Renal Arterial Resistance Index

Publisher Website
Google Scholar
Abstract
Radermacher et al. (July 10 issue)1 assert that the renal arterial resistance index is associated with the prognosis for renal-transplant recipients. This study raises methodologic issues. Survival data are defined in terms of the time from an initiating event (the time origin) to some terminal event.2,3 The date of ultrasonography is a questionable time origin, since it is not a well-defined event; indeed, ultrasonography took place at a random date between 3 and 317 months after transplantation. In our opinion, the time origin should be closely related to the date of transplantation, which would uniformly define time relative to transplantation and provide a clear definition of patient subgroups (since the resistance index is time-dependent). If this approach is not taken, we suggest that survival data not be handled as binary data. Therefore, sensitivity, specificity, and receiver-operating-characteristic curves must not be estimated, except if ad hoc methods are used,4 and survival rates must be derived from Kaplan–Meier curves and expressed as time-dependent percentages. In conclusion, because of these methodologic problems, the results of the study may be confounded.