Click for next page ( 9


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 8
1 Rationale for Assessment of Diagnostic Technology Over the past 15 years, there has been rapid growth in the use of innovative diagnostic technologies, such as digital radiography, ultra- sound, computed tomography (CI), and magnetic resonance imaging (MRI) (Hickman and Schwartz 1985~. The use of diagnostic tests in general is increasing even more rapidly than the cost of medical care (Schroeder 1981~. Furthermore, there is evidence that diagnostic tech- nologies consume a significant portion of our health care resources. For example, Scitovsky has estimated that of aU expenditures for outpatient care in 1975, 29 percent were for laboratory tests and X rays (Scitovsky 1979). In medicine, the primary goal of new technology is to improve the quality of care. Nevertheless, the recent past has been marked by early diffusion of new technologies without adequate measurement of their effects on the quality of care (Abram s and McNeil 1978a,b; Guyatt et al. 1986~. A number of authors have pointed to the lack of prospective, controlled studies (Harris 1981, Sheps and Schechter 1984, Schwartz 1986, Cooper et al. 1988, Kent and Larson 1988~. Although individual diagnostic tests have been studied at length for accuracy, direct compar~- sons of a new test with an older test have been all too infrequent (McNeil et al. 1981, Abram s et al. 1982, Hessel et al. 1982, Alderson et al. 1983, Inouye and Sox 19861. What are the consequences of Me medical profes- sion's failure to evaluate diagnostic tests in a timely and rigorous manner? Why is such an evaluation important? 8

OCR for page 8
RATIONALE 9 These questions may be approached from many perspectives, but Tree are especially relevant: that of the patient, that of society, and that of the physician. We win consider each of these perspectives in terms of safety, efficacy, and economic efficiency aU critical standards for a diagnostic test, but standards whose relative importance depends upon the perspec- tive under consideration. THE PATIENT'S PERSPECTIVE The impact of a diagnostic test is felt most directly by the patient. The patient's concern is improving his or her health status. From the patient's perspective, the most important characteristics of a diagnostic test are its safety and efficacy. The economic aspect of testing may also assume importance, but it is often a secondary consideration. Safety A safe test does not cause an unacceptable degree of direct harm to the patient. It is important to the patient that the diagnosis be made with a minimum of inconvenience and discomfort. If a test is hazardous, he or she win be directly at risk. Obviously, some "safe" tests win have side effects, and the acceptability of such effects wild be determined by weigh- ing their severity against the need for the information that the test pro- vides. Efficacy After safety, the starting point for any assessment of health care tech- nology must be efficacy or effectiveness. Ef5ficacy refers to the potential benefit to patients in a defined population when a test is applied to a specified problem under ideal conditions of use (WiBems et al. 19771. Electiveness is measured under the usual conditions of medical practice. This distinction is important when designing a technology assessment and interpreting its result; in the present discussion, however, we win refer only to efficacy. The efficacy of a diagnostic test should be measured in terms of the test's safety, its technical quality, its accuracy, its therapeutic impact, and its impact on the health of the patient (Fineberg et al. 1977~. The first step in an efficacy assessment is to determine the test's technical capability. Does the test meet the standards attributed to it? For

OCR for page 8
10 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY a diagnostic imaging method, the first stage of assessment might involve using cadaver specimens to see how wed the technique is able to demon- strate the anatomy of various regions in the body. The second stage of an efficacy assessment is to define the test's diagnostic accuracy. In this regard, settee commonly used expressions are true-positive rate, true- negative rate, and accuracy. The true-posit~ve rate, or sensitivity, is a measure of the test's ability to detect disease correctly when it is present. The true-negative rate, or specificity, measures the test's ability to ex- clude disease in those patients who do not have it. Accuracy is the proportion of test results that are correct (true-posit~ve results plus true- negative results divided by the total number of test results) when the test is used in a specified population. Thus, it is a reflection of both the sensitivity and the specificity of the test. To can a test result a true positive or a true negative, one must deter- mine the true state of the patient. This is usually accomplished by doing another test, called the "gold standard," which is considered sufficiently reliable to reveal He true state of the patient, and either confirm or refute Me study test result. For example, coronary angiography has been used to verify the presence of coronary artery disease in patients participating in an efficacy study of the stress electrocardiogram. For an ideal test, Were should be little disagreement between its result and the result of the "gold standard": the test should have bow high sensitivity and high specificity. Although quantitative measures of test performance are important, a study of efficacy should not focus solely on its technical aspects (that is, on the machine). Rawer, an assessment should include data on diagnostic impact and on therapeutic impact, including outcomes that are relevant to the patient. These are the third and fourth levels of an efficacy assessment. The following questions might be asked in the third stage of an assess- ment. Does the result of the technique change the diagnosis? Does the technique add clinically significant information? At the fours level of an efficacy assessment, the question would be: Is the diagnostic impact one that changes the management of the patient? These two levels are often, although not always, interrelated. For example, MRI has an unparalleled capacity to detect demyelinating le- sions in the brains of patients with multiple sclerosis. MR! is therefore an extremely important too] for assessing prognosis and also for evaluating new therapeutic approaches as Hey are developed. Nevertheless, there is currently no effective therapy for this disease. The results from a study that focuses solely on He efficacy of HI in assessing the progression of multiple sclerosis would have limited clinical use, because knowledge of the extent of the lesions does not currently play a part in deciding which

