Blog entry by dianny elizabeth
A decade has now passed since the epoch-making introduction of the concept of a generic “chronic kidney disease”(CKD) based on estimates of glomerular filtration rate (eGFR) and the presence of “kidney damage” persisting for 3 months or more (NKF-KDOQI-2002). An immense amount of epidemiological data, accumulated in millions of subjects, has supported the notion that CKD so defined is described as “common, harmful and treatable.” Much smaller amounts of information, from randomized controlled trials and well-designed observational studies have suggested that treatment, directed at lowering blood pressure or reducing proteinuria, can have a salutary effect on survival and complications, at least in certain conditions (most notably in advanced diabetic nephropathy and in nephrotic syndrome due to primary and other secondary glomerular diseases).
These revelations have generated much enthusiasm for widespread attempts to identify CKD early in its course in the general population, using the vehicle of screening for albuminuria and/or reduced eGFR. While little doubt exists that such programs would enhance the visibility of CKD in society as a whole, questions still linger as to whether such early identification among subjects, not currently under medical care for specific diseases causing progressive CKD, would yield benefits to society in a cost-effective manner.
To my knowledge, no prospective trial of screening vs no screening for generic CKD (i.e. not cause specific), with patient-centered outcomes of prolongation of life expectancy or avoidance of serious (and possibly fatal) complications (such as acute kidney injury or ischemic heart disease), has yet shown such benefits.
Translating statistically valid associations derived from large cross- sectional epidemiological (observational) studies to real-life clinical situations can be hazardous, as many unrecognized biases and confounding variables can creep into the data, however diligent the investigators are in avoiding these shortcomings.
Thus, it seems to me, that the time is ripe to ask some hard questions and to advocate efforts to secure answers to them--- rather than to assume that society needs mass screening for CKD and needs it now.
The Questions are:
1) Are the current estimates of prevalence of CKD (at each Stage and according to age and gender) in the general ambulatory population( i.e. asymptomatic candidates for screening) correct?- here I would choose to define CKD as a generic entity having some meaningful impact on survival or meaningful patient-centered morbidity
2) Is the incidence of newly diagnosed CKD (at each Stage and according to age and gender) increasing in the general population and at what rate? Here, data from developing and developed nations would be desirable.
3) Are the current criteria used to diagnose CKD appropriate and equally applicable to subjects of all ages, genders, ancestral groups and geographic locations? What are the most appropriate thresholds of eGFR and albuminuria for defining CKD as a target for specific interventions to reduce progression or to avoid complications?
4) Will periodic screening of subjects in the general population, without known risk factors such as diabetes or hypertension, using eGFR and/or albuminuria lead to improved outcomes compared to no screening and accomplish this in a cost-effective manner.
5) Will screening of subjects in the general population for high blood pressure or diabetes yield superior long- term outcomes compared to screening for generic CKD using absolute threshold for eGFR and/or albuminuria as CKD defining criteria?.
I submit that until these Questions are answered widespread screening of the general population should not be advised. The risks of an uncertain number of false positives, leading to unnecessary investigations and promoting unwanted anxiety among screenees is too great. Un-validated re-assurance for the false negative screenees is another potential risk. An eGFR of >60ml/min without proteinuria, currently regarded as indicating the absence of CKD in many epidemiological studies, does not exclude renal disease, especially in the young (e.g. Autosomal- dominant Polycystic Kidney Disease).
Simple tools (such as the Framingham Risk Score) are already widely available to identify and quantify, in absolute terms, the time-related risk of cardio-vascular disease. Adding eGFR or albuminuria to these tools contributes only marginally to improving their accuracy in classifying risk in the general population but these same simple tools may underestimate risk in patients treated by dialysis.
CKD, as currently defined, is largely a ”condition” of older and elderly adults, predominantly identified in these groups by a slight reduction in eGFR (using creatinine-based methods) and no proteinuria which might be normal for the subjects age and gender. One could also challenge the appropriateness of including “isolated microalbuminuria” (i.e., normal eGFR and urinary:albumin to creatinine ratio of 30-300mg/gm) as defining a “kidney disease” especially as these values can fluctuate widely and are influenced greatly by air pollution, diet, exercise, inter-current illness, obesity, high blood pressure, underlying atherosclerosis, remote infection and others. No one has yet shown that interventions directed to correcting “isolated microalbuminuria” per se have a beneficial effect on patient-centered outcomes (such as ESRD or cardiovascular disease), indeed some studies have suggested the potential for harm. Amelioration of “Isolated microalbuminuria” is not yet an acceptable surrogate end-point for studies evaluating therapeutic interventions.
Therefore, I urge caution in jumping to conclusions that early identification of many forms of generic CKD, using current diagnostic criteria, will have a net beneficial impact on the overall health of society, at a reasonable cost. We need some clear answers to some hard Questions first. Assuming that universal screening for CKD is without material risk for harm may not be correct. Advocates of such screening do so at their own risk.
Richard J. Glassock, MD
Geffen School of Medicine at UCLA
March 17, 2012