentives of hospital managers to become more prudent purchasers of new technologies. For example, hospital managers are confronted with the threat of antitrust violations if they purchase equipment jointly with other hospitals and the threat of violations of the safe harbor regulations if they become joint partners with physicians (Anderson, 1992~. Certificate-of-need regulations may prevent some hospitals from pur
ROLE OF THE HOSPITAL IN THE ACQUISITION OF TECHNOLOGY 63 chasing equipment that would increase their productivity. Malpractice concerns could arise if a hospital deviates from the trends that are prevailing at the moment and does not purchase certain equipment or use certain drugs. Moreover, aca- demic medical center hospitals may choose to adopt technologies that are not cost-effective to remain in the forefront of research and education (Anderson et al., 1989~. Probably the most important constraint on the hospital manager's discretion with respect to technology acquisition decisions is imposed by the medical staff, a body that is crucial for attracting patients to the hospital. In most cases, the clinical and financial incentives of a physician are to do everything possible for a patient rather than to pay close attention to cost and cost-effectiveness (see chap- ter 5, this volume). In addition, in the case of new technologies that have high public visibility, such as magnetic resonance imaging, a hospital may conclude that it is in its overall economic interest to acquire the technology to protect or enhance its market share of patients, even if it thinks reimbursement for the technology may fall short of its cost (Steinberg et al., 1988~. Thus, although a number of new forces that increase hospitals' incentives to perform technology assessments and monitor medical practices have emerged over the past decade, other incentives cut against these new forces. As a result, it is unclear how extensively hospital managers have actually altered their behav- iors. For example, only 20 percent of hospitals had established a formal technol- ogy assessment committee by 1992 (up from 18 percent in 1990), and only an additional 11 percent were considering the formation of such a committee in 1992 (Johnsson et al., 1991; Lumsdon, 1992~. MODELS OF HOSPITAL BEHAVIOR Economists have developed several theories to explain the behavior of hos- pital managers (Feldstein, 1988~. We present three models of hospital behav- ior models of price, technology, and utility competition and propose that these theories be used to consider how different types of technology assessments could influence the scope of technology diffusion and the nature of medical practice. The price competition model uses traditional economic theory to explain hospital behavior. This model assumes that the hospital manager faces a down- ward-sloping demand curve and evaluates technologies from the perspective of profitability. From that perspective, new technologies are acquired when the expected revenue stream exceeds the expected cost over the useful life of the product. In making these financial calculations, the hospital manager takes into account the fact that each service contributes to the financial viability of the entire hospital; therefore, the purchase of certain "loss leaders" that benefit the entire hospital may be permitted (Steinberg et al., 1988~. Examples of such "loss leaders" are helicopter services, which are money-losing operations in nearly every circumstance, but that can bring visibility to the hospital and attract "profit- able" patients for other parts of the hospital (Anderson, 1990~.
64 GERARD F. ANDERSON AND EARL P. STEINBERG The technology competition model derives from three different theories of hospital behavior: the sales maximization theory (hospitals want to-be the larg- est); the conspicuous consumption theory (hospitals want to show that they are the most technologically advanced); and the physician cooperative theory (hospi- tals will acquire technology that maximizes physician income). In the technolo- gy competition model, physicians and potential patients are assumed to be at- tracted to new technologies and innovative medical practices. To obtain a competitive advantage over other hospitals, it is desirable to be the first hospital in a geographic area to acquire a new technology or to demonstrate a proficiency with a new medical procedure. Even for hospitals that do not strive to be trend- setters, it is still important to maintain technological parity with other hospitals. The third model of hospital behavior is the utility maximization model. Un- der this model, the hospital manager invests in technology, subject to a budget constraint, to enhance the quality or quantity of services that are provided. In this model, technology competes against other services, such as nursing, for a share of the hospital's budget, and new and established technologies are evaluated within this context. All three of these models may explain hospital managers' behavior to some extent. To assess how accurately these theoretical models predict hospital man- agers' actual behavior, it is useful to examine the factors that hospital managers profess to consider when making acquisition and utilization decisions. In chapter 3 of this volume, Luce and Brown reported on the results of their survey of factors that influence hospital managers. In an earlier volume in this series, Paul Griner, general director of the Strong Memorial Hospital, identified eight factors that influence the adoption of new technology: capital financing, hospital pay- ment methods, degree of regulation, degree of competition, hospital capacity, evidence of effectiveness, organizational