Hemoglobin levels in dialysis patients

Hello Dr. Agar,

As you may know, payments for epo have been “bundled” by Medicare here in the US, and on top of this, “experts” and Amgen (who manufactures epo) have stated that lower dosages are better for dialysis patients. I have read that higher hemoglobin levels (13 and above) are dangerous in D patients as they lead to a higher risk of cardiovascular accidents. Can you tell me what it is about dialysis patients/the dialysis process that makes it safer to have lower hemoglobin levels? Thank you!

What a ‘devastating’ question … and one on which I swim against the tide!

I preface by apologising for the long response … but, its a seminal issue and one about which I beleive dialysis patients are being hoodwinked out of a treatment for the wrong reasons … the containment of cost and not the delivery of benefit.

If I am true to my colors, this answer will place me ‘out there’ and ‘at odds’ with many of my fellow nephrologists. I know that the Internet can provide many ‘suspect’ opinions … and I hope you don’t take this as just another one of those … but, what-the-hell … I am generally known for being prepared to say what I believe … and I think there is more evidence than is admitted that suggests those beliefs are fairly founded …

So, here goes …

I remain a sworn skeptic of the studies and data that suggest that low Hb’s are ‘good for dialysis patients’ … note, there, my use of the all-encompassing term ‘all dialysis patients’. Sorry … but I can’t bring myself to accept this sudden change-of-heart.

The skeptic within me whispers that the low Hb ‘push’ was first and foremost the result of a cost-control initiative, primarily in the US, where the high costs of ESA’s were ‘breaking the piggy bank’.

I may be pilloried for saying this but better dialysis helps drive better Hb’s and better Hb’s = less ESA use. There is ample NHHD data which shows that ESA requirements fall by more than half in NHHD.

Much of the dialysis delivered in the US isn’t ideal (or even ‘good’) dialysis. It struggles to even make ‘adequate’ dialysis … and is all-too-often judged on Kt/V criteria alone … and you already know what I think of Kt/V (see my section on ‘good dialysis’. US dialysis is all-too-often short hour, infrequent, and brutal - especially in its fluid shifts and circulatory insults.

This is not a milieu ideally suited to optimum delivery of or response to an ESA. It is hardly surprising then, that ESA use in the US was ‘out of hand’, that costs were burgeoning and that ‘something had to be done to curb the use’. That ignores the odd administrative system that rewarded and incentivized the high use of ESAs with added $’s … but I won’t go there.

Two studies drove the cry for a lower Hb. Some would argue that these were highly selective studies which whittled down two very large patient populations into two quite small study groups of 1432 patients (the ‘CHOIR’ study) and 603 patients (the ‘CREATE’ study) each. Even their conclusions were ‘inconclusive’. While one trial suggested a 34% increase in cardiovascular ‘end-points’ with a higher Hb, the other trial disagreed and did not show this outcome.

Importantly, the ‘CHOIR’ trail – the one that supposedly showed a ‘34%’ increase in cardiovascular ‘end-points’, was done on pre-dialysis CKD patients, not dialysis patients! Furthermore, the patients who were “randomized” to the high Hb group had significantly higher cardiac disease prior to the study! And … no individual cause of death was significantly different - they had to patch together a ‘composite cardiovascular’ outcome to find significance (a ‘solution’ to find a result which was, I think, a solution ‘after the fact’).

My God, you say … if I am going to have a 34% higher rate of risk or cardiovascular trouble with a higher Hb, spare me the ESA!

BUT … and it’s a very big ‘but’ … you need to understand what that ‘34%’ means. It’s not that 34% of patients with a high Hb died! No … far from it …

Of 1432 patients in the trial that showed a ‘high risk’ from a high Hb, there were only… and I use the word ‘only’ advisedly … 52 deaths in the higher Hb group and only 36 deaths in the lower Hb group. There were 64 hospitalizations in the higher and 47 hospitalizations in the lower Hb group. But remember, these were not dialysis patients – they were pre-dialysis – and the high Hb group started out with a significantly higher rate of congestive cardiac failure (read fluid overload) to begin with … before the trial even got underway.

Back to the numbers: 52 + 64 = 116 (higher) and 36 + 47 = 83 (lower). Mind you, his is 116 and 83 out of a randomized total of 1432.

That means that 1432 – (116 + 83) = 1233 had no deaths or hospitalizations … but nothing is said about how ‘well’ those remaining 1233 patients were.

Was the ‘difference’ between the two groups significant?

Was this ‘difference’ … 116-83 = 33 (of 1432 patients) … really due to the higher Hb or was it because the higher Hb group had more cardiac disease to begin with – ie: before they were ‘exposed’ to a high Hb.

Or … finally … was it chance?

And … if and when the patients reached dialysis, what was the dialysis ‘quality’ of all patients like? Was it the same?

And … why was the initial number of patients ‘available’ for the trial far more than 10000 patients – in each trial – and yet the trial participants of 1432 and 603 were a whittled down number less than 10% of the ‘available’ (or at risk) patient population?

Then, this small-difference conclusion, reached from this small cohort of patients which were highly selected out from the total ‘at risk’ population, was converted into a ‘recommendation’ to back-apply to all 400,000 dialysis patients (a different population group anyway) as if they were all the same as the ‘trial subjects’.

It is also quite possible that the poor results seen in some studies of where high doses of ESA have been required are confounded by the fact that sicker patients - wth higher levels of inflammation - need a higher dose of ESAs and it their inflammatory state (whatever the cause) that is the real cause of harm and not the ESA at all.

