Tuesday 27 January 2009

Q & A: CKD and CKD Coding

Q: Hi Donal, I wanted to mention a few problemS with eGFR and the over 80s and coding of CKD.
  • In my practice, 45-60 is the median value for this age group and 30-45 is well inside the "bell shaped curve". Could the normal values or guidelines be adjusted to take account of the normal reduction in renal function with age?
  • I have a problem with the CKD codes. These are all number codes - daughter codes of 1Z1 and all at the same level. All number codes in read are codes to do with "Process of Medicine" and all disease codes start with a letter. The letter code for GU system diseases is K, and impaired renal function disorder is K08 and its daughter codes. Codes starting with the number 1 are all Symptom codes. This means that if I search for people with impaired renal function, no-one with a CKD code would be found.
  • The other problem with the codes all being in the same level is that logical daughter codes do not show. By this I mean CKD 3 would naturally be subdivided into CKD 3 with proteinuria and CKD 3 without proteinuria, these latter two should be daughter codes of CKD 3 rather than on the same code level.
  • I know that I have mentioned this previously but eGFR >90 does get equated to CKD 1 and eGFR 60-90 to CKD 2. Now I know that these would only normally apply in the presence of proteinuria. It seems strange therefore that there are codes for CKD1 without proteinuria (1Z18) and CKD 2 without proteinuria (1Z1A).

Dr Merlyn Wilcox, GP Partner, GP trainer and Hon Senior Lecturer in Primary Care and General Practice, South Birmingham PCT

A: Dear Merlyn, many thanks for your questions. I do appreciate that MDRD formula has limitations and is perhaps best considered as a population tool. Perhaps most important is whether or not the population being identified has increased morbidity and mortality - I think it is likely due to associated cardio-vascular disease.

I completely agree with regard to the coding. The statement "well I wouldn't start here" springs to mind. I think you are probably more familiar with the coding than I am but my understanding is also that codes beginning with 1 are "history/symptoms". Those more familiar with coding have advised that once something is in the hierarchy it is generally best left there. In CKD Stages 1 and 2, there needs to be evidence of kidney damage and while that is often signalled by a raised albumin creatinine ratio in the urine, other markers of damage can occur without proteinuria – such as ultrasound evidence of scarring from Pyelonephritis or Polycystic Kidney Disease. You may or may not know that there is a kidney group actively working on a SNOMED CT approach and I think will be a major advance.

The recent NICE guidance stating that ACR is the proteinuria test of choice and the inclusion of annual ACRs as part of the CKD section of the Quality and Outcomes Framework will soon result in many more quantitative proteinuria tests and at that stage we will be able to move towards considering the CKD population in a more dynamic way. The rate of change of eGFR for individuals is perhaps more important than the absolute value or indeed whether or not that value is in the "normal" range.

The situation is moving rapidly and your comments and observations are helpful to keep the momentum in the correct direction.