arrangements, and the decisionmaking process (Griper, 1992~. Other hospital managers have conceptualized the technology acquisition pro- cess from more of a strategic planning perspective. In making technology acqui- sition and utilization decisions, they consider the need to improve existing clini- cal strengths, provide synergy with existing technologies, be consistent with the hospital's overall mission, minimize financial risk, and recognize the life span of a product. According to a survey of 524 health care managers in 1990, the following criteria were rated as "very important" by more than half of them: the ability to establish or expand services (85 percent), receipt of a return on invest- ment (71 percent), and the ability to reduce operating costs (67 percent). The enhanced image of the hospital (47 percent) and medical staff pressures (43 percent) were also cited frequently (Anderson, 1990~. A comparison of the economist's and manager's perspectives of what moti- vates the hospital manager suggests a number of commonalities. The profitabili- ty of the investment is an important consideration, although it must be viewed from the perspective of the entire institution. Hospitals compete to be the first to
ROLE OF THE HOSPITAL IN THE ACQUISITION OF TECHNOLOGY 65 acquire new technologies to enhance their market share, satisfy the medical staff, and improve the image of the hospital. Increasingly, it is necessary for hospital managers to make trade-offs between the acquisition of new technology and other competing demands for constrained resources. All of these considerations will affect a hospital manager's perspective when he or she evaluates a new technology. TECHNOLOGY ASSESSMENT: SOCIETAL VERSUS HOSPITAL PERSPECTIVE Several factors should be considered when performing any technology as- sessment (Fuchs and Garber, 19901. These include not only clinical consider- ations, such as safety, efficacy, and effectiveness, but also economic, legal, and ethical considerations as well as patient satisfaction and preferences. Tools capa- ble of measuring these endpoints have become increasingly sophisticated. Even so, it is critically important to keep in mind the fact that the weight placed on some of these dimensions may depend on the perspective from which the assess- ment is being conducted. In the past most technology assessments in the United States were conducted by government entities or academicians, who have tended to perform the assess- ment from the perspective of society. (In most other countries, the national government continues to sponsor technology assessments that are performed from the perspective of society.) These assessments tended to be performed long after the technologies had diffused widely. As we have discussed, however, recent public policy in the United States has attempted to induce hospitals to take greater responsibility for conducting tech- nology assessments from their own perspective. The Medicare prospective pay- ment system, for example, uses diagnosis-related groups (DRGs) to pay hospitals and, with few exceptions, DRG payments do not vary by type of technology used. As a result, the hospital is given the financial incentive to conduct technol- ogy assessments and evaluate practice patterns with the knowledge that the pay- ment they receive for hospitalization will not vary according to the types of technology that are used. If more technology assessments are to be conducted from the hospital's perspective, it is important to examine the potential implications of changing the assessment perspective from what benefits society generally to what benefits a specific hospital. In discussing these potential implications, our intention is not to imply that hospital managers do not consider society in making their deci- sions; rather, our objective is to point out how adoption of the hospital's perspec- tive might affect technological assessment and diffusion at the margin. Our views regarding the effect of changing the perspective from which a technology assessment is performed from a societal to a hospital perspective are as follows. Technology assessments conducted from the hospital perspective will:
66 GERARD F. ANDERSON AND EARL P. STEINBERG be performed earlier in the diffusion process, be more responsive to changes in medical knowledge and practice be more sensitive to local medical conditions, place less emphasis on long-term outcomes and total health care costs, place more emphasis on legal liability, and give less consideration to impacts on other providers. First, a shift to a hospital perspective is likely to result in the performance of technology assessments at an earlier point in the technology's development. To maintain a competitive advantage over other hospitals, hospital managers need to conduct their technology assessments at as early a stage as possible in the devel- opment of a new technology. As a result, they may feel compelled to perform technology assessments while the drug or device is still in clinical trials. In contrast, government is prone to wait until more "complete" information is avail- able before conducting an assessment. Second, assessments performed from a hospital perspective are more likely to be responsive to ongoing changes in medical knowledge and practice patterns. In part, this is because a hospital's process for performing an assessment is more streamlined and less bureaucratic; it can therefore respond to new information as soon as it becomes available. The capacity to respond quickly may determine whether a hospital succeeds or fails in attracting a substantial share of patients in a competitive market. Government, in contrast, will