OCR for page 8
RATIONS ~ 1 therapy the patient should receive (F;eeny et al. 1986). Many studies of diagnostic tests fad] to consider the impact of a test result on patient management and clinical outcome. In this sense, the patient's perspec- tive is not always fully reflected in contemporary technology assessment. Economic Issues Although the primary concern of the patient may not be financial, the cost of testing and of receiving high-quality medical care is clearly ~mpor- tant to them. Unnecessary tests, inappropriate treatment, and disability may all prove very expensive. Although most patients are somewhat insulated from the costs of their care by third-party payment mechanisms, they are still usually responsible for paying a deductible portion of their care, that is, a fixed amount that must be paid before insurance coverage begins. This may be quite large in some types of insurance. In addition, some individuals have limited insurance coverage, and 35 million people in the United States have no health insurance (Annas 19861. The RAND health insurance experiment demonstrated that paying patients are sensi- live to the costs of health care and that individuals responsible for almost ad of Heir health care costs incurred expenses about 50 percent less than individuals receiving free care (Newhouse et al. 19811. The Benefits of High-Quality Studies There are many potential benefits for the patient of a test that has been shown to be both safe and effective. By improving the quality of diagnos- tic information, the test result may lead to timely, correct therapy. Earlier diagnosis may result in reduced or postponed morbidity and mortality. Tests may resolve uncertainty (Abrams and McNeil 1978by, reassure both the padent (Sox et al. 1981) and the physician, and increase the patient's confidence in the physician (Marton et al. 1982~. Comprehensive assess- ment provides the patient with information about the accuracy of the test, its risks, and the monetary costs associated with its use. With this infor- mation, He patient is better able to make an informed decision about accepting a recommendation to undergo a diagnostic procedure. The Risks of Poor or Absent Assessment Data Tests are often used when there is no proof of efficacy. Patients may suffer in several ways when their diagnosis depends on a test lacking established validity and known false-positive and false-negative rates.

OCR for page 8
12 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY Tests can lead to unnecessary confirmatory tests or to incorrect ther- apy. Consider the outcome of using a test that has a high false-positive rate How specificity): a large proportion of "normal" individuals may mistakenly receive treatment for the condition in question or may have to undergo further tests to define their true state. They may suffer direct harm from the unnecessary diagnostic interventions or treannent and may experience considerable anxiety. In me initial assessments of radionu- clide ventnculography as a test for coronary artery disease, bias in me selection of study patients led to an overly optunishc conclusion about the test's performance. The test was later found to be far less specific (49 percent) than early studies had shown (93 percent) (Rozanski et al. 1983~. This high false-positive rate may have caused many patients without coronary artery disease to be referred for an unnecessary invasive angi- ographic procedure. Tests can also cause harm through false reassurance. A methodologi- caDy flawed assessment may lead to the conclusion that a test has a much lower false-negative rate than is really true. Many patients who have a disease win have a negative result and be told that the disease is absent. Hey win then have a false sense of confidence, treatment may not be started, and the disease may advance beyond the point where it can be cured. Patients with suspected colorectal cancer could have suffered if Heir physicians had used a normal carcinoembryonic antigen (CEA) level to conclude that there was no cancer. This test had been used widely in the initial diagnosis of colorectal cancer, but later studies showed that CEA levels are often normal in the early stages of the disease. The test may detect as few as 30 percent of patients with early colorectal cancer (Fletcher 1986)e Even if a test provides accurate information, the test result may have no impact on the therapeutic plan or on patient outcome. For example, randomized trials of emergency endoscopy for patients with upper gastro- intestinal bleeding have shown that endoscopy provides additional diag- nostic information, but that this information does not alter surgical rates, length of hospital stay, or patient mortality (Peterson et al. 1981, Dron- field et al. 1982~. On balance, the examination cannot be considered beneficial. Summary: The Patient's Perspective For the patient, the most important attributes of a diagnostic test are its safety and clinical efficacy. A safe, efficacious test should reveal He true

