External validation and comparison of clinical prediction models for cisplatin- associated acute kidney injury: A single-centre retrospective study

External validation and comparison of clinical prediction models for cisplatin- associated acute kidney injury: A single-centre retrospective study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article External validation and comparison of clinical prediction models for cisplatin- associated acute kidney injury: A single-centre retrospective study Kazuki Saito, Satoru Nihei, Junichi Asaka, Kenzo Kudo This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6678420/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 29 Sep, 2025 Read the published version in Journal of Pharmaceutical Health Care and Sciences → Version 1 posted You are reading this latest preprint version Abstract Background: Cisplatin-associated acute kidney injury (C-AKI) is a major complication of cisplatin therapy. Although two clinical prediction models have been developed for the US population, their external validity in the Japanese population remains unclear. This study aimed to evaluate the external validity of these models and compare their predictive performances in a Japanese cohort. Methods: We assessed the performance of two C-AKI prediction models developed by Motwani et al. and Gupta et al. in a retrospective cohort of 1,684 patients treated with cisplatin at Iwate Medical University Hospital. C-AKI was defined as a ≥0.3 mg/dL increase in serum creatinine or a ≥1.5-fold rise from baseline. Severe C-AKI was defined as a ≥2.0-fold increase or renal replacement therapy initiation. Model performance was evaluated using discrimination (area under the receiver operating characteristic curve [AUROC]), calibration, and decision curve analysis (DCA). Logistic recalibration was applied to adapt the model to the local population. Results: The discriminatory performance for C-AKI was similar between the Gupta and Motwani models (AUROC, 0.616 vs. 0.613; p = 0.84). However, the Gupta model showed better discrimination of severe C-AKI (AUROC, 0.674 vs. 0.594; p = 0.02). Both models exhibited poor initial calibrations, which improved after recalibration. The recalibrated models yielded a greater net benefit in the DCA, with the Gupta model demonstrating the highest clinical utility in severe C-AKI. Conclusions: Both models demonstrated discriminatory ability, with the Gupta model showing particular utility in predicting severe C-AKI. Recalibration may be necessary before applying these models in Japanese clinical practice. clinical decision rules cisplatin acute kidney injury validation study Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Background Cisplatin (CDDP) is an important chemotherapeutic agent with proven efficacy against various solid tumours ( 1 ). One of its major dose-limiting toxicities is nephrotoxicity ( 2 ), with cisplatin-associated acute kidney injury (C-AKI) occurring in 20–30% of patients and in approximately 10% even after the first dose ( 3 , 4 ). C-AKI is associated with poor renal outcomes ( 4 ), treatment interruptions, poor prognosis ( 5 – 7 ), prolonged hospital stays, and increased healthcare costs ( 8 , 9 ). Preventive strategies, such as hydration, forced diuresis with mannitol or furosemide ( 10 ), and magnesium supplementation ( 11 ), are commonly employed to mitigate CDDP-induced nephrotoxicity. However, these measures did not completely eliminate the risk of C-AKI. Given the frequency and potential severity of C-AKI, assessing the associated risks in advance is important. Clinical prediction models, tools that estimate the probability of a specific outcome using multiple predictors, are widely used in medical practice. In 2018, Motwani et al. developed a prediction model for C-AKI using data from the US population ( 12 ). This scoring system estimates C-AKI risk based on readily available clinical variables, such as age, CDDP dose, serum albumin level, and history of hypertension. In 2024, Gupta et al. proposed a new model that incorporated additional predictors, including blood cell counts, haemoglobin levels, and serum magnesium concentration (Table 1 ) ( 5 ). In their development and validation cohorts, the Gupta model demonstrated superior discriminatory performance compared to the earlier model developed by Motwani et al. Table 1 Clinical prediction models for cisplatin related acute kidney injury Motwani et al. Gupta et al. Definition of AKI Creatinine ≥ 0.3 mg/dL in 14 days Creatinine ≥ 2.0-fold or RRT in 14 days Predictors Age (years) ≤ 60: 0 points ≤ 45: 0 points 61–70: 1.5 points 46–60: 2.5 points > 70: 2.5 points 61–70: 3.5 points > 70: 4.5 points Hypertension 2 points 1 point Diabetes – 1 point Smoker – 1 point CDDP dosage (mg) ≤ 100: 0 points ≤ 50: 0 points 101–150: 1 point 51–75: 2 points > 150: 3 points 76–100: 2.5 points 101–125: 3 points 