Criteria From General Population to Special Field till Individual

We read with great interest the paper by Zeng et al¹ The authors' innovative approach to redefining acute kidney injury (AKI) criteria in the cardiac surgery setting is commendable, particularly their identification of a higher serum creatinine change (ΔSCr) threshold (0.55 mg/dL) for stage-1 AKI, which significantly enhances the prognostic value compared with the KDIGO-(Kidney Disease: Improving Global Outcomes) defined threshold.

This study provides important insights into the need for context-specific diagnostic criteria in cardiac surgery-associated AKI (CSA-AKI).

The authors' identification of a 0.55 mg/dL ΔSCr threshold is compelling, as it aligns better with cardiac surgery-specific pathophysiology. This refinement improves prognostic accuracy for short- and long-term mortality while reducing the potential for overdiagnosis. By challenging the “one-size-fits-all” approach of the KDIGO criteria, the authors address a major gap in the field.²

The recalibrated threshold has significant implications for resource allocation and clinical care. Patients identified using the new threshold are more likely to represent clinically significant AKI, potentially improving the focus of interventions. This is particularly relevant given the increasing burden of AKI diagnoses in an era of high-frequency electronic medical record-based monitoring.³

The study underscores the unique perioperative factors influencing CSA-AKI, such as fluid shifts and ischemia-reperfusion injury. Incorporating these nuances into diagnostic criteria represents a step forward in precision medicine.

However, several issues deserve further discussion and exploration:

While the study focuses on ΔSCr, the KDIGO AKI definition also incorporates urine output criteria. Ignoring this component may oversimplify the complexity of AKI diagnosis in the perioperative setting. Urine output is particularly relevant in the cardiac surgery population, where fluid shifts can obscure serum creatinine changes.

The reliance on ΔSCr alone, even at a recalibrated threshold, may not fully capture the multifactorial nature of CSA-AKI. Emerging biomarkers such as neutrophil gelatinase-associated lipocalin and kidney injury molecule-1 could complement ΔSCr and provide a more comprehensive assessment of kidney injury.⁴

The study does not adequately explore the variability of the proposed threshold across subgroups, such as patients with chronic kidney disease, off-pump surgeries, or differing fluid management strategies. Subgroup analyses could provide valuable insights into the generalizability of the findings.

While the study demonstrates that a higher ΔSCr threshold improves prognostic accuracy, its diagnostic utility in identifying “real” kidney injury remains unclear. A more granular analysis of clinical outcomes specific to reclassified AKI cases is needed to validate the clinical relevance of this threshold.

The authors utilized restricted cubic splines, receiver-operating characteristic analysis, and survival analysis to rigorously evaluate thresholds. While restricted cubic splines effectively captures nonlinear dose-response relationships, the exclusive use of a 30-day mortality endpoint for threshold determination may be limiting. Survival analyses extended to 5 years indicate substantial long-term prognostic implications, suggesting that employing a longer-term endpoint for threshold optimization might yield clinically robust criteria.

An important clinical phenomenon absent from the discussion involves patients experiencing acute SCr elevation postcardiopulmonary bypass surgery whose postoperative SCr levels subsequently decline, coinciding with improved cardiac output and hemodynamics.⁵ Were such cases observed in your dataset? If so, their implications for the identified thresholds merit discussion. This phenomenon could underscore transient, reversible renal dysfunction rather than significant AKI, complicating prognosis and management.

The study's sensitivity analysis revealed differential 48-hour ΔSCr thresholds for on-pump (0.628 mg/dL) vs off-pump (0.480 mg/dL) procedures. Given these distinctions, is a unified threshold appropriate for assessing both surgical types? Moreover, the threshold's reliability in cases with combined procedures, where complexity and operative risk escalate, deserves further scrutiny. For such scenarios, should a higher threshold be considered to balance specificity and clinical relevance?

Furthermore, while the authors report superior prognostic performance of the proposed threshold relative to KDIGO criteria, the observed AUCs (eg, 0.757 for 30-day mortality) indicate room for improvement. Incorporating additional variables such as perioperative fluid balance or biomarkers like neutrophil gelatinase-associated lipocalin might enhance predictive accuracy.

In conclusion, the recalibrated ΔSCr threshold proposed by Zeng et al¹ offers promising improvements in prognostic utility, particularly in the cardiac surgery population, and highlights the importance of refining diagnostic criteria to address context-specific nuances. Building on these findings, future research may consider prioritizing multicenter validation of the proposed threshold across diverse cardiac surgery populations, exploring the integration of biomarkers into AKI definitions, and assessing the utility of combined diagnostic criteria incorporating ΔSCr, urine output, and other relative biomarkers. Additionally, investigations into the cost-effectiveness and real-world applicability of these revised criteria are essential to optimize clinical impact. We highly commend the authors for their valuable contribution to this critical area of research and anticipate that their work will inspire further advancements in the diagnosis and management of perioperative kidney injury.