The Impact of Economics on Changing Medical Technology with Reference to Critical Care Medicine in the United States
2003; Lippincott Williams & Wilkins; Volume: 96; Issue: 2 Linguagem: Inglês
10.1213/00000539-200302000-00023
ISSN1526-7598
Autores Tópico(s)Hemodynamic Monitoring and Therapy
ResumoHigh Costs of Health Care in the United States Exploding Health Care Costs Over the past 20 yr, health care expenditures in the United States have exploded, resulting in a 4.5-fold increase per year when adjusted for inflation and population growth. In 2000, 13.2% of the US gross national product was spent on health care (1), but life expectancy is not as long in the United States as in countries such as Japan, which spend much less of their domestic output on health care (Table 1) (2). Within the United States, there are also large discrepancies in life expectancies among racial groups (Fig. 1). One percent of the US gross national product is being spent on critical care services, accounting for roughly 5% of all patients cared for and 20% of all health care expenditures (3). Different states vary in amounts spent on health care (Fig. 2). The national per capita expenditure for health care amounted to $4637 in the United States in 2000 (1). As “baby boomers” age and require more health care resources, this figure could increase to as much as $25,000 by the year 2020.Table 1: Female and Male Life Expectancy at Birth, 1992Figure 1: US life expectancy at birth. From Kindig (2). Reprinted with permission from the University of Michigan Press.Figure 2: US per capita expenditure for health care for various individual states. From Kindig (2). Reprinted with permission from the University of Michigan Press.Health care economists such as Folland et al. (4) believe that the increasing costs of health care can be attributed to three factors: the moral hazard of insurance, increased technology, and asymmetry of information. Newhouse (5) describes five factors associated with increased health care expenditures: increased insurance usage, aging, increase in income, increases in supplier-induced demand, and miscellaneous factors producing problems in the service sector. The purpose of this article is to examine how these factors interact to drive critical care costs and to identify ways to make technology more cost-effective. We will consider the roots of the problem, the necessity for technology assessment, and possible solutions. Moral Hazard of Insurance As defined by Folland et al. (4), “moral hazard” is an insurance term that represents the disincentives created by insurance for individuals to take measures that would reduce the amount of care demanded. In the health services literature, it is more often used to express the additional quantity of health care demand resulting from a decrease in the net price of care attributable to insurance. The concept of moral hazard was advanced by the RAND health insurance experiment (5), which held technology constant so that the effect of co-payment of insurance on moral hazard could be sorted out. This study divided a patient population into five groups with different types of health coverage. Group 1 had a free plan, Group 2 had a plan with a 25% co-payment, Group 3 had a 50% co-payment, Group 4 had an individual deductible plan, and Group 5 had a 95% co-payment plan. The study showed 39% increased medical expenditures for the free plan compared with the 95% plan. The biggest difference was between the free plan and the 25% co-payment plan. The study did not include data on patients in intensive care units (ICUs), but the results clearly showed that if health care services were free, then patients would demand more. Technology The technological age has blossomed during the past 30 yr, with the development of the computer and the “high-tech” electronic industry. The field of biomedical engineering has helped to fuel the fire for research and development for technological advances. The American public as a whole perceives technology as something useful and beneficial. To health care workers, technology appears to play an ever-expanding role, encompassing all drugs, devices, processes, and procedures that enable us to deliver more sophisticated health care services. To health care economists such as Weisbrod (6), technology can be addressed on three levels, depending on what it achieves. 1) “Non-technology” helps prevent disease and is relatively inexpensive. An example of this would be the blood pressure cuff, which enables a relatively inexpensive determination of a patient’s physiologic status. 2) “Halfway” technology is supportive, but it does not cure the disease and is expensive. For example, patients with renal failure may remain on hemodialysis for extended periods and may need intensive care because of multiorgan system failure, but the outcome is usually death. Thus, the amount of economic resources expended tends to be large, but outcome is poor. 3) “High” technology is also expensive but addresses the disease mechanism directly. An example would be thrombolytic therapy for myocardial infarction, which has saved lives, although the expense is enormous compared with more conservative treatment. Scitovsky (7) has suggested that technological change has increased cost in many ways. There has been an emphasis on introducing costly technologies, but questions should be asked about which technologies increase costs and by how much. At issue is whether we can develop technologies that decrease the overall cost and increase the quality of care, rather than increasing costs without improving outcomes. Over the past 300 yr, human physiology has undergone changes induced by technology (8). Although there have been many improvements because of technology, according to the production curve defined by Kindig (2) (Fig. 3), if output were plotted on the y axis against the total amount of investments on the x axis, a second-order function would be observed, with a plateau showing decreased output as more dollars were invested in health care technology. In certain cases, investing more dollars in technology