OCR for page 8
RATIONALE 13 state of the patient, with a minimum of inconvenience and adverse effects. The benefits of a true-positive finding include timely diagnosis and treat- ment. The benefits of a true-negative include reassurance and protection from unnecessary treatment and further procedures. A correct test result should increase patients' confidence in their care. Patients may also be very concerned about Me costs of testing, particularly if they do not have health insurance. The role of good technology assessment is to identify tests Cat are reliable, that provide useful incremental information, and that may have a positive effect on patient outcome, as wed as to single out those tests that fail to measure up to these standards. SOCIETY'S PERSPECTIVE Safety and Efficacy Society's concern about the safety and efficacy of technology has two components. First, our government's role as "guardian of the public safety" (Foote 1987 - traditionally quite limited in the past has been expanded over recent years. The power of the Food and Drug A~ninistra- tion to regulate the safety and efficacy of drugs and, more recently, the safety and efficacy of medical devices is evidence of this expanded obligation (Foote 19861. Thus, society's interest in safety and efficacy stems pardy from a perceived ethical duty. A second component of this interest is economic. Society must be concerned about test efficacy because tests that are not efficacious are not likely to be economically efficient. Tests with high false-positive rates (low specificity) expose patients to the risk of a needless workup and increase the direct costs for their care. Missed diagnoses, which result from the use of tests win a high false-negative rate, may increase total health care costs if treatment that is begun at a later stage of the illness consumes more resources. Society's concern about the safety and effi- cacy of technology is matched by an equally strong interest in economic efficiency. Economic Issues Many analysts agree that new medical technology has been an impor- tant factor in the rise of health care costs, although they disagree about the magnitude of its contribution (WaIdman 1972, Feeny et al. 1986~. The Office of Technology Assessment (OTA) has estimated that the technol-

OCR for page 8
14 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY ogy components of care are responsible for nearly 30 percent of We rise In Medicare costs for the period between 1977 and 1982 (OTA 1984~. Because resources are limited, and because government at all levels has assumed a larger role in paying for health care, society has an interest In assuring that the available resources are utilized efficiently. A cost-effectiveness analysis is Me method used to measure the effi- ciency with which dollars are translated into health outcomes. The OTA defines cost-effectiveness analysis as a comparison of the positive and negative consequences of using alternative technologies (OTA 1980a,b). The key to cost-effectiveness analysis is that it is comparative. It com- pares the cost and outcome of using one test for a diagnostic problem with the cost and outcome of using another test. If there is no existing diagnostic technology, the new technology can be compared to doing nothing. The results of a cost-effectiveness analysis are usually expressed as the cost per unit of outcome (average cost-effectiveness) or the change in costs per change in unit of outcome (marginal cost-effectiveness). Compared win existing technology, an efficient new technology would achieve the same outcome at a lower cost, a better outcome at the same cost, or a better outcome at a lower cost. From the societal perspective' cost-effectiveness analyses help poli- cymakers decide which technologies should be encouraged by reimburse- ment policy. In principle, given a fixed budget, the use of this approach to allocate resources to programs results in the greatest impact on clinical outcome. A complete cost-effectiveness analysis will measure monetary and other costs, the effectiveness-of the technology in achieving its intended objectives, and the positive and negative effects from bow intended and unintended consequences (A~nstein 19771. Without com- plete assessments and physician education, certain technologies will be overutilized, while others may be underutilized (Abrams 1979, Doubilet and Abrams 1984~. The net result is wasted resources and lost opportuni- ties. Summary: Society's Perspective Society necessarily has a deep interest in the costs of diagnostic tech- nology, although safety and efficacy are also important. As government has assumed a much larger role in financing heady care, the importance of efficient use of diagnostic procedures has also grown. Policymakers need accurate information about which technologies consume the least re- sources for a given outcome so that they can allocate limited health care