be more likely to wait until a general consensus has been reached before performing or revising an assessment. Third, in performing a technology assessment a hospital is more likely than government to consider the implications of local medical conditions, such as the strengths of the physicians on their staff, the institution's mission, the character- istics of the patients in the hospital's catchment area, and the behaviors of other hospitals in the geographic area. An assessment performed from a societal per- spective, in contrast, must consider issues from a more aggregated geographic perspective. Fourth, whereas an assessment from a societal perspective will tend to con- sider both long-term and short-term outcomes, hospital managers are likely to place greater emphasis on short-term outcomes than on long-term outcomes. For example, a hospital will tend to consider costs incurred during a hospital stay as opposed to the costs incurred over the long term. In addition, they will be less concerned about technologies that prevent readmissions, because under most pay- ment systems readmissions offer the hospital an opportunity for a second pay- ment. The hospital perspective thus could increase long-term health care costs as hospital managers focus on short-term cost implications. The outcomes emphasized in assessments performed from a societal per- spective may differ from those emphasized in technology assessments conducted from a hospital perspective in other ways as well. For example, a hospital may pay more attention to the impact of a technology on patient satisfaction and
ROLE OF THE HOSPITAL IN THE ACQUISITION OF TECHNOLOGY 67 quality of life than on long-run mortality or morbidity. A hospital is also more likely to consider carefully the malpractice and liability implications of ad-opting versus not adopting a new technology. In addition, indirect costs associated with morbidity, costs incurred by a patient outside the hospital setting, and intangible costs of suffering are more likely to be considered in an assessment from a societal perspective. Finally, hospital managers also may tend to be less concerned than society about the impact of their decisions on other providers. For example, one of hospitals' responses to the incentives created by the Medicare prospective pay- ment system was to shorten the length of stay. This increased the number and complexity of patients who were discharged to nursing homes and other provid- ers. Except to the extent that this may have affected hospitals' ability to dis- charge patients, the changes were of less concern to hospitals than they would be for an assessment conducted from a societal perspective. In addition, we believe the total cost of multiple hospitals performing inde- pendent technology assessments is likely to be greater than the cost of a single assessment conducted by a single entity. Although individual hospitals are un- likely to devote considerable resources to any single assessment, the effect of many hospitals performing assessments on the same technology could result in more total resources being devoted to technology assessment. In view of the cost of performing technology assessments, hospital managers are forming technolo- gy assessment consortia. Societal Versus Hospital Perspective: Some Examples What is the aggregate effect of these differences in perspective? Because no single technology assessment is likely to illustrate how all of these differences in perspective might manifest themselves, we believe there is value in considering how assessments of a couple of technologies might differ when performed from a hospital versus a societal perspective. The use of high-dose chemotherapy with bone marrow transplant for treat- ment of various types of malignancies, such as metastatic breast cancer, is a good example. Although few data are available and one cannot yet determine whether this treatment is effective in patients with metastatic breast cancer, many hospi- tals already offer this treatment. Their decision to offer the treatment is presum- ably based on their own consideration of the potential value of establishing them- selves as a leader in the adoption of new cancer treatments. When paid for, this treatment also generates substantial revenue, although little is known about the short- or long-term cost-effectiveness of this treatment. Even though sufficient data to evaluate this technology from a societal perspective are clearly not avail- able, some hospitals have made a decision, on the basis of what few data are available, to adopt this treatment. Their views regarding the appropriateness of continuing to offer this treatment may be revised several times as new data be
68 GERARD F. ANDERSON AND EARL P. STEINBERG come available, well before an assessment performed from a societal perspective is ever undertaken. The case of laparoscopic cholecystectomy is also illustrative. Early data have suggested that the short-term outcomes associated with this procedure are quite favorable compared with those associated with traditional cholecystectomy. Patients' lengths of hospital stay are shorter, short-term morbidity is lower, and return to work is reported to be earlier after laparoscopic cholecystectomy com- pared with that after traditional cholecystectomy. As a result, surgeons and hos- pitals have rushed to adopt this new procedure. Data regarding longer-term clinical outcomes and the costs associated with laporoscopic cholecystectomy are just now becoming available. These data sug- gest that the total costs associated with laparoscopic cholecystectomy, that is, payments to hospitals and physicians, may be higher than those associated with