The Authors Reply

We thank Dr. Rosansky for his comments.¹ First, we are in full agreement that dialysis should not be started until the patient requires renal replacement therapy. Patients in our study² were started on conventional hemodialysis (HD) 3 times per week, not on incremental HD, by their nephrologists who believed that initiation of dialysis was clinically needed, although we could not determine the exact reason for each dialysis initiation due to the retrospective nature of our study. Moreover, our study patients (based on 386 of the 410 study patients in whom immediate predialysis start serum creatinine values were available) had a mean modification of diet in renal disease estimated glomerular filtration rate (eGFR) of 9.7 ml/min per 1.73 m² (SD of 4.5), and 38% of the study cohort initiated dialysis with eGFR ≥10 ml/min per 1.73 m². These statistics are similar to those reported in the most recent United States Renal Data System report,³ in which the mean eGFR at initiation of dialysis in 2014 was 10.2 ml/min per 1.73 m² and 39% of incident end-stage renal disease cases started with eGFR ≥10 ml/min per 1.73 m². Therefore, our study group did not appear to have been started on dialysis any earlier than what was observed nationally.

Our study aim was to assess the proportion of a clinically stable incident HD cohort who could have theoretically started dialysis on a twice weekly regimen. For each of the subjects, we determined twice weekly HD clearance needed to “complement” the residual renal function, not to suggest that twice weekly HD be initiated to “supplement” the renal function of individuals who might not have required dialysis initiation. In addition, the basis for clearance calculations did not exceed the standard weekly urea clearance target of 2.3 that is generally recommended in clinical guidelines for HD.⁴ Notably, our theoretical ideal twice weekly dialysis group had a measured standard weekly urea clearance of 1.02 volumes, which is approximately one-half of desired weekly clearance, and therefore, this group needed HD.

Second, although we agree with Dr. Rosansky that failure to control volume with diuretics might be a reason to initiate dialysis, we disagree that all patients in this scenario should start conventional thrice weekly treatments. Our calculations suggested that the volume removed on twice weekly HD, even with a tight upper limit of dialysis ultrafiltration rate <13 ml/kg per hour, may allow adequate weekly fluid control in many patients with residual renal function. Optimization of diuretic use in such patients, even when dialysis is initiated, may further aid in decreasing the ultrafiltration needs.

Finally, we agree that preserving residual kidney function is important and perhaps not emphasized enough in the care of HD patients. To that end, twice weekly HD in patients where it is feasible, may allow for greater native renal function longevity. We also hope that our study provides insight to incremental HD, so that it can be a real option for the appropriate patient, rather than just a default prescription for patients with limited resources.