OCR for page 8
RATIONALE 15 resources. Nevertheless, the first considerations in deciding whether a diagnostic technology is a good societal investment must always be safety and efficacy. THE PHYSICIAN'S PERSPECTIVE The importance of diagnostic tests and technology assessment to the physician must be examined in the context of the physician's dual role as the agent of both the patient and society. Safety and Efficacy The physician's role with respect to a patient is traditionally that of a fiduciary: the principal (the patient) entrusts the fiduciary (the physician) with the power to act on his behalf. For example, a patient usually undergoes a diagnostic procedure at the request of a physician, and therefore both have the same interest In knowing that a test is bow safe and efficacious. Tests are done when the pa-dent's history is consistent with a particular illness but the true disease state remains uncertain. The purpose of a diagnostic test in this coccal setting is twofold. First, it should provide reliable information about the patient's condition. Sec- ond, the result of the test should influence the physician's plan for manag- ing the patient. A test can serve these functions only if the physician knows how to interpret its result. Adequate assessment of diagnostic technology is important to cl~- cians because it provides the data needed to interpret test results. The result of a test whose sensitivity and specificity have been measured reliably can be used as the basis for sound clinical decisions. For ex- ample, in cases where there is an effective treatment for a disease, a positive test result may raise the probability of disease sufficiently to convince the clinician to start treatment. Tests that have not been adequately assessed are not as useful to the physician because the meaning of their results is ambiguous. For example, if a test result is negative, should the physician trust that result and assume that the disease is not present? When the sensitivity of a test is unknown, the physician has no way of knowing the proportion of patients who have the disease despite a negative result. Frequently, the false- negative rate and false-positive rate of a test are stated but are inaccurate. The physician may believe that a second test is needed when it is not, or may think that it is unnecessary when in fact it should be done.

OCR for page 8
16 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY Without comparative assessment data, a physician cannot be sure whether a new test should replace an older test, should be used in conjunc- tion with the older test, or should not be adopted at ah. The routine use of intrapartum electronic fetal monitoring (EFM) in place of periodic auscul- tation for ah delivenes provides an example. Because EFM is a very accurate diagnostic tool, it met win early acceptance. Its use, however, may lead to a higher rate of operative delivenes. -For low-nsk pregnan- cies, periodic auscultation provides adequate information with fewer ad- verse side effects (Thacker 1987~. Similarly, a recent comparative trial has demonstrated that even in high-r~sk pregnancies, auscultation yields equivalent results with EM in terms of both the fetus and the mother Mushy et al. 1987~. Comparative assessments, such as that conducted by Luthy et al., along with physician education, are needed to prevent irra- tional and inefficient use of diagnostic technology. Ideally, the interests of the physician and the patient can be equated, because, In principle, the physician is acting solely for the benefit of the individual patient. In addition to patient benefit, however, physicians may also be concerned about the financial and legal repercussions of diagnostic or therapeutic errors. A good deal of '`defensive" medicine is practiced with consequent overutilization~ecause of the fear that the omission of a diagnostic test may be construed as malpractice. Errors might be caused by using a test whose efficacy is uncertain. A test Mat yields many false-negative results may lead to missed diagnoses. One that yields many false-positive results may lead to excessively complex workups with untoward effects. Thus, a new technology that has made its way into clinical practice without adequate assessment may adversely affect the health care pro- vider as well as the patient. Conversely, when a thorough assessment indicates that a test is both highly sensitive and specific, both the physi- cian and the patient benefit. Economic Issues Physicians as well as hospitals must be concerned about economic efficiency. In an increasingly competitive environment, economic suc- cess will depend on efficiency as well as on quality of care. Adopting a new technology may attract more patients and increase a physician's competitive advantage. Some new technologies may reduce the costs of providing health care, but many new technologies are more expensive than Hose they are designed to replace. The cost of diagnostic technology