may even worsen outcome, for example, when iatrogenic complications are induced. Furthermore, in some circumstances, the technology itself may be the incentive to use more health care interventions, resulting in increased investments without a corresponding improvement in outcome.Figure 3: Health investment and health outcome. From Kindig (2). Reprinted with permission from the University of Michigan Press.Asymmetry of Information Folland et al. (4) define “asymmetric information” as “situations in which the parties on the opposite sides of a transaction have differing amounts of relevant information.” Quite often, patients do not have perfect information regarding their choices in health care. As providers, physicians also often lack perfect information as to optimal treatments for their patients’ illnesses. If adequate means are not available for evaluating outcomes, diagnostic and therapeutic techniques may be used with little outcome benefit and, in some cases, with high cost. Health care differs from other markets because it is concerned primarily with the preservation of life. It abounds with many informational asymmetries (6). Problem Analysis Interrelationship of the Moral Hazard of Insurance, Technology, and Health Care Costs Many nonmarket factors are important in the relationship between insurance, health care technology, and health care costs. The United States is the world leader in research and development of technology. At the same time, insurance and managed care are expanding. These factors are interdependent. Weisbrod (6) has argued that health insurance has induced a technological explosion. The moral hazard of insurance results in increased demand for technological development because low out-of-pocket costs for the insured induce increased use of health care services. The key point is that if insurance pays for the additional costs when new technology is introduced, then insurance will drive demand for technological development, leading to higher and higher costs. Thus, the question is whether insurance-driven incentives that increase the development of costly technologies can lead to economically efficient and life-saving developments in health care. A model to explain the association of technology, health care costs, and health insurance was developed by Goddeeris (9), who showed that decreasing the amount of a patient’s co-payment for health insurance tends to increase demand for medical care resources that use expensive forms of technology. Increased use of expensive technology leads to increased profits for the industry, and Goddeeris hypothesized that this factor is a major reason for the increase in technological development. Newhouse (5) also thought that insurance tends to induce increased use of technology and increased change, but that the marginal benefits of the increased technology do not outweigh the marginal costs. Accordingly, the availability of technology would influence the demand for health insurance. Because the inflation caused by technology is influenced by the amount of co-payment of insurance, greater insurance coverage would lead to increasing technology costs. Influence of Type of Reimbursement. The health care reimbursement system in the United States has changed from a retrospective fee-for-service system to a prospective, capitated system. The retrospective payment fee-for-service system has been blamed for fueling the fires of technology development and health care inflation. It was thought that a prospective payment system under managed care would slow this growth. Initially it did; however, managed care has failed to deliver its promise of controlling health care costs (10,11). A capitated health care system may not hold down health care costs that stem from increases in technology. Vicious Circle Resulting from Increased Demand for Health Care. A major question is how such changes in health care reimbursement influence technological development and health care costs. Schwartz (12) hypothesized that increased demand for health care stimulates the search for “magic pills” that would cure all medical ills. The fight against cancer proposed in the 1960s by President Lyndon B. Johnson, with the aim of finding a cure by 1969, represents an example. Such a search, in turn, can stimulate increased technological cost, which increases health insurance costs. Thus, a vicious circle is created in which increased demands for insurance, health care, technological services, and the magic pill are intertwined, ultimately increasing the costs of health care services for society. Even when progress is made, if the magic pill is never found, this phenomenon is equivalent to the top portion of the production function curve defined by Kindig (2), in which a decrease in health care output occurs with increased investments over time (Fig. 3). Lack of Free Market Competition Managed care came of age with the idea of controlling both health care costs and moral hazard. It was believed that free-market competition could hold down managed care prices so that marginal costs would equal marginal revenues for most goods and services. However, certain conditions must be met for a perfect competitive market to exist. Arrow (13) lists the following five factors. 1) There are sufficient buyers and sellers of each economic good such that no single player has any power over the price. 2) The good is homogeneous; all producers produce exactly the same good so that the market cannot be segmented on the basis of different goods. 3) There is perfect information; all buyers and sellers have reliable information on all relevant variables, such as prices and quantities. 4) There are no barriers to entry and exit as producers start producing items on terms that are equivalent to those already in the industry. 