NCDS: Revisited Three Decades Later

By Peter Laird, MD

The debate over length of chronic dialysis sessions began nearly as soon as Dr. Scribner solved the access issues with the Scribner shunt in 1960.  Fifty years later, it is still a topic of debate. Dr. Chertow, et al, recently published an observational study based on the USRDS data base looking at mortality and treatment length per session:

Shorter dialysis times are associated with higher mortality among incident hemodialysis patients

On primary marginal structural analysis, session lengths <4 h were associated with a 42% increase in mortality. Sensitivity analyses showed a dose–response relationship between session duration and mortality, and a consistency of findings across prespecified subgroups. Our study suggests that shorter hemodialysis sessions are associated with higher mortality when marginal structural analysis was used to adjust for time-dependent confounding. Further studies are needed to confirm these findings and determine causality.

Glenn Chertow will report on these findings further at the Annual Dialysis Conference in Phoenix as the key note speaker.  Certainly, this is an important study in its conclusions, yet why has it taken 50 years to resolve this rather simple concept in dialysis outcomes? Sadly, The National Cooperative Dialysis  Study (NCDS) from 1981 should have already lead to accepting longer treatment times and to avoid shorter dialysis sessions than 4 hours.  

The NCDS was the first dialysis RCT with four treatment arms: Group I) high BUN clearance and time of 4.5-5 hours, Group II) low BUN clearance and time 4.5 - 5 hours, Group III) high BUN clearance but short time of 3-3.5 hours, Group IV) low BUN clearance and short time of 3-3.5 hours.  The study was stopped prior to the targeted completion date because of significant adverse outcomes in Group II and Group IV which were the lowest BUN clearance groups.

The best outcome in the NCDS was high BUN clearance and long treatment sessions of 4.5-5 hour length seen in Group I. Despite this, the dialysis industry gravitated to shorter and shorter dialysis sessions justifying their use by "adequate" Kt/V clearances derived from a post hoc analysis of the NCDS.  The authors of the NCDS did discuss briefly the element of time as a factor in improving dialysis outcomes, yet continued the main focus of the article on urea kinetics, as did the dialysis industry in general for the next three decades.

Treatment duration appeared to contribute somewhat to morbidity, since Group II had more hospitalization then Group I, and Group IV had more than Group II. Fluid removal is limited by the duration of dialysis treatment. Although there are marginal effects suggesting that short-duration therapy may be inferior to longer-duration therapy, these effects are minor when compared with the effects of TAC_urea. Interpreting the contribution of dialysis TIME to morbidity, however, is made more difficult by the observation that TAC_urea was slightly higher in the two short-TIME groups than it was in the two longer-TIME groups. Furthermore, we cannot make inferences about the effect of treatment times that exceed the limits of those employed in this study (six hours or two hours).

I am taken aback by the NCDS authors discussion on duration of dialysis sessions especially in light of the nearly 30 years of adverse dialysis outcomes in America by adhering to the Kt/V as the single best determinant of "adequate" dialysis.  Clearly, the NCDS showed a profound impact of duration of dialysis sessions on outcomes in this relatively small and short duration RCT when viewed by the hospitalization rates analysis.

Time was clearly an independent factor of improved outcomes. Yet the authors chose to focus mainly on the biochemical clearance which no doubt is also an independent variable in patient outcomes, but not alone. Both improved urea clearance and time are important factors and neither should be ignored, yet history records that those that promoted longer and more frequent sessions continued to be maligned for decades despite many studies showing its benefit even before the NCDS was published.

I take issue with the authors of the NCDS who, in my opinion, belittled and essentially ignored the effect that time and total duration of each session had on the outcomes. At the same time, they elevated biochemical measures, specifically urea clearance as the most important factor alone. This allowed the dialysis industry to justify what had been termed ultra short dialysis to become the new standard of care as long as the Kt/V met minimum goals.  In addition, the dialysis industry did not conduct any further RCTs looking at the independent factor of time or frequency of dialysis until the Toronto group published a small RCT on nocturnal dialysis in 2007 and 2010 when the Frequent Hemodialysis Network Trial Group study was published.  

While Dr. Chertow, et al present a very welcome analysis of dialysis outcomes based on duration of dialysis sessions more than 4 hours, in my opinion, it is nearly 30 years too late for the hundreds of thousands, perhaps millions of patients who suffered in the American dialysis industry while their nephrologists calmly assured them that their dialysis was adequate based on Kt/V. We cannot go back in time, but assuredly we can stop the future debate on dialysis session length. What more do we need to know than the established fact that more frequent and longer duration dialysis saves lives.  Isn't that what Dr. Scribner told us back in the 1960's?

Comments

[Bill Peckham]

Thanks for sending me the full article Peter. Looking back at the provision of dialysis's history I think this time period - the late 70s to early '80s - was when the industry took a wrong turn.

This embrace of urea only makes sense from a business perspective - the only people to benefit from urea driven dialysis were the owners of for profit dialysis units.

National Medical Care - the original for profit dialysis provider based out of the Boston school of nephrology - had people to game out a successful business strategy. Then, as now, it was professionals against hobbyists; the professionals easily won.

[Peter Laird, MD]

Bill, I like your term, urea driven dialysis, it says it all right there. Instead of being patient centered, focusing on restoring health and well being, the entire dialysis industry focuses on urea, debatably a non-toxic solute that is easy to measure, but little consequence other than a marker of renal failure.

Some may debate that point, but the issue is how the NCDS and the follow up article introducing Kt/V by Gotch and Sargent in 1985 did just what you state, change the industry into a urea driven dialysis system. I am not too hopeful at this point that the FHN will change the status as well. We have all of the evidence in our corner, but are still losing the battle in the centers and academic halls.

[Bill Peckham]

The alternative to urea driven dialysis is to ask people how they feel and to offer enough dialysis so that they don't feel sick. FHN showed, again, that urea is an inappropriate measure of dialysis efficacy.

[Mwesten333]

Without frequent home hemodialysis, I would be dead. I have lived 8 years since ESRD diagnosis. The 3xaweek system is designed to keep you from dying, not keep you healthy nor alive for an extended period. The vascular system takes a beating when fluid isn't removed, particularly during that 3-day off period in the clinics. I dialyze 5-6 times a week, 3.5 hours per treatment. There is no shortcut to living with ESRD.

[Peter Laird, MD]

What a great statement, no shortcut to living with ESRD. That is likely something I will repeat again if that is OK with you.

[MooseMom]

"No shortcut living with esrd." That pretty much sums it up if you want to thrive and not merely survive.

[Tejas Desai, MD]

I am currently working on a presentation highlighting the NCDS, HEMO, and recently released FHN data. Would you mind if I link our web resource to this posting? Please send me an email either way letting me know.

Thanks.