In this nationwide real-world study of practice patterns for kidney function assessment in Brazilians covered with medical insurance, we found that 41% of individuals undergoing creatinine testing lack an accompanying urinary test, a combination of tests that would comply with international guidelines. This incomplete CKD assessment is more common in patients with higher eGFR, who are often not followed by kidney specialists, and presumably are at lower risk of CKD. Most laboratory evaluations of kidney function are performed by non-nephrologists, namely internists, cardiologists and OBGYNs, even among patients with presumable advanced CKD. Finally, up to 80% of patients with at least two eGFR assessments have a follow-up creatinine test within 12 months of the first; the lower the eGFR, the more frequent the creatinine assessment.

Although our study was not designed to evaluate the prevalence of CKD, our results suggest that approximately 11% of patients in Brazil have CKD, assuming no false positive results due to single eGFR and pACR assessments. These results are consistent with the estimates of 10-12% worldwide CKD prevalence [9]. They are also similar to two prior investigations of CKD prevalence in Brazil; one estimated an 8% prevalence in a cohort of civil servants [10], and the other reported an 11.4% prevalence in a convenience sample from a working setting [11].

In Brazil, a country with universal public coverage of healthcare services, nearly 150,000 (640 per million population) patients are on kidney replacement therapy (KRT) as of 2020 [12]. The number of patients on KRT in the country have steadily increased by 6% per year over the last two decades [13]. In this period, mortality rates among patients on hemodialysis, the most frequent KRT modality in the country, increased from 17% to 20% [12]. CKD and its associated conditions consume about 13% of all public health resources in Brazil [14].

Our results imply that approximately 21 million persons are living with CKD in Brazil, a country with roughly 4,000 registered nephrologists. Thus, it is expected that non-nephrologists will screen and assist most CKD patients in Brazil, even those with more advanced stages [13],a problem not exclusive to non-high-income countries [15].In fact, our analysis suggests that, in absolute numbers, nephrologists contribute to approximately 1% of eGFR and pACR assessments in the private setting in Brazil. In general, OBGYN physicians ordered a large absolute number of eGFRs and pACRs, which may be explained by the fact that OBGYNs lead an important fraction of primary care among women in Brazil. In absolute numbers, nephrologists ordered most eGFR assessments only among patients with single eGFR ≤ 15 mL/min/1.73 m2. In stages G3a and G3b, in which new therapies that prevent kidney failure can be implemented [5, 6], cardiologists and internal medicine specialists were the most frequent physicians to monitor or screen for CKD. The relatively low absolute contribution of nephrology to CKD screening or monitoring may be explained by the referral structure in Brazil, in which patients with eGFR below 30 mL/min/ 1.73 m2 are referred to nephrologists. Additionally, our results suggest that several patients with eGFR below 15 mL/min/ 1.73 m2, who were mostly monitored by nephrologists, were tested for urinary albuminuria. The value added by albuminuria measurements in this advanced CKD population is uncertain because no interventions to reduce albuminuria have been proven to delay CKD progression in this context [4]. Thus, this practice may not necessarily reflect the most cost-efficient strategy to monitor advanced CKD patients. Our analysis suggests that efforts to increase CKD awareness and promote screening should target non-nephrologists in absolute terms but should also focus on incentivizing best practices among nephrologists. However, there are conflicting recommendations for CKD screening among distinct societies representing medical specialties.

Diverse guidelines from cardiology, internal medicine, and nephrology have shaped the varied implementation of CKD screening, mostly diverging on priorities. The American College of Physicians (ACP) advises against CKD screening in the general population due to a lack of evidence on its risks and benefits, warning it may lead to overdiagnosis and added costs [16]. KDIGO highlights that using cystatin-C to measure eGFR can reduce misclassification and overdiagnosis risks, advocating for CKD screening in patients with a wide array of risk factors for CKD [4]. Emphasizing the clinical advantages of early interventions, KDIGO supports the prioritization of efforts to implement CKD screening [4]. Meanwhile, the 2021 European Society of Cardiology (ESC) Guidelines recognize CKD as a significant cardiovascular risk factor, suggesting its evaluation in broader cardiovascular screening [17]. They underline the value of albuminuria and low GFR results, which can inform strategies, especially with new treatments that reduce cardiovascular events and preserve kidney function [17]. Reflecting the ESC's stance, the European Renal Association now urges a more proactive CKD screening in the wider population [18].