conventional cholecystectomy (Legorreta, 1993~. In addition, there is some indi- cation that rates of readmission for procedural complications were initially higher with laparoscopic cholecystectomy than with traditional cholecystectomy, per- haps as a result of the learning curve involved. Even though the latter data may decrease the attractiveness of this technology from a societal perspective, it may not lessen the attractiveness of the technology from a hospital's perspective. POLICY CONSIDERATIONS What, then, is the aggregate effect of these differences in assessment per- spective on the overall rate of diffusion of medical technology? We believe the answer is "very little to date." Prior to the implementation of prospective pay- ment, hospitals were rapid adopters of new technologies, primarily because sev- eral forces promoted the early adoption of new technology and the cost of using those technologies could be recovered easily. Few hospitals deferred acquisition decisions until a technology was mature and an assessment of it had been per- formed from a societal perspective. Although several changes in regulatory, legal, and payment policies that increase hospital managers' incentives to per- form technology assessments have been implemented over the past decade, we believe hospitals continue to be early adopters of new technologies, largely be- cause several forces promoting the early adoption of technology counterbalance the incentives to perform careful technology assessments. If policies that result- ed in the performance of more technology assessments from a societal perspec- tive were implemented, with acquisition decisions made on the basis of those assessments, then we believe the rate of adoption of new technology could slow substantially. If technology assessment conducted from a societal perspective is consid- ered to be the "gold standard," then it is worthwhile to consider various strategies to encourage the use of this perspective. One option is to alter the perspective of hospitals to give them more of an incentive to adopt a societal perspective. For
ROLE OF THE HOSPITAL IN THE ACQUISITION OF TECHNOLOGY 69 example, the Health Care Financing Administration's 30-day mortality index is an attempt to measure the impact of the hospital visit over a time per~od-longer than most hospital stays. Similarly, the emphasis of peer review organizations on readmission rates requires hospitals to take a longer-term perspective. Increased use of payment for a bundle of services or use of a capitated payment for a year could give hospitals more of an incentive to consider the impacts of their deci- sions on long-term outcomes, total health care costs, and other providers. In- creased competition between hospitals on the basis of data regarding their long- term outcomes or a focus on such outcomes by the Joint Commission on the Accreditation of Health Care Organizations might have similar effects. CONCLUSION Recent changes in health care financing have given hospitals more of an incentive to evaluate new technologies from their own perspective. In light of this trend, it is important to consider how this perspective might affect the diffu- sion of technology. We believe that technology assessment conducted from a hospital perspective instead of a societal perspective promotes the more rapid diffusion of medical technology, gives less weight to long-term outcomes and long-run health care costs, and increases the overall cost of conducting technolo- gy assessments. Public policy could mitigate some of these effects by establish- ing payment systems that emphasize total health care costs and information sys- tems that emphasize the longer-term impacts of different treatment modalities. REFERENCES Anderson, G. 1992. The courts and health policy: Strengths and limitations. Health Affairs 11 :95-110. Anderson, G., and Steinberg, E. 1984. To buy or not to buy: Technology acquisition under prospective payment. New England Journal of Medicine 312:1349-1353. Anderson, G., Lave, J., Russe, C., and Neuman, P. 1989. Providing Hospital Services. Baltimore, Md.: The Johns Hopkins University Press. Anderson, H. J. 1990. Survey identifies trends in equipment acquisitions. Hospitals 64:30-35. Darling v. Charleston Community Memorial Hospital, 33 Ill. 2d. 326, 211 N.E.2d. 253 (1965~. Davis, K., Anderson, G., Rowland, D., and Steinberg, E. 1990. Health Care Cost Con- tainment. Baltimore, Md.: The Johns Hopkins University Press. Feldstein, P. 1988. Health Care Economics. New York: John Wiley & Sons. Fuchs, V., and Garber, A. 1990. The new technology assessment. New England Journal of Medicine 323:673-677. Griner, P. 1992. New technology adoption in the hospital. In: Institute of Medicine. Medical Innovation at the Crossroads. Vol. 3, Technology and Health Care in an
70 GERARD F. ANDERSON AND EARL P. STEINBERG Era of Limits. A. C. Gelijns, ed. Washington, D.C.: 123-132. National Academy Press, pp. Hall, M., and Anderson, G. 1992. Health insurers' assessment of medical necessity. Pennsylvania Law Review 140(5~:1637-1712. Johnsson, J., Anderson, H. J., and Burke, M. 1991. Technology acquisition: Trends in imaging. Hospitals 65:26-36. Legorreta, A. P., Silber, J. H., Costantino, G. N., Kobylinski, R. W., and Zatz, S. L. 1993. Increased cholecystectomy rate after the introduction of laparoscopic cholecystec- tomy. Journal of the American Medical Association 270: 1429-1432; Lumsdon, K. 1992. Beyond technology assessment: Balancing needs, strategy. Hospi- tals 66:20-26. Steinberg, E. P., Stason, W. B., diMonda, R., and Schroeder, S. A. 1988. Determinants of acquisition of MR imaging units in an era of prospective payment. Radiology 168(1):265-270.