126–150: 5 points 151–200: 7.5 points > 200: 9.5 points Hemoglobin (mg/dL) – ≥ 12.0: 0 points 11.0–11.9: 1 point 12.0: 1.5 points Albumin (mg/dL) > 3.5: 0 points > 3.8: 0 points ≤ 3.5: 2 points 3.3–3.8: 1 point < 3.3: 1.5 points Magnesium (mg/dL) – ≥ 2.0: 0 points < 2.0: 1 point Acute kidney injury (AKI) was defined as the degree of increase in serum creatinine levels from baseline. RRT, renal replacement therapy; CDDP, cisplatin; WBC, white blood cell count. The ultimate goal of such clinical prediction models is to inform clinical decision-making in real-world practice. To enable this, external validation is essential to confirm that a model performs reliably in populations or settings different from those in which it was originally developed, whether geographically or temporally ( 13 , 14 ). However, to date, neither of these C-AKI prediction models have been evaluated outside Western populations ( 5 , 15 ), including the Japanese population. The Motwani and Gupta models were developed using different outcome definitions. Severe C-AKI, the primary outcome targeted by the Gupta model, represents a more clinically significant event associated with a worse prognosis than mild C-AKI ( 5 ). Therefore, directly comparing various outcome thresholds may offer valuable insights into their suitability for clinical applications. This study aimed to evaluate the external validity of the Motwani and Gupta models in a single-centre Japanese cohort and compare their predictive performance. Methods This study followed the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis + artificial intelligence guidelines (16). The complete checklist is presented in Table S1. The patients and members of the public were not involved in the design or conduct of this study. As this was a retrospective analysis of existing clinical data without any prospective intervention, no formal study protocol was prepared, and the study was not registered in a public trial registry. All statistical analyses were performed using the R version 4.3.1. OpenAI's GPT-4o language model was utilized to assist in the development and refinement of the R code used for these analyses. The code used for the data analysis is provided in Table S2. Statistical significance was defined as p < 0.05. Data source This study used data from patients who received CDDP at Iwate Medical University Hospital between April 2014 and December 2023. The exclusion criteria were as follows: (i) age < 18 years at the time of administration, (ii) CDDP outside the study period or at another institution, (iii) treatment with daily or weekly CDDP regimens, and (iv) missing baseline renal function or outcome data. All eligible cases were included, and no a priori sample size calculations were performed. Data on patient characteristics (age, sex, height, and weight), smoking history, concomitant medications, dates and doses of CDDP administration, and relevant laboratory values (creatinine, albumin, white blood cells, platelets, haemoglobin, and magnesium) were extracted from the electronic medical records. The estimated renal glomerular filtration rate was calculated using the Japanese Society of Nephrology (17). Diabetes and hypertension were defined based on the use of the corresponding medications. The baseline laboratory values were defined as the most recent measurements obtained within 30 days of CDDP administration. Missing values were supplemented using regression-based imputation. Definition of outcomes In this study, C-AKI was defined as an increase in serum creatinine levels within 14 days of CDDP exposure. Based on the Kidney Disease: Improving Global Outcomes (KDIGO) criteria (18), C-AKI was defined as either an increase of ≥0.3 mg/dL or a ≥1.5-fold rise from baseline. Although an increase of ≥0.3 mg/dL over 14 days is a slight deviation from the strict KDIGO definition, this criterion was included to align with the outcome definition used in the Motwani model (12). Severe C-AKI was defined as a ≥2.0-fold increase from baseline or the initiation of renal replacement therapy (KDIGO stage ≥ 2). Calculation of scores and predictive probabilities The scores were calculated according to the scoring criteria listed in Table 1. For the Motwani model, predictive probabilities were derived using the baseline risk and odds ratio reported in their development cohort—specifically, an incidence rate of 0.04 at a score of 0, and an odds ratio of 1.49 per one-point increase in score. As the simple model by Gupta et al. does not provide predictive probabilities for individual scores, the predictive probabilities were estimated using the primary model, which incorporates a cubic spline. The codes used in the calculations are listed in Table S2. Statistical performance The statistical performance of the clinical prediction models was evaluated from three key perspectives: discrimination, calibration, and