5) There are no significant externalities or public goods; the goods must not produce any secondary benefits or costs to others. Obviously, the health care market does not meet these criteria, particularly with regard to technology. In an analogy to economic systems, patients would be thought of as “buyers” and hospitals and physicians as “sellers.” (The analogy is, of course, limited. Patients are not simply buyers, and physicians are also patient advocates and, thus, not true sellers.) Concerning the first point, buyers of health care (patients) are not actually looking to buy new technology as such. With the exception of plastic and cosmetic procedures, they want to buy health and health care services; but health care technology and health care services are not the same. New technology cannot always give consumers the outcomes they desire. Second, health care markets are definitely not homogeneous. There is not a “Super Clinic” on every street corner. People do not have access to the same quality or quantity of care around the country. Third, there is a lack of information on both sides. Often, the buyers (patients) do not or cannot fully understand what they are buying and make decisions on emotional rather than rational grounds, for example, because they fear loss of life. Sellers (physicians or hospitals) often ignore costs because they feel obligated to give what they consider the highest quality of care to patients. Because the true chances for a positive outcome are unknown, this often involves using technology that has little benefit, and in some cases with high cost. Thus, marginal costs exceed marginal benefits. This may even lead to so-called supplier-induced demand—that is, greater supply might induce physicians to use higher-cost technology items in an effort to deliver higher quality of care. Use of high-tech items would lead to the filling of beds. Even though prices may decrease through increased supply, the demand for service increases dramatically, offsetting the price base and leading to higher overall costs. Fourth, there are barriers to entry for the uninsured. Lack of access to forms of health care that entail use of expensive technology items results partly from the widening socioeconomic gradient within this country. Fifth, there are significant externalities for health care. The cost of medical care for the uninsured is often passed on to the insured in the form of more expensive health care premiums. In particular, the use of “halfway” technology that prolongs life and is costly is paid for by those with insurance for those who do not have insurance. Thus, we do not have a perfectly competitive market for health care. Despite the tendency of health maintenance organizations (HMOs) to control costs by decreasing the use and quality of care (11), there is no evidence that HMOs have underused technology in their efforts to stay competitive. This has become evident as managed health care costs have skyrocketed, for example, in Minnesota, where HMOs are prevalent (10). Although managed care there did tend to hold down health care costs in the past, a dramatic increase in health care costs has occurred recently as a result of the use of technology. Cost of Technology Versus Value Thamer et al. (14) showed that low-cost technology items can account for a considerable proportion of hospital costs. Shine (15) discussed the difficulties that arise when low unit costs of technology increase demand, thus shifting the demand curve upward and increasing total costs. Shine therefore recommended that not only the cost, but also the value, of the technology be evaluated. As defined by most economists, value is equal to quality divided by cost. By assessing value instead of looking solely at the cost of technology, an optimal production function curve can be obtained in terms of what is the best quality for a limited cost. Traverso (16) has discussed the concept of value in technology to differentiate a true advance from a “gimmick.” A gimmick affords no increase in value—only a possible increase in cost and possibly a decrease in quality, setting in motion the vicious circle of increasing health care costs due to increasing technology, with no beneficial output quality. However, if new technology does increase quality, then it will increase value and may, as a consequence, decrease cost. Supply-and-Demand Factors Other factors associated with increased health care expenditures are related to supply and demand. For example, according to Newhouse (5), increasing the number of physicians may play a modest role in increasing health care costs because with more physicians there would be more use of technology, which in turn would increase the cost of health care. Asymmetry of Information Hospitals and physicians strive to remain competitive in the market by acquiring the latest in technology (17). In the United States, advertising and marketing to the public probably enhance this. The public sees that Hospital A has a new gimmick in comparison with Hospital B. This may lead to increased purchases and increased costs. This results in information asymmetry because hospitals often do not have any means to evaluate outcomes of the different forms of technology (18). To provide a basis for decision-making, new technology should be assessed properly before its implementation. With respect to high-cost services such as critical care, fair resource allocation based on outcome data is needed. Thus, to achieve an optimal production function curve, a new procedure should be evaluated as to what it achieves (i.e., whether it is “non-technology,” “halfway technology,” or “high technology”) and should be rated on a value scale. Thus, technology assessment is needed not only to improve cost-effectiveness and cost-benefit ratios but also to improve quality of health care. Technology Assessment in Critical Care Medicine Technology assessment in the critical care environment can encompass all diagnostic and therapeutic devices, medications, information systems, procedures, and clinical and research protocols (19). In some countries, such as New Zealand, technology is looked at carefully before it is implemented (20). In others, such as England, underfunding of the national health care system may lead to underutilization of technology that could be beneficial to patients (21). In the United States, we may have two extremes to