Since the participating centers in this study cover both rural and urban segments of the Greek population, it is possible that some of these differences directly reflect some heterogeneity in the distribution of risk factors for CKD, local practice patterns of referring WY 14643 physicians and nephrologists and models of cooperation between them. Nevertheless, the fact that such differences do exist, suggest a missed opportunity for standardizing care at a national level by e.g. CKD diagnosis and treatment educational programs aiming at non-specialists and promotion and adoption of guidelines specifying thresholds for appropriate referral and blood pressure targets. The findings and interpretations in this report should be interpreted in light of certain limitations. First, the cross-sectional design of the study precludes drawing conclusions about the rate of renal function decline over time among study participants. Second, we did not have access to referring physicians’ records so we cannot explain the apparent channeling bias in referring patients with diabetes to a nephrologist. This pattern may reflect the limited understanding of the subtle manifestations of a wide spectrum of renal pathology by non-specialists, the sensitization of non-nephrologists to the renal complications of diabetes, or possibly a skewed view by non-specialists about the benefits of nephrologist co-management. A study of referring physicians could help us understand these referral patterns, suggesting one possible way for filling this knowledge gap. Third, even though the MDRD and CKD-Epi equations have been evaluated in a number of different populations and cohorts, no direct validation exists for the Greek population, so that the comparison between these equations should not be viewed as one of a method against a gold standard. Finally our study was conducted before the 2012 KDIGO classification of CKD stages along the two dimensions of eGFR and albuminuria categories, and thus we did not collect data about abnormal urinary biomarkers in our patients. Hence, we cannot exclude the likelihood that such abnormalities drive nephrology referrals at higher levels of eGFR especially for patients with glomerular disease who are more likely to manifest proteinuria and/or hematuria. In summary, we have undertaken the first national crosssectional evaluation of non-dialysis dependent CKD patients in outpatient nephrology clinics of the Greek National Health System. We found that many patients appear to be referred late by physicians, while self-referred patients consult a nephrologist at a higher level of renal function possibly due to direct access to test results. To reduce the burden of ESRD in Greece, with the 8th highest incidence rate in the world, which during the current financial crisis bears the cost of providing unfunded dialysis services to a large number of uninsured individuals including illegal immigrants future initiatives should focus on the adoption of eGFR reporting in order to facilitate early detection, appropriate confirmatory testing, and prescription of reno-protective medications in order to reduce the progression to dialysis dependency.