Do we need randomised trials to evaluate diagnostic procedures?

Excerpt

The VU Medical Centre serves a population of 2.5 million people and so, in 1996, when it was decided that the Centre would acquire a dedicated positron emission tomography (PET) scanner, it was anticipated that diagnosticians and clinicians would soon be confronted with a lack of scan capacity. Decisions would need to be made on which clinical indications and, within those indications, at what point in the diagnostic work-up PET would be most beneficial. Our main concern was to try to ensure equity of access to this new technology. We insisted that routine use of PET should be restricted to those applications where it would have a clinically meaningful impact, reasoning that waiting lists are not compatible with routine patient care. In cancer care, most of the literature on the use of PET focussed on non-small cell lung cancer (NSCLC), which was also the most common cancer in terms of incidence. Moreover, a retrospective study in our region identified a relevant clinical problem in lung cancer diagnosis [1]. …