There are many more ‘holes’ in these trials … too many to argue about here (either pro or con) … but, the cost containers had their studies! Aha! … a reason to reduce ESA use (and cut costs). That is not to say that all ESA use was appropriate or necessary … but that wasn’t the target. The target was ‘all dialysis patients’ and ‘all Hb levels’. But, as they say … the rest is history.

I am an old guy. I still kind of like registry data. Why? …

Registries record all comers, they are warts and all, cover the the good and the bad, the sick and the well, the old and the young. They have no selection, no introduced biases from the over-selection of trial-bunnies or the use of ‘exclusion criteria’ which just may ‘exclude’ things that might just be important. Registry data is just good, whole-of-practice data. Yes, registries have their drawbacks too … but no more, or less, than do the revered RCTs … and I think we have prostituted decision-making to the revered RCT far too much. OK, that’s a personal view, but I hold to it.

Time and again, the Brits (and others) have published their registry data. This seems to have fallen on deaf US ears – or, more likely, wasn’t read … after all, it wasn’t a US study, was it.

Ian MacDougall and colleagues from the UK registry reported in NDT, Volume 25, (3) March 2010 914-919, the following abstract … which you may get the gist from, but I have copied it verbatim …


Background. Much controversy has been generated in recent times over the optimal target haemoglobin in chronic kidney disease patients receiving erythropoiesis-stimulating agent therapy. This has arisen from the paradoxical conclusions obtained from large retrospective or epidemiological studies versus interventional randomized controlled trials.

Methods. Data from haemodialysis patients in the UK Renal Registry from 1999 to 2005 were analysed year by year for the relative risk of death at different haemoglobin concentrations, compared with a reference of 1011 g/dl. The population size varied from 2291 in 1999 to 8209 in 2005. The data were analysed by chi-square tests, and a multivariate analysis was performed.

Results. Across the years 1999 to 2005, there was a consistent relationship between the haemoglobin achieved and the risk of death (P < 0.0001). In 2005, the relative risk (RR) of death = 1.32 for a haemoglobin (Hb) of 910 g/dl; RR = 0.44 for Hb > 13 g/dl. The relationship between Hb and the RR of death is nevertheless remarkably consistent across the 7 years of study, with an S-shaped correlation (polynomial) between an Hb range of <9 g/dl and an Hb range of >13 g/dl (P < 0.0001). Multivariate analysis also showed age, time on dialysis and diabetes to be strongly predictive of death across all 7 years analysed (P < 0.0001 in all cases).

Conclusion. There is a significant relationship between achieved haemoglobin and mortality across the 7 years analysed, with no increase in risk seen with higher Hb levels

It is really worthwhile looking at this paper … and it is representative of registry data from many other sources outside the US … which shows, consistently, a better survival with a better Hb. You can see the whole MacDougall paper at:

For mine … I remain an Hb 12 g/L target nephrologist.

For mine … there is zero evidence that a naturally occurring higher Hb level is harmful for those on dialysis with a naturally ‘good’ Hb … like many patients with polycystic kidney disease. We don’t bleed PKD patients to get their Hb down to 11.

For mine … an Hb of 12.5 is ‘getting up there’ and I will accept that an Hb of 13 is ‘getting too high’ … and I would back of the ESA until back under 12 then recommence at a lower dose … but, equally, I am not convinced (at all) that ‘less is better’. What is better is better dialysis. More dialysis lessens ESA use.

I refuse to believe that having a reasonable level of oxygen in your blood (red cells carry the oxygen and less red cells = less oxygen-carrying capacity) can harm you!

There are more important things than death, like your daily QOL. If there truly is a choice between length of life and quality of life, only the patient and his/her MD should be making it! This should not be the subject of regulation.

My reading is that US Hb levels are plummeting. They now hover close to 10, I think, and I wouldn’t be surprised if a ‘9’ is on the way. Many more patients are at risk from too low levels than too high ones.

In addition, your problem in the US, is not the Hb level (or the ESA usage) … it never has been … the real problem is the level of administered dialysis.

Deliver better dialysis and ESA use will fall – probably dramatically – but the use of ESA will fall as a consequence of better dialysis and not as a result of ham-fisted regulation and ESA denial.

But then, few US nephrologists will believe that.

I don’t care. It’s what I think.

Take it or leave it.

I’ll take it. Thank you!

Oh, btw, there does seem to be a pattern here in that the answer to most of our queries does appear to be “more dialysis.” lol!

I guess I must sound like a broken record to many … but, in a word, yes.

More dialysis isnt the answer to the Universe … 42 is … according to the Hitchhikers Guide to the Galaxy!

see … http://en.wikipedia.org/wiki/Phrases_from_The_Hitchhiker’s_Guide_to_the_Galaxy

But it IS the answer to many of the issues raised at this site like: dialysis symptoms, volume control, BP control, biochemical control, well-being, quality-of-life, survival … just to name a few.

And I do hope my answer was intelligible. I know that it is against ‘current views’ in the US … but I can’t help if if I think those views are ill-founded and based more in funding ideology than in common sense. Your god - if you happen to believe in one - gave us Hb for a reason. I see no over-riding reason to buck nature.

No, you don’t sound like a broken record at all, and I apologize if my comment sounded more snarky than true. We can all faff about debating this point or that, but it’s just common sense that the closer your renal replacement therapy can replicate the function of a normal kidney that is working 24/7/365, the fewer dialysis-related problems you will have. Far from sounding like a “broken record”, you sound like a neph who is trying to shake some common sense into those many, many people who are more interested in making a buck than they are in actually rehabilitating a dialysis patient.

And yes, your answer was intelligible, and your skepticism is shared by many renal patients.