deal with. The fee-for-service system has fueled part of the technology explosion in critical care medicine (3). The pendulum tried to swing the other way with the capitation of care, which may in turn lead to underutilization (3). At present, hospitals in the United States have little incentive to look at technology assessment for their ICUs. For example, a well known technology device in the ICU is the pulse oximeter. There is no good model for assessing cost-benefit and cost-effectiveness ratios for pulse oximetry (22). The Swan-Ganz catheter is another example. There is no good scientific evidence from randomized trials to guide practice or to assess the cost-effectiveness of its use (23). Physician assessment of the effect of right-heart catheterization in treatment decisions and patient outcomes has not been deemed reliable (24). Uniform Standards to Guide Resource Allocation One reason no uniform technology assessment system has been instituted is the existence of differences in the epidemiology of critical care (25). It can be hard to compare differences in outcomes from different ICUs in the same country (26), and comparisons between different countries are even more problematic. As we become more interested in quality improvement in medicine, we should embrace principles calling for more standardization of delivery of medical and nursing care (27). Standardized approaches to such issues as using technology in critical care, dealing with nursing shortages, and assessing outcomes can all play a role in controlling costs. Standardized delivery of health care in the ICU should be by individuals trained in critical care medicine, but currently only one third of patients in ICUs are cared for by physicians with critical care training (27). It seems ironic that ICU care in this country should be delivered by individuals who are not trained in critical care medicine and who are using high-cost equipment that may not provide any real benefit for patients. The question arises as to whether and how we can best assess technology and develop uniform standards for resource allocation. Developing uniform standards for resource allocation has hints of rationing. Although rationing of health care is not something anyone wants to do, as Kindig (2) points out, we already have a form of rationing. The goal must be to make resource allocation as fair as possible. Guidelines for Technology Assessment in Critical Care In 1995, the Society of Critical Care Medicine (SCCM) established a critical care coalition, with several suggestions for assessing monitoring devices in critical care (28). Monitoring devices are a very big, essential part of any ICU environment. The recommendations called for 1) product guidelines developed by the Food and Drug Administration for technology, in particular for monitoring devices in the ICU; 2) more multidisciplinary research; 3) commonly accepted standards for technology assessment, identified by the SCCM; and 4) an independent council of clinicians to be made available to manufacturers for recommendations on research and development. These suggestions could be particularly useful when devices are in the early stage of development. The SCCM has also suggested the development of a closed-loop information system (29), which could provide a system of evaluation based on human outcome trials before release of the device. Although prognostic scoring systems cannot predict outcome for a given individual, they represent a valuable tool for evaluating outcome when assessing new technology, particularly if they address morbidity and quality of life rather than mortality only. Prognostic scoring systems have been around for the past 20 yr. One of the most noted has been the APACHE (Apache Physiologic and Chronic Health Evaluation) system developed by Knaus et al. (30,31). One strong piece of evidence that APACHE has given us is that patients can have different outcomes in different ICU settings (26). This speaks to quality of care issues that could be addressed by greater standardization of care. However, no one scoring system has been validated for care. Existing systems still need work. Solutions The solution to increasing costs from health care technology will be very difficult to find, especially with regard to critical care expenditures. A market course alone, through a system of managed competition, has not been able to curtail increasing technology costs in health care. As a matter of fact, managed competition never was able to get off the ground. We are at present in the scenario described by Schwartz (12), in which the cost of technology continues to increase and probably will do so into the next decade. The five factors that Newhouse (5) described (increasing use of insurance, aging, increasing income, supplier-induced demand, and factor productivity problems in the service sector) have all come into play to fuel costs in critical care. Sibbald et al. (32) evaluated the effect of technology on costs in critical care medicine and called for 1) assessment of which patients would benefit and at what cost, 2) determination of technology that would increase benefit as judged by total quality improvement guidelines, and 3) evaluation of benefits of new technologies by health care funding agencies. In essence, this echoes from a physician’s standpoint the views put forth by Newhouse (5) from an economic standpoint. Marsden (3) listed the following recommendations: 1) improve the ability of critical care medicine to evaluate technology, 2) encourage the development of national and international bodies with a specific mandate for technology assessment, 3) establish funding sources for technology evaluation, and 4) encourage hospitals and physicians to develop strong liaisons between health care and industry. Restraining Demand for Health Care In addition to maximizing the returns from technology, Goddeeris (9) suggested that by increasing the amount of co-payment that the consumer has to pay for health care insurance, one could see a readily obtainable isoprofit line. Furthermore, because insurance benefits and