A thorough and objective knowledge of the natural history of vascular disease (and of vascular interventions) is the essential underpinning of vascular surgery. This knowledge enables us to dispassionately select the patients most likely to benefit from intervention and the interventions most likely to help our patients. Pearce and associates have presented a well conceived, but ultimately flawed, study designed to clarify the natural history of renal atherosclerosis. The data show a low rate of incidence and progression of renal artery disease in the study population.
The authors’ thoughtful discussion points out the three major difficulties in this report: the low rate of follow-up (14.3%), the “healthy cohort” effect, and the challenges inherent in longitudinal observational studies using noninvasive technology, even in the best of hands. There is no question that the group under study in this report was a healthy cohort, and, therefore, the prior probability of significant renal arterial disease in the study group was low. Thus, the likely high false-positive rate seen in the initial renal duplex scans in this report would be expected by Bayes’ theorem. The authors cite other potential methodologic problems, including drift in diagnostic criteria and changes in personnel and equipment. The combined effect of these considerations is an underpowered study with some degree of diagnostic uncertainty.
Despite these shortcomings, these data are among the best available today to define the incidence and rate of progression of renal arterial disease in unselected subsets of the elderly. This article clearly supports the contention that there is no justification for duplex screening of the renal arteries in the elderly in the absence of clinical features suggesting a higher incidence of disease. Unfortunately, these data shed no light on the natural history of the renal artery stenoses that vascular specialists encounter frequently in practice: those occurring in patients with severe hypertension, moderate renal insufficiency, and/or significant symptomatic systemic atherosclerosis. Published reports of highly selected subsets of these patients show much higher incidences and rates of progression of renal artery disease but suffer from analogous selection bias, as well as methodologic and statistical issues.
In short, there are few solid data on which to base our clinical judgments in this area. This article should stimulate vascular surgeons to take the lead and harness our collective expertise to design sufficiently powered, well-controlled, multicenter studies using state-of-the-art duplex technology to address these fundamental gaps in our collective knowledge. Single-center studies in isolation, even those performed at centers with great skill and experience, will not provide the answers we need to properly guide our practice.
FULL TEXT