Clinicians managing chronic kidney disease (CKD) have long relied on estimated glomerular filtration rate (eGFR) as the cornerstone of both diagnosis and staging. Over the past two decades, substantial effort has been devoted to refining the equations that generate these estimates, producing tools that have standardized CKD classification and improved cross-center communication. Yet a growing body of evidence suggests that for practicing nephrologists, a single eGFR value, however precisely calculated, may be less actionable than the trajectory it describes. A linked multicentre prospective study published in The BMJ, conducted by Scandrett and colleagues, examined this question directly in a large United Kingdom cohort, with findings that challenge how eGFR monitoring is currently structured in clinical practice.

The central argument advanced in the accompanying editorial is straightforward: in CKD care, the clinically consequential question is rarely what the glomerular filtration rate (GFR) is at a single time point, but rather how rapidly it is declining. Treatment decisions, risk stratification, and preparation for kidney replacement therapy (KRT) all depend on reliable longitudinal assessment rather than cross-sectional precision. In that context, the accuracy of the slope of GFR decline carries more clinical weight than the exactness of any individual measurement.

The Cohort and What It Revealed

Scandrett and colleagues followed a large UK cohort of patients with moderate CKD, specifically those classified at stage 3, over an extended follow-up period. Stage 3 CKD, defined by an eGFR between 30 and 59 mL/min/1.73m², encompasses a broad and heterogeneous population. Some patients within this range will remain stable for years; others will progress to end-stage kidney disease (ESKD) within a relatively short interval. The clinical challenge is distinguishing between these trajectories early enough to intervene.

The study focused on the eGFR slope as a predictor of outcomes and assessed how monitoring frequency and measurement variability influenced the reliability of slope estimates. The findings, consistent with prior observational data from the Chronic Kidney Disease Prognosis Consortium (CKD-PC), indicate that eGFR slope calculated from routine clinical measurements carries substantial prognostic value, but that this value is highly sensitive to the number and timing of measurements used to construct it.

A slope derived from two measurements separated by a brief interval is substantially more susceptible to biological variability and measurement error than one derived from five or six measurements over two to three years. This observation has direct implications for how nephrology services are organized and how frequently patients at different stages of CKD should be seen.

Limitations of Current Monitoring Frameworks

Current Kidney Disease: Improving Global Outcomes (KDIGO) guidelines recommend monitoring frequency based on CKD stage and risk category, with higher-risk patients seen more frequently. This framework is rational, but it was designed primarily to guide staging and medication adjustment rather than to optimize the statistical reliability of slope estimation. The distinction matters.

If the clinical goal shifts from capturing a point estimate to characterizing a trajectory, the monitoring schedule must be designed with that goal in mind. Measurements need to be spaced appropriately to separate true functional decline from the noise introduced by day-to-day physiological variation, acute illness, dietary changes, and the effects of medications such as renin-angiotensin-aldosterone system (RAAS) inhibitors, which can acutely lower eGFR without reflecting true nephron loss.

The 2024 KDIGO CKD guidelines acknowledged the prognostic utility of eGFR slope but stopped short of providing specific operationalized recommendations for slope-based monitoring protocols. The Scandrett et al. study provides empirical data that may help fill that gap, offering a basis for more prescriptive guidance in future guideline iterations.

eGFR Equations and Their Limitations

The debate over eGFR equations has not been fully resolved. The CKD-EPI 2021 creatinine equation removed race as a variable following sustained scrutiny of the biological assumptions underlying its inclusion, a decision that generated both support and methodological concern. The CKD-EPI cystatin C and combined creatinine-cystatin C equations offer improved accuracy in certain populations, particularly those with reduced muscle mass, obesity, or dietary patterns that confound creatinine-based estimates.

In Hawaii and across the broader Pacific Basin, these concerns are not abstract. Native Hawaiian and Pacific Islander patients have historically been underrepresented in the cohorts used to derive and validate eGFR equations. Dietary patterns, body composition, and comorbidity profiles in these populations differ in ways that may systematically bias creatinine-based estimates. Clinicians practicing in Hawaii face a practical question that national guidelines do not fully address: which equation provides the most reliable slope estimate in their specific patient population.