payments to employees are made with pretax dollars, changing tax codes so that insurance benefits were paid with posttax dollars may also decrease moral hazard and thereby decrease the demand on technology. Such measures may provide more incentive for choosing health care services wisely. However, insurance coverage must be portable and guaranteed. Redirecting younger generations of health care seekers toward medical savings accounts should also decrease the inelastic demand for health care and help to decrease demands for technology, thereby decreasing costs. The problem of the increasing demand for health care caused by the increasing age of the population cannot be solved at present. Only through preventive services in a younger generation can this be changed for future generations. A further aspect relates to supplier-induced demand. Physicians tend to use expensive items in the—sometimes false—hope of providing better outcomes. If thorough assessment of technology took place and information technology provided easy access to outcome information, inappropriate use of high-end technologies that have little benefit could be avoided (33). Insurance companies should work with physicians to evaluate high-cost technologies to see whether their marginal benefits will outweigh their marginal costs. Redirecting Technology Technology needs to be redirected to the point where the marginal benefits of technological change are equal to or greater than the marginal costs. Rather than attempting solely to decrease health care costs, we need to streamline technology into development of what will give us better health. Because information is such a vital necessity in making appropriate medical care decisions, technology at present should be redirected to improve information systems (34). It is not uncommon in other industries for 7% of budgets to be spent on information technology, and information technology is known to decrease transaction costs. Yet information technology is still underused within the health care industry, and few health care dollars are spent in this arena. At present, improving information technology will probably give us the best marginal benefit at the best marginal cost. Electronic patient records should help to improve efficiency of health care, but the validity and usefulness of such information systems need to be tested (35,36). Improvements in information technology should also facilitate evaluation of the outcome of other technology, so that we can then look at curtailing health care costs from more expensive items. Ethical Issues How can we reconcile these issues with the major ethical issues in critical care medicine? Emphasis should always be placed on giving the best care possible. Issues to address are 1) foregoing life-sustaining technology (37), 2) recognizing the importance of preventing patient dehumanization, and 3) allocating resources in the ICU (3). Regarding the allocation of resources, in view of the aging of our population in critical care, preventive health care will be our best alternative, coupled with validated prognostic scoring systems. Patients need to be aware of their choices and outcomes. Things to help them could be 1) congressional passage of a self-determination act, 2) development and acceptance of validated prognostic scoring systems for critically ill patients, and 3) possibly transferring chronically ill, ventilated patients to skilled nursing facilities that might meet needs appropriately at lower cost rather than simply shifting health care costs to another facility. For example, Zimmerman et al. (38) have developed markers to identify patients who may benefit from transfers to more cost-effective step-down units. Conclusions Controlling health care costs that arise from increasing technological costs is a multifaceted problem. In view of the models showing that higher insurance coverage leads to increasing costs of technology (9), one part of the solution should be the institution of medical savings accounts that would increase co-payments for individuals, thus potentially decreasing moral hazard and decreasing the demand for higher-cost technological items. As the second part of the solution, health care insurance premiums should be taxed, and the tax revenue from this could be further used to provide coverage for those who are uninsured. Managed care organizations are not controlling costs through competition. There is no true competition, and technology is driving health care costs. Instead of rationing care or giving up totally and having the government nationalize the health care system, a system of assessment of all technologies must be implemented to allow development of what might be called “value” technologies, which can be demonstrated to give the highest quality benefit at the lowest marginal costs. To achieve this goal, technology at the present moment should be more oriented toward health information systems. Information is the key to solving many of our problems. Information technology could give us a better production function and in turn lead to a better marginal benefit to outweigh the marginal costs from increases in technology. However, we must also validate this step to see whether it will enable us to provide more cost-efficient and cost-beneficial health care services. Finally, we must go back to the basics. If we want to solve the problem through free-market competition, then we must satisfy the criteria for a free health care market. The main stumbling blocks are lack of production of a homogeneous good, lack of adequate information, and lack of equal access to health care. The effects of moral hazard and information asymmetry have not been looked at with reference to the economics of spending in critical care medicine. This may be difficult to do. Patients need to be better educated about all of their options, with definitive outcome data. Technology assessment is necessary to ensure that all of our patients receive the best, most cost-effective, and most cost-beneficial care that we can offer in this world of limited resources.
Referência(s)