Evidence and the Role of Diagnostic Imaging

The Grading of Recommendations, Assessment, Development and Evaluationworkinggroup,while recognizing that diagnostic tests can play a number of roles, states that when used for primary diagnosis, randomized control trials of patient outcomes remain the highest standard of evidence for evaluating diagnostic tests [1,2]. If this evidence is not available, welldesigned studies of the accuracy of diagnostic tests can be used as a proxy for patient outcomes [1,2]. However, in practice, high-quality evidence of either of these 2 types may not be available to groups that are trying to develop guidelines on the appropriate use of diagnostic imaging, such as the ACR Appropriateness Criteria [3]. When determining the type of evidence that may be available and is most relevant, it is important to look at the role diagnostic imaging will play. Diagnostic imaging plays 4 important roles in the care of patients. It is widely used to assist in the primary diagnosis of a patient’s illness. However, it is also used to help in the management of a patient by assessing the changing state of disease in response to treatment or by aiding in the diagnosis of complications. It can be used as a screening tool. It is also used as an integral part of many interventional procedures. Interventional radiologic procedures are treatment modalities and should be evaluated like any other treatment intervention [4]. Ideally, randomized clinical trials should be used to compare their efficacy with that of other treatment modalities [5]. However, if randomized controlled trials are not available or cannot practically be carried out, other well-designed studies, such as cohort studies and assessments of benefits and harms, may also be used as evidence [5]. Randomized controlled trials may be considered to be the best evidence for assessing the value of diagnostic imaging for screening [6]. However, because of the rapid advances that occur in diagnostic imaging, the results of randomized controlled trials may become out of date, and it may become impossible to repeat such trials. It then becomes necessary to look at other types of evidence [7,8]. When diagnostic imaging is used as a screening modality, outcome studies are extremely important. These must include measures of accuracy, patient outcomes, and cost-effectiveness measures [9,10]. These studies are difficult to design and carry out [11], and the results, as we know from the ongoing debate about the appropriate age range for mammographic screening [12,13] or whether it is effective at all [14,15], are often controversial. Part of the problem is that the rapid advances in diagnostic imaging technology frequently render obsolete the results of the large-scale trials necessary to assess screening [16]. Nevertheless, the best current evidence should be used to support screening recommendations [7,13]. When diagnostic imaging is used to assist in the management of patients, outcome studies to identify the value of the specific role of diagnostic imaging would be extremely difficult to carry out because diagnostic imaging represents only a part of the whole management strategy, and the actual treatment usually has more impact on a patient’s outcome than does the diagnostic imaging. Accuracy studies may be available, and if they are available and of high quality, they are of course important evidence