OCR for page 8
RATIONALE 17 is clearly of concern to physicians practicing within an organization where income is based on a fixed payment per patient rather than fee-for- service, because they may be at direct financial risk for expenditures over this fixed amount. The ultun ate net effect of a new technology cannot be assumed to be beneficial; once clinical efficacy has been established, the cost-effectiveness of the technology should be evaluated. These com- parative assessments should then influence hospitals' decisions to pur- chase new technology, as well as physicians' decisions to use it. Economic efficiency is important to the physician at yet another level. After the primary responsibility to the patient, the physician also has a societal obligation to help to contain the costs of health care. Physicians play the principal role in controlling the services patients receive, and they have a large influence on aggregate health care expenditures. Be- cause diagnostic technologies are an important component of these costs, clinicians can exert their influence by using diagnostic technologies effi- ciently. Without high-quality technology assessments, practice habits may change inappropriately. For example, a new, expensive diagnostic method may replace an older, less expensive but equally efficacious technol- ogy. The case of intrapartum EFM in place of periodic auscultation of the fetal heart is again illustrative. Initial studies of EFM documented its high level of technical and diagnostic accuracy and suggested that its use was associated with a reduction in pennatal morbidity and mortality. Wide- spread diffusion of this costly technology followed. Nevertheless, two recent cntical examinations of the literature on EFM concluded that there is little rigorous evidence that routine use of the method leads to a beneficial impact on patient outcome. The conclusion is that EFM should have been thoroughly evaluated by comparative studies at an early stage of its diffusion into routine practice before it replaced the less costly alternative of traditional auscultation (Shy et al. 1987, Thacker 1987~. Summary: The Physician's Perspective This perspective reflects the concerns of bow the patient and the society, because the physician serves as a crucial link between the two. For example, physicians play a large role in controlling the flow of society's health care resources. Our society expects physicians to make responsible decisions on how often and under what circumstances expen- sive diagnostic technologies win be used. To this end, they must have reliable, accurate, and comparative information on test performance. In

OCR for page 8
18 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY addition, accurate tests help to avoid errors that could lead to legal and financial difficulties. Economic efficiency is an important personal con- ce~n for physicians as well, and those In private practice or in large hospitals that use diagnostic technology inefficiently may be unable to compete effectively in the health care market. Most important, the physi- cian needs information about diagnostic technology in order to provide high-quality health care to every patient. Technology assessment is one mechanism for obtaining this information. SUMMARY High-quality, timely assessment is the prerequisite for the safe, eff1ca- cious, and economically efficient use of diagnostic technology. The data to be obtained win depend on the perspective adopted for the purpose of the assessment. Nevertheless, the three pardes most directly affected by the use of diagnostic technology share many of the same concerns. Although there is a well-def~ned methodology for assessing diagnostic technology, few studies have satisfied aU of the methodological criteria. Even well-designed studies have encountered problems in the course of collecting primary data. The following examples should serve to illustrate this point. In a cooperative study of computed tomography (Cl ) and radionu- clide ~ studies on the brain, data were collected by five hospices over a five-year period. Of 3,000 patients who entered into the study, only 136 patients had technically adequate and available Or and RN studies Hat could then be used in the final data analysis (McNeil 1979~. One of the few prospective, comparative studies of diagnostic imag- ing techniques to date ran into similar difficulties. The authors of the cooperative study on computed tomography, ultrasound, and gallium imaging in patients with fever stated: "We spent 17 months collecting data from two major teaching institutions [the Peter Bent Brigham and Johns Hopkins hospitals]. Fug-time research assistants at each institution tried to obtain aD cases in random order. Yet even this concerted effort produced only 156 cases, and then only 50 percent of them included objective proof of disease" (McNeil et al. 19811. Another class of problems is highlighted in Philbrick's analysis of studies of exercise testing in the diagnosis of coronary artery disease. He found a wide range of both sensitivity (35 percent to 88 percent) and specificity (41 percent to 100 percent). The results of this review of 33