overall model fit (14). For each metric, the mean and 95% confidence intervals (CI) were estimated using 1,000 bootstrap samples. Discriminatory ability, the capacity of the model to distinguish between individuals who do and do not experience an event, was assessed using the area under the receiver operating characteristic curve (AUROC). Differences between the two AUROCs were tested using the bootstrap method. Calibration, which reflects the agreement between the predicted and observed risks, was evaluated using the following metric: calibration-in-the-large, which indicates systematic over- or underestimation across the model. The ideal value is 0; values 0 suggest an underestimation. The calibration slope represents the steepness of the calibration curve. A slope of 1 indicates perfect calibration, and values < 1 indicate extreme predictions in certain probability ranges. The model fit was assessed using the Brier score, which quantifies the mean squared difference between the predicted probabilities and observed outcomes. A lower Brier score indicates a better model performance, with 0 being ideal. Risk stratification with the Gupta simple model In the original study by Gupta et al., a simple model score was evaluated to stratify patients into risk groups, with accompanying expert recommendations on clinical management. By contrast, the Motwani et al. model does not propose a risk stratification system or clinical action guidelines. Therefore, we conducted an additional risk stratification analysis based on the Gupta model to explore how the risk grouping and associated clinical recommendations proposed in the original study were applied to our cohort. Patients were categorised into four groups according to their total score following the definitions from the original study as follows: low (0–5.5 points), moderate (6–9.5 points), high (10–15.5 points), and very high (≥16 points). The incidence of severe C-AKI was calculated for each group. Recalibration Recalibration is a model-updating technique that adjusts regression coefficients to reflect the target population better (19). Logistic recalibration was performed for each scoring system (20). Specifically, the intercept and slope were re-estimated using logistic regression with the outcome as the dependent variable and each prediction score as the independent variable. Recalibration did not enhance the discriminative ability because the model structure remained unchanged. Although structural updates could potentially lead to substantial improvements in model performance, model updating was limited to recalibration given that this study was based on a single-centre cohort. Clinical utility Decision curve analysis (DCA) was performed to evaluate the clinical utility of each model. In DCA, the net benefit, defined as the benefit of correct classification minus the harm of misclassification, is plotted on the vertical axis, whereas the threshold probability, the probability at which the benefits of treatment and non-treatment are considered equal, is shown on the horizontal axis. At any given threshold probability, a higher net benefit indicated greater clinical usefulness. Conversely, a negative net benefit suggests that model-guided decision-making may cause more harm than benefits. Results A total of 2,184 patients who received CDDP during the study period were identified. After applying the exclusion criteria, 1,684 patients were included in the final analysis (Fig. 1). Patient characteristics and outcome frequencies are summarised in Table 2. The primary outcome, C-AKI, was observed in 186 (11.0%) patients, and severe C-AKI occurred in 36 (2.1%) patients. The statistical performances of the clinical prediction models are summarised in Table 3, with a full comparison of all metrics provided in Table S3. Receiver operating characteristic (ROC) curves and calibration plots are presented in Figs. 2 and 3. The discriminatory performances of the Gupta and Motwani models for C-AKI showed no significant difference (AUROC, 0.616 [95% CI, 0.575–0.658] vs. 0.613 [95% CI, 0.570–0.656]; p = 0.84). By contrast, the Gupta model demonstrated significantly better discrimination of severe C-AKI (AUROC, 0.674 [95% CI, 0.584–0.768] vs. 0.594 [95% CI, 0.482–0.697]; p = 0.02). The calibration slopes and other indices revealed that both models were poorly calibrated. In addition, risk stratification based on the Gupta simple model score is shown in Fig. 4. This risk stratification analysis was exploratory and intended to illustrate the distribution of risk categories in our cohort rather than serving as a primary evaluation of model performance. Although the model demonstrated reasonable discrimination across risk groups, the overall risk distribution was lower than that of the original development cohort, with most patients categorised into moderate- or high-risk groups. Logistic