Wednesday, July 7, 2010

Follow the patient, not the number

As a nephrology fellow, I find it somewhat challenging to follow a patient with advanced CKD (eGFR, MDRD 8-10). However, I do find this process rewarding as it truly allows for the establishment of a substantial relationship with your patient. I follow the patient every 1-2 months and ask at each clinic visit if they are experiencing any uremic symptoms. Most of the time they say no and I go about performing an exam and review their labs. With the lack of randomized controlled trials to answer the question about what the optimal eGFR is to start renal replacement therapy we truly have to "listen" to the patient. However, as the symptoms of uremia are often insidious, a patient may not recognize how bad they really feel. I often wonder if I am putting a patient at risk by "waiting" for the development of uremic symptoms. Several recent observational studies have suggested that starting renal replacement therapy early may, in fact, be harmful.


Enter the IDEAL trial (
Initiating Dialysis Early and Late). Published in the NEJM on June 27th, 2010. This trial randomized 828 patients (in Australia and New Zealand) with advanced CKD (cockcroft eGFR 10-15ml/min) into 2 groups according to when renal replacement therapy (RRT) is initiated.
  1. Early RRT initiation group- 10-14 ml/min
  2. Late RRT initiation group- 5-7 ml/min
However, the decision on when to initiate RRT was left to the discretion of the treating physician in the late initiation group. The primary outcome was death from any cause and secondary outcome were cardiovascular events, infections, quality of life and dialysis related complications. The patient population was mostly white (70%), male (64%), average age of 60, DM accounted for 1/3 of the ESRD, PD was the planned form of RRT initiation in 195 patients in the Early RRT group and 171 in the Late RRT group. Hemodialysis was the method in 188 in the Early RRT and 215 in the Late RRT groups. Both groups had been followed by a nephrologist for an average of 2.5 years before being enrolled in the study. Average creatinine at time of enrollment was 6 mg/dL.
There was no statistical difference in either the primary or secondary outcome in both groups. Interestingly 322 patients (76%) in the late RRT initiation group were started on some form of RRT secondary to mostly uremic symptoms or fluid overload before reaching an eGFR below 7.
(cockcroft gault/MDRD) eGFR at randomization and initiation of RRT-
  • Early group start eGFR (13/9.8) --> 1.5 months average before initiation, eGFR (12/9)
  • Late group start eGFR (13/9.9) --> 7.8 months average before initiation, eGFR (9.8/7.2)
In conclusion, I agree with editorial that accompanied this article. The majority of patients in the late group were initiated on RRT secondary to symptoms of uremia. Waiting for this to occur did not adversely affect the outcome in the late initiation group. Providing excellent pre-ESRD care to all patients with CKD is paramount. Getting timely access and providing medical therapy for complications of hypertension, fluid overload, electrolyte/acid-base derangements while listening closely to the patients symptoms of uremia, as we already do, seems to trump the lab value (in this case the estimated glomerular filtration rate). The decision to initiate renal replacement therapy needs to be individualized, not simply generalized by a lab number. Dr. Simon Prince has written a nice blog and video-blog piece on this article as well.

9 comments:

Anonymous said...

I think an important question is quality of life that is not addressed in this study. does starting a patient early allow an easier transition to dialysis, save on cost of hospitalization for uremic sx/vol overload. Mortality is a very important but lets not forget quality of life.

Matt Sparks said...

actually a quality of life score was measured during this trial. "No significant difference was observed between the two groups in quality of life, as measured by the Assessment of Quality of Life instrument during the follow-up period of the trial"- taken from the article.

Matt Sparks said...

Actually, the late dialysis group had more temporary catheter use than the early group. However, I do not know about the hospitalization issue for uremic sx/volume overload. Good point.

Unknown said...

Thanks Matt. Excellent post and well put.

Like a lot of my colleagues, I am usually very comfortable waiting 'as long as possible'. Of course, it is necessary to make sure that the patients symptoms (if any) are mild / very much tolerable, their labs are non-life threatening and they are reliable with their diets / follow up appointments.

I often explain this to patients as the "gray-zone" and explain the pros / cons of waiting longer... I find it is important to encourage a partnership with the patient in regards to the decision to start RRT.

This study has limitations.. But, a study like this I believe can make us feel more comfortable about waiting longer, which as we know.. may patients prefer.

Thanks again for the mention.. and keep up the great work here!

Simon

Tarun said...

The QOL index was not used to measure quality of life during the 'wait period' for the late initiation group (about six months). It appears it was for the post initiation follow-up of both the groups after they had been on HD for a total follow-up of about 3 years. I think it would have been important to see how the QOL compared between early initiation and waiting, during the 'wait period'.-Tarun

Matt Sparks said...

From my read of the study rationale and design

http://www.pdiconnect.com/cgi/content/abstract/24/2/176?ijkey=d1115ded3ca257ddbafbf92583b679da4c0f019e&keytype2=tf_ipsecsha

"QOL data was collected in both groups every 3 months". Although the data was not expressed in table/figure form. It was mentioned in the NEJM text that no difference was seen in both groups."No significant difference in fluid and electrolyte disturbances was seen between the study groups, and no difference in quality of life was observed between the groups at any stage of the study"

Anonymous said...

how sensitive is the MDRD equation in differentiating between 12cc/min and 9.8 cc/min? to me, the delta in GFR between the 2 groupd was not enough to justify any real conclusions.

Matt Sparks said...

Agreed completely with you about this. I think this is also a good point to make. Looking at an ESTIMATED GFR value, especially at this low of renal function, is likely not very accurate. The editorial points this out as well. Again, another reason we cannot just look at the number. I have seen in many instances were the MDRD eGFR can fluctuate between 2-4 points in a few months. I think the real point to make is that picking an arbitrary number to start RRT is incorrect, but looking at multiple issues and choosing to start RRT is more of an individual patient by patient decision. The paper has flaws for sure, but this trial represents the first real randomized attempt to look at this fundamental question.

Anonymous said...

Let me add a developing country perspective too. When RENAL CARE appears to be impossible to be funded by Insurance (public or private), imagine funding it out of pocket. It is the reality in India. Add a higher prevalence of vegetarian diet - like prevalent in India. You can actually stretch conservative care quite a number of years. IDEAL trial only vindicated patient choice in this part of world and also assured Nephrologists that they may making the right choice scientifically too.