OCR for page 8
RATIONALE: 19 studies of patients undergoing both a stress ECG and coronary ang~ogram 'isuggest that a principle source of variation may be methodological defects in research design" (Philbr~ck et al. 1980~. These defects included bias in patient selection, referral for the coronary angiogram, and test and gold-standard interpretation, as well as inadequate reporting. There are many obstacles to good studies of diagnostic technologies. The examples above touch on only a few. The problems of conducting clinical teals of therapeutic technologies have been weD documented, and considerable research has been done with We aim of improving thera- peutic teals (for example, see Meinert 1986~. Until recently, however, there has been less interest in teals of diagnostic technology. Although me science of assessment of diagnostic technology has made considerable progress over the last decade, me art of conducting this type of study remains underdeveloped. This monograph focuses, therefore, on the prob- lems that arise in the attempt to coldect high-quality primary data for diagnostic technology assessment. REFERENCES Abrams, H.L. The "overutilization" of x-rays. New England Journal of Medicine 300:1213-1216, 1979. Abrams, Ho., and McNeil, B.~. Medical implications of computed to- mography ("CAT" scanning). New England Journal of Medicine 29X:255-26 1, 1978a. Abrams, H.~., and McNeil, B.~. Medical implications of computed tomography ("CAT" scanning). New England louInal of Medicine 298:310-318, 1978b. Abrams, H.~., SiegeLnan, S.S., Adams, D.F., et al. Computed tomogra- phy versus ultrasound of the adrenal gland: A prospective study. Radiology 143:121-128, 1982. Alderson, P.O., Adams, D.F., McNeil, B.~., et al. Computed tomography, ultrasound, and scintigraphy of the liver in patients with colon or breast carcinoma: A prospective comparison. Radiology 149:225- 230, 1983. Annas, G.~. Your money or your life: "Dumping" uninsured patients from hospital emergency wards. American Journal of Public Health 76:74-77, 1986. Arnstein, S.R. Technology assessment: Opportunities and obstacles. IEEE Transactions on Systems, Man and Cybernetics Health SM-7:571-582, 1977.

OCR for page 8
20 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY Cooper, L.S., Chalmers, T.C., McCally, M., et al. The poor quality of early evaluations of magnetic resonance imaging. Journal of Me American Medical Association 259:3277-3280, 1988. Doubilet, P., and Abrams, H.~. The cost of underutilization: Percutane- ous transluminal angioplasW for peripheral vascular disease. New England Joumal of Medicine 310:95-102, 1984. Dronfield, M.W., I-angrn an, M.~.S., Atkinson, M., et al. Outcome of endoscopy and barium radiography for acute upper gastrointestinal bleeding: Controlled tnal of 1,037 patients. British Medical Journal 284:545-548, 1982. Feeny, D. New health technologies: Their effect on health and the cost of health care. In Feeny, D., Guyatt, G., and TugweE, P., eds., Health Care Technology: Electiveness, Efficacy and Public Policy. Mon- treal, The Institute for Research on Public Policy, 1986. Fineberg, H.V., Bauman, R., and Sosman, M. Computerized cranial tomography: Effect on diagnostic and therapeutic plans. Joumal of the American Medical Association 23X:224-230, 1977. Fletcher, R.H. Carcinoembr~ronic antigen. Annals of Internal Medicine 104:66-73, 1986. Foote, S.B. From crutches to CT scans: Business-govemment relations and medical practice innovation. In Post, I.E., ea., Research in Corporate Social Policy and Performance. Greenwich, Conn., lAI Press, 1986. Foote, S.B. Assessing medical technology: Past, present and future. The Milibank Quarterly 65:59-80, 1987. Guyatt, G., Drummond, M., Feeny, D., et al. Guidelines for me clinical and economic evaluation of health care technologies. Social Science and Medicine 22:393-408, 1986. Harris, J.M. The hazards of bedside B ayes. Journal of the American Medical Association 246:2602-2605, 1981. Hessel, S.~., Siegetman, S.S., McNeil, B.~., et al. A prospective evalu- ation of computed tomography and ultrasound of the pancreas. Radi- ology 143:129-133, 1982. Hickman, Am., and Schwartz, I.S. The adoption and diffusion of Or and MR! in the United States: A comparative analysis. Medical Care 23:1283-1294, 1985. Inouye, S.K., and Sox, H.C., Jr. Standard and computed tomography in the evaluation of neoplasms of the chest. Annals of Intemal Medi- cine 105:906-924, 1986. Kent, Do., and Larson, E.B. Diagnostic technology assessment: Prob- lems and prospects. Annals of Internal Medicine 108:759-76 1, 1988. Cushy, D.A., Shy, K.K., van Belle, G., et al. A randomized trial of electronic fetal monitoring in preteen labor. Obstetrics and Gynecol- ogy 69:687-695, 1987.

