Advances in science and technology are generally the product of multiple investigations. This article discusses statistical methods for combining empirical results from a series of different experiments or clinical investigations. We delineate the steps an assessor might take in combining data from different studies and provide references for topics not discussed in detail. The article reviews some of the most commonly used statistical techniques for combining results in the medical and social sciences.

The expression meta-analysis refers to the practice of using statistical methods to combine the outcomes of a series of different experiments or investigations. The terminology arises because a meta-analysis uses a study as the unit of observation, and, thus, the data point for each study is itself a statistical summary based on an analysis of primary data.

This article describes a collection of meta-analysis techniques based on Bayesian statistics for interpreting, adjusting, and combining evidence to estimate parameters and outcomes important to the assessment of health technologies. The result of an analysis by the Confidence Profile Method is a joint posterior probability distribution for the parameters of interest, from which marginal distributions for any particular parameter can be calculated. The method can be used to analyze problems involving a variety of types of outcomes, a variety of measures of effect, and a variety of experimental designs. This article presents the elements necessary for analysis, including prior distributions, likelihood functions, and specific models for experimental designs that include adjustment for biases.

The complexities and nuances of evaluating the costs associated with providing medical technologies are often underestimated by analysts engaged in economic evaluations. This article describes the theoretical underpinnings of cost estimation, emphasizing the importance of accounting for opportunity costs and marginal costs. The various types of costs that should be considered in an analysis are described; a listing of specific cost elements may provide a helpful guide to analysis. The process of identifying and estimating costs is detailed, and practical recommendations for handling the challenges of cost estimation are provided. The roles of sensitivity analysis and discounting are characterized, as are determinants of the types of costs to include in an analysis. Finally, common problems facing the analyst are enumerated with suggestions for managing these problems.

This article reviews the methods of economic evaluation of health technology. The different forms of evaluation are outlined and the importance of assessing costs and benefits at the margin stressed. The approaches used by economists to analyze priorities for health technologies are discussed and a number of key managerial and policy issues identified. It is concluded that, while further methodological improvements can be made, a clear methodology for economic evaluation of health technologies has been established. However, attention needs to be paid to the timeliness and relevance of economic evaluation in order to ensure its maximum impact

Cost-effectiveness analysis (CEA) is a method of economic evaluation that can be used to assess the efficiency with which health care technologies use limited resources to produce health outputs. However, inconsistencies in the way that such ratios are constructed often lead to misleading conclusions when CEAs are compared. Some of these inconsistencies, such as failure to discount or to calculate incremental ratios correctly, reflect analytical errors that, if corrected, would resolve the inconsistencies. Others reflect fundamental differences in the viewpoint of the analysis. The perspectives of different decision-making entities can properly lead to different items in the numerator and denominator of the cost-effectiveness (C/E) ratio. Producers and consumers of CEA need to be more conscious of the perspectives of analysis, so that C/E comparisons from a given perspective are based upon a common understanding of the elements that are properly included.

Cost-effectiveness analysis compared four treatments of end-stage renal disease in Brazil: continuous ambulatory peritoneal dialysis (CAPD), in-center hemodialysis (HD), cadaver donor transplantation (CD-Tx), and living related donor transplantation (LR-Tx). After 2 years, the costs per year of survival were CAPD, $12,134; HO, $10,065; CO-Tx, $6,978; and LR-Tx, $3,022. The HD cost was lower than CAPO partially because of the reuse of hemodialyzers in Brazil. Although less cost-effective, both dialysis treatments yielded more years of survival after 2 years. This analysis reveals a trade-off between cost per year of survival and years of survival.

This article considers technology assessment (TA) to be a comprehensive form of policy research. Technology assessment must then have a relation to policy-making; in the area of health care, TA must relate to such political decisions as resource allocation. Since policies are determined politically, i.e., by factors such as power and influence, technology assessment is, or should be, part of a political process. Technology assessment seems to be developing predominantly as a technical and professional activity, carried out in centers with no relation to the policy-making process. While the impact of technology on health, as well as such broader impacts as those on financial costs, is a key concern, political considerations and political decision-making must always be an important dimension in health care TA.

Direct costs were estimated for the treatment, rehabilitation, and nursing of 125 patients with first stroke in the Lund and Orup health districts (population of 200,191). Patients were followed from the onset of stroke in 1983 until October 31, 1985. The data were used to calculate the present value of the expected lifetime direct costs for an individual contracting his or her first stroke at various ages. For example, at the age of 72, these costs were estimated at SEK 283,000 for a man and SEK 561,000 for a woman. Hospital care was the major cost component (75% for males and 89% for females) at this age of onset. The results could be used to estimate the economic benefits of preventing new strokes, hence, forming part of a cost-effectiveness or cost-utility analysis. Together with prognoses of the expected developments in the incidence of stroke, they could also serve as the basis of forecasts of future costs of the health care and social service sectors.

By introducing discounting in the DEALE method, we reanalyze a recent publication on the cost-effectiveness of early thrombolytic therapy with intracoronary streptokinase after acute myocardial infarction. We show that failure to discount gains in life expectancy results in a serious overestimation of the cost-effectiveness of this technology.

Evince-I is a desktop expert system for the differential diagnosis of dementia, implemented on a personal computer. It is intended to assess the effectiveness of this new technology in modeling a psychiatrist who uses international guidelines for diagnosing dementia. EVINCE-I was tested in diagnosing 19 patients with varying stages of dementia and 10 patients showing other disorders except dementia. EVINCE-I and the human expert were in perfect agreement on the diagnosis of dementia and correlated highly on the diagnosis of dementia of the Alzheimer type and multiple infarct dementia. EVINCE-I thus offers important possibilities as a tool in investigating the data and procedures used by the human expert.

Our present-day health sectors are increasingly characterized by both rapid technological change and growing visibility of its considerable consequences in terms of the quality and costs of health care. Improved health care decision-making in such an environment requires adequate and timely information on the benefits, risks, and costs of health care technology. Whereas traditionally, technology assessment focused predominantly on evaluating efficacy and (short-term) safety in a more or less “controlled” environment, an interesting shift in emphasis is now occurring. There is a growing recognition that improved information is also needed on the health and economic outcomes of technology as used in everyday clinical practice, i.e., on its effectiveness, cost-effectiveness, and long-term safety in normal “uncontrolled” daily medical life. Moreover, because the treatment of particular clinical conditions (such as stable angina, gallstones, or prostatism) increasingly involves choices among alternative diagnostic and therapeutic options, these assessments should ideally provide effectiveness and cost-effectiveness information for all the various technological alternatives involved.

During the past 5 years, hospitals in several countries have experienced a significant increase in the use of platelet concentrates. In the Netherlands, e.g., the number of donor units of platelet concentrates rose from 59,000 in 1981 to almost 97,000 in 1984 at a cost of 5 million guilders (US $2,800,000) annually. This increased utilization has been attributed to such factors as chemo- and radiotherapy, major surgery, severe trauma, and bone marrow transplantation.

However, there was concern that in a significant number of patients, platelet concentrates are administered for no good reason, that the effect of platelet transfusions is often not adequately controlled, that insufficient care is taken to prevent side effects (notably alloimmunization), and that the quality control applied to platelet suspension is not optimal.