LAUSR

We were disappointed to see the Canadian Association of Radiologists Journal (CARJ) published the recent article, ‘‘Choosing Wisely Canada and Diagnostic Imaging: What Level of Evidence Supports the Recommendations?’’ [1], without reviewing the wording of the current guidelines. The current wording of the Choosing Wisely Canada (CWC) Family Medicine guideline on screening mammography is actually, ‘‘Don’t routinely do screening mammography for average risk women aged 40e49. Individual assessment of each woman’s preferences and risk should guide the discussion and decision regarding mammography screening in this age group.’’ The term ‘‘low risk’’ was changed to ‘‘average risk’’ within weeks of the original release at the request of the Canadian Association of Radiologists to comply with evidence of risk of breast cancer. It is frustrating to see that the inaccurate wording of the first version has been perpetuated in the CARJ itself several years later. It can only be inferred that the erroneous and misleading wording of the original CWC document has had a lasting impact. Also troubling, the rather impressive-appearing level of evidence (LOE) for the screening mammography section belies many problems with the quality of the evidence. LOE, after all, should include consideration of both study design and quality of evidence. For example, the early studies from the 1960s to the 1980s were superficially well designed, but the largest study designed to study 40-50-year-olds was contaminated by poor randomization and patient crossover from non-screening to screening limbs and was the only study that failed to show a mortality benefit as compared with seven other randomized controlled trials (RCT). This has been well-documented. The other studies were not designed to distinguish between 40e50 years and 50 years and older. Moreover, in radiology, a field highly sensitive to advances in technology, the quality of a study from the 1960s is severely eroded by obsolete imaging, even if the study had been well-designed and executed. Imagine publishing a brand-new paper on driving safety recommendations based on RCT evidence from cars made in the 1960s. Yes, there are some truths which may have stood the test of time, but the