Cystatin C, while more expensive and less universally available than creatinine, is less susceptible to muscle mass confounding and may yield more stable slope estimates in populations where creatinine-based estimates are inherently noisier. The American College of Physicians and the American Society of Nephrology have both encouraged broader cystatin C access, and the 2021 NKF-ASN task force recommended its use as a confirmatory test. Whether Hawaii-based nephrology practices have moved toward routine cystatin C measurement as a complement to creatinine-based eGFR remains an open institutional question.

Slope as a Surrogate Endpoint

Beyond its clinical utility, eGFR slope has gained traction as a surrogate endpoint in clinical trials. The U.S. Food and Drug Administration (FDA) issued guidance in 2023 supporting the use of eGFR slope as a surrogate endpoint for accelerated approval in CKD trials, contingent on a sufficiently large treatment effect. This regulatory shift has accelerated the design and initiation of trials in diabetic kidney disease and IgA nephropathy, two conditions with substantial disease burden and limited approved therapies until recently.

The approval of sparsentan for IgA nephropathy, finerenone for chronic kidney disease associated with type 2 diabetes, and the expanded use of sodium-glucose cotransporter-2 (SGLT2) inhibitors such as dapagliflozin and empagliflozin across CKD stages have all been supported by slope data. The DAPA-CKD and EMPA-KIDNEY trials both demonstrated statistically significant reductions in eGFR slope in their respective treatment arms, with corresponding reductions in the composite endpoint of sustained decline in eGFR, ESKD, and kidney-specific or cardiovascular death.

For clinicians, the implication is that therapies now exist with demonstrated capacity to modify the slope of GFR decline. This makes accurate slope assessment not merely a prognostic exercise but a means of evaluating therapeutic response. A patient on an SGLT2 inhibitor who continues to show a steep decline in eGFR slope despite treatment adherence may warrant reassessment of diagnosis, additional workup, or escalation to combination nephroprotective therapy.

Practical Implications for Nephrology Practice

The Scandrett et al. findings, in conjunction with the broader literature, support several practice-level conclusions. First, eGFR monitoring schedules should be designed with longitudinal slope reliability in mind, not simply as periodic checkpoints for staging. This may require more frequent measurements in early follow-up for newly diagnosed CKD patients, followed by spacing calibrated to the individual’s apparent trajectory.

Second, a minimum of three to four eGFR measurements over at least one year should be considered before drawing firm conclusions about the rate of decline. Decisions of major consequence, such as referral for KRT evaluation or initiation of dialysis access planning, should not rest on a slope derived from two data points, particularly if those measurements were obtained during a period of acute physiological stress.

Third, the documentation and visualization of eGFR slopes in the electronic health record (EHR) remain inconsistent across institutions. Many EHR systems display eGFR as a series of discrete values rather than as a plotted trajectory with a calculated slope. Nephrology practices and health systems should consider whether their current EHR configuration supports the kind of longitudinal visualization that makes slope-based clinical reasoning practical.

Fourth, patient education should reflect this longitudinal framework. Patients often interpret a stable eGFR as reassurance and a declining one as cause for alarm, without context about the rate or clinical significance of the change. Structured conversations about trajectory, rather than threshold, may improve patient engagement with monitoring and adherence to nephroprotective therapies.

What Remains Unresolved

The Scandrett et al. study, along with the editorial commentary it accompanies, advances the field but does not resolve all outstanding questions. The optimal measurement interval for slope construction, the minimum number of observations required for a clinically reliable estimate, and the best statistical method for fitting the slope in the presence of acute perturbations all require further investigation. Comparative data on slope reliability across eGFR equations in diverse populations, including those represented in Hawaii’s patient mix, are notably limited.

Regulatory and payer frameworks have not yet fully aligned with a slope