STARZ Neonatal AKI Risk Stratification Cut-off Scores for Severe AKI and Need for Dialysis in Neonates

Sidharth Kumar Sethi, Rupesh Raina, Sanjay Wazir, Gopal Agrawal, Ananya Vadhera, Nikhil Nair, Kritika Soni, Abhishek Tibrewal, and on behalf of the TINKER Working Group Pediatric Nephrology, Kidney Institute, Medanta, The Medicity Hospital, Gurgaon, Haryana 122001, India; Akron Children’s Hospital, Akron, Ohio, USA; Department of Neonatology, Cloudnine Hospital, Gurgaon, Haryana 122001, India; Maulana Azad Medical College, New Delhi, India; Case Western Reserve University School of Medicine, Cleveland, Ohio, USA; and Akron Nephrology, AGMC Cleveland Clinic, Cleveland, Ohio, USA


INTRODUCTION
N eonatal acute kidney injury (AKI) is a significant pathology associated with higher mortality rates, longer neonatal intensive care stay, and worse clinical outcomes. 1,2 In order to mitigate the avoidable outcomes, it is important to identify AKI early and start early therapeutic measures. 2,3 There have been previous attempts to derive illness severity scores, such as the Clinical Risk Index for Babies, the Simplified ageweight-sex score, Pediatric Risk of Mortality and Pediatric Index of Mortality among neonates and children. 4 These scores assess the illness severity, and cannot be used for the risk stratification for AKI or mortality. Risk of mortality in low birth weight neonates has been predicted by the NMR-2000 score, which was validated for use in low to middle income countries. 5 A specific score for AKI risk stratification in older children, Renal Angina Index, which uses the reduction in estimated creatinine clearance, fluid balance and high-risk disease states, has been shown to predict AKI accurately among various high-risk disease states. 6 Neonatal AKI risk scores are imperative to help predict which neonates are at a high-risk and should have early directed interventions. The STARZ score predicts the risk of AKI in neonates with high sensitivity (92.8%), specificity (87.4%), positive predictive value (80.5%), negative predictive value (95.6 %,) and accuracy (89.4%), which allows for its validation for use in low to middle income countries to facilitate the rapid identification of at-risk neonates. 7,8 The variables of the STARZ score are shown in Table 1. This research letter reports cut-off scores required for identifying risk of severe AKI and dialysis need in neonates. The methodology and statistical analysis of the study is provided in the Supplementary Material.
The comparison of different variables among neonates with severe AKI (AKI stage 3) versus mild to moderate AKI (AKI stages 1-2) is presented in , P < 0.001) were significantly higher among those with severe AKI than among those mild to moderate AKI. The median (IQR) STARZ score was found to be significantly higher (67 [54-77] vs. 50 [40-61]. P < 0.001) among neonates with severe AKI than among those with mild to moderate AKI. The best cut-off value STARZ score for severe AKI was found to be 59 with a sensitivity of 71% and specificity of 70% and with an area under the receiver operating characteristic curve of 0.755 (95% CI: 0.704 -0.806, P < 0.001) ( Figure 1). The median (IQR) time to AKI was observed to be significantly lower (1 [1][2][3] vs. 3 [1][2][3] days, P < 0.001) among neonates with severe AKI than among those with mild to moderate AKI. were significantly higher among those with severe AKI than among those with mild to moderate AKI. As expected, the median (IQR) STARZ score was observed to be significantly higher (77 [71-84] vs. 64 [50 -77], P < 0.001) among stage 3 AKI neonates treated with PD than among those treated without PD. The best cut-off value STARZ score was found to be 66 with a sensitivity of 97% and specificity of 52% for PD use with an area under the receiver operating characteristic curve of 0.804 (95% CI: 0.738 -0.870), P < 0.001) (Supplementary Figure S2) Table 2 and Supplementary Table S2).
To summarize, we found the following cut-offs for neonatal AKI prediction: STARZ score less than 31.5 predicts low probability of AKI; STARZ score less than 59 predicts low probability of severe AKI, and STARZ score less than 66 predicts low probability of severe AKI with the need for PD. The cut-off scores were found to increase with increased AKI severity. Similar studies to derive cut-offs to predict severe AKI and need for dialysis have been done with urine neutrophil gelatinase-associated lipocalin at admission in adults. 9 To our knowledge, this is the first of its kind study to use a scoring system that can easily be replicated in NICU. Nevertheless, further studies are needed to validate the cut-off scores. These cut-offs can help a clinician to determine the need for dialysis requirement, anticipate severe neonatal AKI and acts as a beneficial and easy clinical adjunct to neonatal intensive care units of all types.

DISCLOSURE
All the authors declared no competing interests.

AUTHOR CONTRIBUTIONS
All authors made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data; drafting the article or revising it critically for important intellectual content. All authors gave final approval of the version to be published.

ACKNOWLEDGMENTS
This manuscript is the result of support from the ISN Clinical Research Grant.

SUPPLEMENTARY MATERIAL
Supplementary File (PDF) Figure S1. Flow of the study Figure S2. Area under the ROC curve for stage 3 AKI with peritoneal dialysis Table S1. Demographic profile of the neonates included in the study Table S2. Comparison of different variables among stage 3 AKI neonates with versus without peritoneal dialysis.