In Ontario, between 1993 and 2003, the annual number of MRI scans performed increased by more than 600 percent (Iron et al. 2003), and the number of CT scans increased threefold (Tu et al. 2005). Despite these massive increases, the Fraser Institute reported a median wait of five weeks for CT and thirteen weeks for MRI scanning in 2004 (Esmail and Walker 2004), and Canadians are increasingly concerned about the length of time they wait for diagnostic imaging. Because of this, politicians have made decreasing wait times for diagnostic imaging one of their top priorities (Health Canada 2004). This raises several interesting questions. Have the indications for CT and MRI really expanded that rapidly, or was there just a huge pent-up demand because Canada had fallen so far behind in acquiring modern imaging machines? Are physicians relying more on diagnostic imaging technologies and less on clinical skills? Are an increasing number of patients undergoing scans when there is a small likelihood that the results will change their management or improve their outcomes? Supporters of the view that Canada needs to expand its diagnostic imaging capacity point to the fact that we rank well behind many developed countries in terms of the number of diagnostic imaging machines per population (Canadian Institute for Health Information 2003: 33), and that improvements in imaging quality have expanded the indication for imaging. Supporters of the view that there is an increased and inappropriate reliance on technology over clinical skill point to the findings of a recent American study showing that the regions that spent the most on healthcare did not have better outcomes than the regions that spent less--indeed, the trend was toward poorer outcomes in the highest-spending regions (Fisher et al. 2003a, 2003b). One of the greatest differences between the highest- and lowest-spending regions was their expenditure on a variety of diagnostic tests, suggesting that more testing did not lead to better outcomes on a population basis. It may in fact have led to iatrogenic illnesses because of the workup of false positive results, and diverted attention away from simple interventions that have been shown to be effective (Fisher et al. 2003a). The truth is likely a combination of many factors. Some patients with clear indications for diagnostic imaging undoubtedly wait too long for their tests in Canada. At the same time, a number of patients undergo tests whose results have a very small likelihood of changing their management, which itself contributes to the access problem. Unfortunately there are no evidence-based benchmarks for the appropriate rate of diagnostic testing that can be used to determine the optimal supply of diagnostic machines and radiological personnel. In this article we discuss the reasons it has been so difficult to determine the optimal imaging capacity needed for a population, describe some factors that are "inappropriately" increasing the rate of imaging and suggest some solutions. Although many of our examples deal with CT and MRI scanning, our remarks apply more broadly to many other diagnostic tests.