recalibration was performed to adjust both models to the characteristics of our cohort; the updated coefficients are presented in Table S4. The clinical utility was evaluated using DCA for both the original and recalibrated models (Fig. 5). For C-AKI, the non-recalibrated Gupta model (dashed red line) demonstrated lower net benefit than the model-independent strategy (black line) across most threshold probabilities, whereas the non-recalibrated Motwani model (dashed blue line) yielded net harm at higher thresholds (approximately ≥12.5%). By contrast, both recalibrated models exhibited positive net benefits across a wide range of thresholds with comparable clinical utility. A similar pattern was observed for severe C-AKI; both uncalibrated models showed net harm, which improved after recalibration. The recalibrated Gupta model demonstrated the highest net benefit for predicting severe C-AKI. Discussion In this study, we conducted the first external validation of two existing clinical prediction models for C-AKI developed by Motwani et al. and Gupta et al. in a Japanese patient population. The discriminative ability of the two models for C-AKI was comparable. However, the Gupta model demonstrated superior performance in predicting severe C-AKI, which was clinically significant. Both models initially exhibited poor calibration, and the DCA further indicated that using unadjusted models for clinical decision making did not produce a net benefit. After logistic recalibration, both models showed improved clinical utility, with the recalibrated Gupta model offering the greatest net benefit to patients with severe C-AKI. These results highlight the importance of recalibrating the prediction models before their application to new populations, particularly in different geographic and clinical contexts. Clinical prediction models are typically optimised using development cohorts and perform poorly when applied to different populations. Discriminatory performance declined when the Motwani and Gupta models were applied to the Japanese cohort. White individuals accounted for approximately 80% of participants in the original development cohorts (5,12), suggesting that these models were primarily trained on non-Asian populations. Demographic differences, along with variations in healthcare delivery systems and baseline risk profiles, may have contributed to the reduced discrimination. Additionally, compared to the original studies, our cohort was older, had a smaller body size, and exhibited less variability in CDDP dosing (Table 2). This relative homogeneity in patient characteristics and risk profiles may have constrained the ability of the models to distinguish between patients who did and did not develop C-AKI (21). Another potential source of performance variation is the differences in the definitions of AKI between studies. The Motwani model used C-AKI, whereas the Gupta model targeted severe C-AKI as the primary outcome (Table 1). None of the models was originally evaluated for transportability across different outcome definitions. Predicting severe AKI is particularly relevant for guiding treatment decisions because it is associated with worse renal prognosis and overall survival (12,22,23). However, even modest elevations in creatinine levels may affect renal outcomes (24). Therefore, differences in outcome definitions must be carefully considered when interpreting the model performance across populations. The decline in calibration performance was more pronounced than the reduction in discriminatory ability. Calibration, which reflects the agreement between the predicted and observed risks, is critical for the reliable clinical application of prediction models. Recent studies have consistently reported that models developed in Western populations show greater discrepancies in calibration than discrimination when applied to Asian settings (25–27). Similar to discrimination issues, this may be attributed to demographic differences, healthcare system variations, baseline patient characteristics, and differences in the definitions of C-AKI. Poor calibration can be particularly detrimental when clinical decisions are based on predicted risk estimates (28). Underestimation may lead to missed cases of C-AKI, whereas overestimation may prompt unnecessary therapeutic interventions, such as switching from cisplatin to alternatives (e.g. carboplatin or oxaliplatin (29)), reducing chemotherapy doses, or intensifying supportive care. These interventions can increase healthcare costs and compromise treatment efficacy. The DCA (Fig. 5) showed that both models, prior to recalibration, provided either no net benefit or net harm across a wide range of clinically relevant threshold probabilities. Furthermore, risk stratification based on the original simple Gupta model showed that the observed incidence of severe C-AKI was lower in our