However, barriers to broader CKD screening and monitoring prevail. Our analyses indicate that up to 41% of patients who had an eGFR evaluation did not undergo a reassessment of kidney health within 12 months. This pattern echoes findings from US and UK studies among CKD patients [19, 20]. The proportion of patients undergoing evaluations for eGFR and pACR grows with the severity of CKD stages, likely due to the increasing involvement of nephrologists in advanced CKD care. In fact, a US survey found that non-nephrologists often underestimate the significance of albuminuria in early CKD stages [21]. This uncertainty can hinder timely education, planning, and interventions critical for CKD progression and cardiovascular risk reduction [4]. Importantly, even among patients with a low risk for kidney failure (e.g., those with eGFR > 60 mL/min/ 1.73 m2), many may still have CKD according to KDIGO guidelines [4]. Our results suggest that approximately 54% of patients with eGFR > 60 mL/min/ 1.73 m2 were tested with a urine test. Among this tested population with eGFR > 60 mL/min/ 1.73 m2, approximately 2.3 % had results consistent with CKD (N= 59,818). The absence of an albuminuria measurement in this context could result in missed diagnoses of CKD.

This study has several noteworthy limitations. First, we relied on secondary data extraction from an administrative database for our study. We did not have detailed clinical information on participating patients, nor were we able to assess the reasons for laboratory evaluation. Therefore, we were unable to evaluate whether creatinine or albuminuria were ordered for screening or monitoring purposes. Second, our evaluation of medical specialties is limited to the reported information on the laboratory requisition. In Brazil, internal medicine may broadly cover unspecialized physicians working on primary care, specialists in primary care (family medicine), or specialists in internal medicine. Third, our database covers primarily insured patients, not capturing users of the unified public healthcare system in Brazil. This can explain the lack of family medicine practitioners as a common medical specialty requesting laboratory tests in our sample, as these physicians are more often leading the care for patients in the public healthcare system. Fourth, the patient population in this study represents relatively young and healthy patients, who are presumably under a low risk for CKD. Therefore, our results do not necessarily generalize to higher risk settings in Brazil. Fifth, our sample mainly comprised patients from the Southeast region in Brazil, which may limit the external validity of our findings. Sixth, the GFR was estimated by the CKD-EPI equation although serum creatinine was measured by the Jaffé method instead of the isotope dilution mass spectrometry (IDMS)-traceable Jaffé method. Finally, we used single assessments of creatinine and pACR to define CKD. Despite these limitations, our results have key strengths. Our large sample covers a wide geographic distribution in Brazil, which increases the external validity of our findings. Additionally, our results are highly consistent with prior studies, which further reinforces the advantages of using large real-world databases for epidemiological research. Therefore, we believe our data can support action plans to reduce the growing burden of CKD and its complications in Brazil.

In conclusion, in this real-world study evaluating practice patterns for laboratory evaluation of biomarkers of kidney damage and function, we found that non-nephrologists are the key drivers of CKD assessment in Brazil and that CKD assessment is largely non-compliant to current guideline recommendations, since more than one-third of patients with eGFR evaluation lack urinary tests. Our results suggest that a large proportion of patients with CKD remain undiagnosed in Brazil, particularly those with proteinuric CKD with eGFR above 60 mL/min/1.73 m2. Furthermore, most patients at risk of progressing to kidney failure and suffering poor outcomes are sub-optimally screened and risk-stratified in Brazil. Increasing the awareness of the importance of CKD diagnosis and coordinating multidisciplinary efforts for screening, risk stratification and referral strategies remain unmet goals in CKD care that require urgent attention.