EDITORIAL
SA JOURNAL OF DIABETES & VASCULAR DISEASE
4
VOLUME 7 NUMBER 1 • MARCH 2010
urinary albumin/creatinine ratio should be determined from
the first morning specimen, to avoid effects of posture on
albuminuria. Although microalbuminuria is defined as two
or three abnormal samples, most clinicians from a practical
perspective base decisions on a single sample. It is unnecessary
to do routine urea, sodium, chloride and bicarbonate testing
as part of a creatinine, urea and electrolyte test. These
additional tests are seldom necessary and raise the cost of
screening nearly threefold.
Several studies in South Africa show that screening in high-
risk patients, both in the private and public sector is sporadic.
6,7
In the private sector, a survey showed that at least 10.4 and
32.2% of diabetics had unrecognised macroalbuminuria (or
overt nephropathy) and microalbuminuria, respectively.
6
Which eGFR?
There are several methodological issues relating to the
determination of eGFR. The two commonly used equations
are the Cockcroft-Gault (CG) and the modification of diet
and renal disease (MDRD), each with its strengths and
weaknesses.
8
The CG was established using creatinine
clearance as a reference, which tends to overestimate the
GFR, particularly in more advanced CKD, due to increased
tubular excretion of creatinine. It also does not take into
account the changes in standardisation of serum creatinine
measurement and the new normal ranges. It requires the
weight of the patient and as this is seldom added to request
forms, laboratories cannot automatically calculate the eGFR.
On the other hand, the patient’s weight may provide better
accuracy for determination of eGFR, particularly in wasted
patients, who have lower production of creatinine, but will
overestimate it in oedematous patients.
The MDRD was largely developed in the USA and a
correction factor has been added for African-Americans.
There are also differing models but importantly, weight is
not part of the equation, enabling automatic calculation by
laboratories, and it provides for standardisation of creatinine
estimation.
Van Deventer
et al
. compared the MDRD and CG formulae
to calculate eGFR with chromium-labelled EDTA clearances in
black South Africans.
9
They found good correlation with the
chromium-labelled EDTA and the CG eGFR, but the MDRD
showed poor correlation if the correction factor for African-
Americans was applied. Good correlation was found if this
was not used and their recommendations were to omit the
correction factor in black South Africans.
In summary, both the MDRD and CG forumulae can be
used to calculate eGFR, but the MDRD is more useful for
laboratory-based calculations as the patient’s weight is
unnecessary. Another critical point not appreciated is that
eGFR calculations must be used only in patients with stable
renal function, and eGFR may fluctuate, especially in the
45- to 60-ml/min range because of the effects of high meat
protein intake.
3
Patients should avoid eating meat prior to
estimation of creatinine levels.
Monitoring of CKD
Lewis provides practical guidelines for frequency ofmonitoring
of CKD. In essence, patients with early stages are seen less
frequently and those with advanced disease more frequently.
Patientswith stage 5CKDneed tobe seen at leastmonthly. The
important goals of monitoring are to ensure blood pressure
control, reduction of proteinuria or albuminuria, stabilisation
of eGFR and avoidance of treatment-related complications,
for example, hyperkalaemia and acute deterioration in renal
function due to ACE inhibitors and other nephrotoxic agents
such as radiocontrast.
The introduction of ACE inhibitors or angiotensin
receptor blockers (ARBs) in patients with CKD may result in
a deterioration in renal function, especially if the creatinine
is above 126
µ
mol/l.
10
If this is less than 20 to 30%, the ACE
inhibitor/ARB may be continued, provided the eGFR does not
further deteriorate and the serum potassium (K
+
) is normal.
If there is a more severe deterioration, these drugs should
be withdrawn, and renal artery stenosis or over-treatment of
blood pressure must be considered.
Referral to a nephrologist
Lewis presents a good practical outline of reasons for referral
to a nephrologist.
3
Firstly, there needs to be early referral of
patients for treatment of advanced CKD, and evaluation and
assessment for dialysis and transplantation. Too often, referral
occurs when urgent dialysis is needed and this is associated
with a very high first-year mortality rate.
1
Ideally, referral
should occur when the eGFR is between 30 and 60 ml/min.
Early referral has been associated with better outcomes
as these patients receive appropriate management of renal
bone disease and anaemia, prevention of cardiovascular
disease, and early creation of vascular access for dialysis.
11,12
In addition, pre-emptive renal transplant with a living donor
often provides the best outcomes.
13
The second reason for the referral is to seek clarification
when there is diagnostic uncertainty about the cause of CKD.
Diabetics can develop other primary renal diseases and they
may also be more prone to renal artery stenosis.
Clues to primary renal disease are shown in Table 2. These
are similar but more extensive than those recommended by
Lewis and include the presence of co-existent HIV infection.
Prevention of progression of diabetic nephropathy
Wright and Vardhan provide practical advice for the
Table 2.
Clues to primary renal disease
Presence of haematuria and red cell casts
•
Heavy proteinuria – urinary protein/creatinine ratio > 0.2 g/mmol
•
Abnormal renal ultrasound
•
Rapid deterioration in renal function
•
Sudden deterioration of renal function after ACE inhibitor or ARB
•
Resistant hypertension
•
HIV infection
•