cohort than in the original development cohort (Fig. 4). Although Gupta et al. recommended reducing the cisplatin dose or switching to an alternative treatment for patients classified as high or very high risk, the lower incidence observed in our cohort suggests that the direct application of these recommendations could lead to overtreatment in Japanese clinical practice. Altogether, these findings suggest that directly applying the original models without adjustment may lead to inappropriate decision-making in Japanese clinical practice. In the present study, we attempted model recalibration to improve the performance. Although recalibration often improves calibration, it is not expected to enhance discrimination. In our analysis, recalibration improved both the calibration and the clinical utility of the Motwani and Gupta models. This improvement may be attributed to the fact that both models retained a certain level of discriminatory ability despite being applied to different populations. However, the cutoffs for continuous variables and the weighting of individual predictors in existing models were originally optimised for development cohorts using data-driven approaches. Consequently, these thresholds and score ratios may not be optimal for Japanese clinical settings. Furthermore, suboptimal discrimination may reflect the absence of important, unmeasured predictors relevant to the local population. As our study was based on a single-centre cohort, model updating was limited to recalibration. Future efforts should aim for more comprehensive model updates, such as incorporating additional predictors, refining cutoff values, and validating across multiple centres, to enhance the applicability and accuracy of prediction models in Japanese clinical practice. This study has some limitations. First, the representativeness of the study population is limited. Owing to data availability, external validation was conducted using a single-centre cohort from a university hospital located in a rural area, where the prevalence of hypertension is higher than that in urban settings (30). In addition, our cohort may be skewed towards older patients because of regional demographic characteristics. These factors could have influenced both baseline risk and model performance, underscoring the need for cautious generalisation of our findings. Second, the proportion of missing magnesium values (61.2%) was high. Serum magnesium is a predictor in the Gupta model, and the large amount of missing data may have led to an underestimation of the model’s performance. As magnesium measurement is not technically difficult, routine assessment of serum magnesium levels is recommended when applying the Gupta model in future clinical settings. Lastly, no a priori sample size calculations were conducted. Although the commonly cited criterion of 100 events for model validation has been met (31), the recent literature suggests more precise sample size recommendations for external validation studies (32). Based on these updated criteria (the parameters used in this calculation are shown in Table S5), the sample size in this study may have been insufficient to ensure precise calibration estimates. Therefore, conclusions regarding the calibration should be interpreted with caution. Future research should aim to validate and refine these prediction models using larger and more diverse cohorts, ideally through national databases or multicentre observational studies, to enhance their generalisability and clinical utility in Japanese practice. Conclusions To the best of our knowledge, this study is the first to externally validate the Motwani and Gupta models for predicting C-AKI in a Japanese population. Although both models demonstrated modest discriminatory performance, the Gupta model showed superior ability to identify patients at risk for severe C-AKI, a clinically important outcome. However, both models demonstrated poor initial calibration, limiting their direct clinical utility. Importantly, recalibration substantially improved both calibration and clinical utility, underscoring the need for population-specific model adaptation prior to clinical application. These findings highlight the potential utility of existing models in the Japanese setting, but only after appropriate recalibration. Future research should focus on model refinement through structural updates and validation in large multicentre Japanese cohorts to support reliable, individualised risk stratification and optimise the safe use of cisplatin in clinical practice. Abbreviations AKI: Acute Kidney Injury AUROC: Area Under the Receiver Operating Characteristic curve BMI: Body Mass Index C-AKI: Cisplatin-Associated Acute Kidney Injury CDDP: Cisplatin CI: Confidence Interval CITL: Calibration-In-The-Large DCA: Decision Curve Analysis eGFR: Estimated Glomerular Filtration Rate KDIGO: Kidney