The ultimate goal of imaging research, indeed, all medical research, is the improvement of human health. However, the technologic nature of our discipline often results in research that assesses a technology or the application of a technology in terms of the information it provides rather than its relationship to care of patients. Dr. Rosenquist’s paper [i ] describes an example of how assessing a technology solely on the basis of the information it provides can be misleading with regard to the effects of that technology on patients’ health. In a broad sense, nearly all of our literature deals with imaging technology assessment. An unfortunate proportion of imaging publications are retrospective, without suitable controls or verification of diagnoses, and they employ dubious or irreproducible measures of efficacy. In evaluating the other publications, we can consider a five-stage hierarchy of imaging technology assessment: imaging efficacy, diagnostic efficacy, therapeutic efficacy, evaluation of the patients’ outcomes, and cost-effectiveness assessment. Imaging efficacy is the level of assessment most commonly encountered in prospective imaging research. Evaluation of imaging efficacy is typified by the work of Abu-Yousef et al. [2], who sought to determine the potential of sonography in identifying patients with appendicitis. This level of assessment is an important and necessary first step in assessing the value of an imaging technology. However, as Dr. Rosenquist has shown, little relationship may exist between imaging efficacy and providing benefits to patients. Indeed, only a few imaging studies have sought to develop a link between advances in imaging technology and benefits to patients. This lack of data has provided a rationale for often misguided technology regulation and reimbursement conmisguided technology regulation and reimbursement constraints. Moreover, the absence of reliable information puts radiologists at a disadvantage in selecting and using technology, especially in an environment that increasingly puts us and our associates at financial risk for applying technologies that do not result in cost-effective care. Levels two through five of the assessment hierarchy potentially could contribute to an information base that would correct these deficiencies. Level two is diagnostic efficacy. Studies dealing with diagnostic efficacy seek to determine the importance of the use of a technology in establishing or ruling out a diagnosis. An example of a measure of diagnostic efficacy that has been used successfully in assessing applications of imaging technology is the log likelihood ratio [3]:
Search from 250 M+ research papersSearch from 250 M+ research papers
Jul 17, 1996
Open Access