Letters to the Editor
To the Editor:
In their recent article, Whittier et al. compared two access flow (Qa) measurement approaches: flow dilution (FD) (Transonic Systems, Inc., Ithaca, NY) and in-line dialysance (DD) (Fresenius Medical Care North America, Waltham, MA). Contrary to the authors' conclusion, I find that the presented data illustrates that DD measurements are not equivalent to DD in the range <2,000 ml/min and lead to substantially different referral indications.
KDOQI Surveillance of AV Access: Low-Flow Threshold Guideline
The authors report that their DD measurements suggested patient referrals for AV grafts in nine versus six instances for FD and patient referrals for AV fistulae in three instances when tested by DD versus two for FD. In other words, this study yielded a 50% higher referral rate for DD than for FD. In view of the cost of false-positive referrals and needless patient suffering during superfluous interventions, I think the authors' conclusion that these two methods may be used interchangeably is unwarranted. The authors correctly point out that larger studies are needed to accurately test referral differences between the two methods, but until such studies prove otherwise, the starting conclusion from the subject study should be that DD will have an unwarranted higher referral rate under the Low-Flow Threshold KDOQI Guideline than FD.
KDOQI Surveillance of AV Access: Sequential Trend Analysis Guideline
The second K/DOQI Guideline for AV access surveillance states that further evaluation of access dysfunction is suggested if Qa had decreased >25% during the course of 4 months if the Qa reading becomes <1,000 ml/min—a 4-month drop from 1,200 to 900 ml/min. The authors do not discuss whether the Trending Guideline (developed in studies using FD) can be implemented at all using DD, even though literature studies have suggested that of the two, this is the more predictive diagnostic test.
It concerns me that in the authors' Bland Altman chart, the difference between the two methods in the 600–1000 ml/min range of interest is >200 ml/min in at least 10 of the cases. This suggests that sequential DD measurements may show differences of 25% or larger even where FD would indicate a stable access, i.e., false positive referrals. It is difficult to expect that a method with such a poor repeatability can be deployed to perform reliable trending, and the authors most certainly did not present that DD can replace FD for this KDOQI Guideline.
In summary, we find that the authors have not made the case that Transonic access flow by flow dilution and Fresenius access flow by in-line dialysance may be used interchangeably.
Michael Levine, MD
Wisconsin Vascular Access and Nephrology