OCR for page 8
RATIONALE 21 Marton, K.I., Sox, H.C., Jr., Alexander, J., and Duisenberg, C.E. Atti- tudes of patients toward diagnostic tests: The case of the upper gastrointestinal series roentgenogram. Medical Decision Making 2:439-448, 1982. McNeil, B.J. Pitfalls in and requirements for evaluations of. diagnostic technologies. In Wagner, J., ea., Proceedings of a Conference on Medical Technologies. DREW Pub. No (PHS) 79-3254, pp. 33-39. Washington, D.C., U.S. Government Printing Office, 1979. McNeil, B.J., Sanders, R., Alderson, P.O., et al. A prospective study of computed tomography, ultrasound, and gallium imaging in patients with fever. Radiology 139:647-653, 1981. Meinert, C.~. Clinical Trials: Design, Conduct and Analysis. New York, - Oxford University Press, 1986. Newhouse, J.P., Manning, W.G., Morns, C.N., et al. Some interim results from a controlled trial of cost sharing in health insurance. New England Joumal of Medicine 305: 1501- 1507, 1981. Office of Technology Assessment, U.S. Congress. The Implications of Cost-Effectveness Analysis of Medical Technology. Stock No. 05 ~ - 003-00765-7. Washington, D.C., U.S. Government Printing Office, 198Oa. Office of Technology Assessment, U.S. Congress. The Implications of Cost-Effectiveness Analysis of Medical Technology. Background paper #1: Methodological issues and literature review. Washington, D.C., U.S. Government Printing Office, 1980b. Office of Technology Assessment, U.S. Congress. Medical Technology and the Costs of the Medicare Program. OTA-H-227. Washington, D.C., U.S. Government Printing Office, 19g4. Peterson, W.L., Barnett, B.S., Smith, Ho., et al. Routine early endoscopy in upper gastrointestinal tract bleeding. A randomized trial. New England Journal of Medicine 304:925-929, 1981. Philbrick, I.T., Horwitz, R.~., and Feinstein, A.R. Methodologic prob- lems of exercise testing for coronary artery disease: Groups, analysis and bias. American Joumal of Cardiology 46:807-~12, 1980. Rozanski, A., Diamond, G.A., Berman, D., et al. lathe declining specific- ity of exercise radionuclide ventriculography. New England Joumal of Medicine 309:518-522, 1983. Schroeder, S.A. Medical technology and academic medicine: The doctor- producers' dilemma. Journal of Medical Education 56:634-639,1981. Schwartz, I.S. Evaluating diagnostic tests: What is done What needs to be done. Journal of General Intemal Medicine 1:266-267, 1986. Scitovsky, A.A. Changes in the use of ancillary services for"common" illness. In Altm an, S.H., and Blendon, R., eds., Medical Technol- ogy: The Cuipnt Behind Health Care Costs? Proceedings of the 1977 Sun Valley Forum on National Health, pp. 39-56. DHEW Pub. No.

OCR for page 8
22 ASSESSMENT OF DIAGNOSTIC TECHNOLOGY (PHS) 79-3216. Washington, D.C., U.S. Government Printing Of- fice, 1979. Sheps, S.B., and Schechter, M.T. The assessment of diagnostic tests: A survey of current medical research. Joumal of the American Medical Association 252:2418-2422, 1984. Shy, K.K., Larson, E.B., and L~uthy, D.A. Evalua~dng a new technology: The effectiveness of electronic fetal heart rate monitoring. Annual Review of Public Health 8:165-190, 1987. Sox, H.C., Ir., Margulies, I., and Sox, C.H. Psycholog~caBy mediated effects of diagnostic tests. Annals of Intemal Medicine 95:680-685, 1981. Thacker, S.B. The efficacy of intrapar~n electronic fetal monitoring. American loumal of Obstetrics and Gynecology 156:24-30, 1987. Waldman, S. The effect of changing technology in hospital costs. U.S. Department of Health, Education and Welfare, Social Security Ad- m~stration, Office of Research and Statistics. Research and Statis- tics Note No. 4-1972 (February 28~. Willems, I.S., Banta, H.D., L`ukas, T.A., and Taylor, C.A. The computed tomography scanner. In Altman, S.H., and Blendon, R., eds., Medi- cal Technology: The Culprit Behind Health Care Costs? Proceedings of the 1977 Sun Valley Forum on National Health, p. 131. DHEW Pub. No. (PHS) 79-3216. U.S. Government Printing Office, Wash- ington, D.C., 1977.