Disease: Improving Global Outcomes PLT: Platelet Count ROC: Receiver Operating Characteristic RRT: Renal Replacement Therapy WBC: White Blood Cell Count Declarations Ethics approval and consent to participate The study was conducted in compliance with the Ethical Guidelines for Medical and Biological Research Involving Human Subjects issued by the Ministry of Health, Labour, and Welfare of Japan, as well as the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the Iwate Medical University (MH2024-110). Due to the retrospective nature of the study, which involved the analysis of existing clinical data, the requirement for individual informed consent was waived by the ethics committee. Consent for publication Not applicable Availability of data and materials Data supporting the findings of this study are available upon request from the corresponding author. The data are not publicly available because of privacy and ethical restrictions. Competing interests The authors declare that they have no conflict of interest. Funding No funds have been received from external agencies for this study. Authors' contributions This study was designed by KS and SN. KS performed the statistical analysis and wrote the manuscript, which was reviewed by SN; JA and KK supervised the study. Acknowledgements We would like to thank Editage (www.editage.jp) for English language editing. References Dasari S, Bernard Tchounwou P. Cisplatin in cancer therapy: molecular mechanisms of action. Eur J Pharmacol. 2014 Oct 5;740:364–78. Mcsweeney KR, Gadanec LK, Zulli A. Mechanisms of cisplatin-induced acute kidney injury. Cancers (Basel). 2021;13(1572):1–42. Trevisani F, Di Marco F, Quattrini G, Lepori N, Floris M, Valsecchi D, et al. Acute kidney injury and acute kidney disease in high-dose cisplatin-treated head and neck cancer. Front Oncol. 2023;13(June):1–10. Patimarattananan T, Nongnuch A, Pattaranutaporn P, Unwanatham N, Jiarpinitnun C, Ngamphaiboon N. Risk and impact of delayed renal impairment in patients with locally advanced head and neck squamous cell carcinoma receiving chemoradiotherapy with cisplatin. Support Care Cancer [Internet]. 2021 Feb;29(2):877–87. Available from: http://www.ncbi.nlm.nih.gov/pubmed/32524284 Gupta S, Glezerman IG, Hirsch JS, Chen KL, Devaraj N, Wells SL, et al. Derivation and external validation of a simple risk score for predicting severe acute kidney injury after intravenous cisplatin: cohort study. BMJ [Internet]. 2024 Mar 27;384:e077169. Available from: http://www.ncbi.nlm.nih.gov/pubmed/38538012 Kuo IC, Chu YC, Chen YH, Chan TC. Association between rapid renal function deterioration and cancer mortality in the elderly: a retrospective cohort study. Cancer Med. 2023;12(8):10008–19. Caddeo G, Williams ST, McIntyre CW, Selby NM. Acute kidney injury in urology patients: Incidence, causes and outcomes. Nephrourol Mon. 2013;5(5):955–61. Vinagolu-Baur J, Javaid S, Frasier K, Li V, Ortega Yaguachi N, Rodriguez E, et al. Hospital related outcomes of acute kidney injury in patients with bladder cancer: A nationwide inpatient study. J Clin Oncol [Internet]. 2024 Jun 1 [cited 2025 Mar 30];42(16_suppl):e16624. Available from: https://ascopubs.org/doi/10.1200/JCO.2024.42.16_suppl.e16624 Cheng Y, Nie S, Li L, Li Y, Liu D, Xiong M, et al. Epidemiology and outcomes of acute kidney injury in hospitalized cancer patients in China. Int J Cancer [Internet]. 2019 Jun 1 [cited 2025 Mar 30];144(11):2644–50. Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/ijc.31993 Crona DJ, Faso A, Nishijima TF, McGraw KA, Galsky MD, Milowsky MI. A systematic review of strategies to prevent cisplatin‐induced nephrotoxicity. Oncologist [Internet]. 2017 May 1 [cited 2025 Mar 30];22(5):609–19. Available from: https://dx.doi.org/10.1634/theoncologist.2016-0319 Yoshida T, Niho S, Toda M, Goto K, Yoh K, Umemura S, et al. Protective effect of magnesium preloading on cisplatin-induced nephrotoxicity: a retrospective study. Jpn J Clin Oncol. 2014;44(4):346–54. Motwani SS, McMahon GM, Humphreys BD, Partridge AH, Waikar SS, Curhan GC. Development and validation of a risk prediction model for acute kidney injury after the first course of cisplatin. J Clin Oncol. 2018;36(7):682–8. Ramspek CL, Jager KJ, Dekker FW, Zoccali C, Van DIepen M. External validation of prognostic models: What, why, how, when and where? Clin Kidney J. 2021;14(1):49–58. Riley RD, Archer L, Snell KIE, Ensor J, Dhiman P, Martin GP, et al. Evaluation of clinical prediction models (part 2): how to undertake an external validation study. BMJ [Internet]. 2024 Jan 15;384(part 2):e074820. Available from: https://www.bmj.com/lookup/doi/10.1136/bmj-2023-074820 Burns C V., Edwin SB, Szpunar S, Forman J. Cisplatin-induced nephrotoxicity in an outpatient setting. Pharmacotherapy. 2021;41(2):184–90. Collins GS, Moons KGM, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ [Internet]. 2024 Apr 16 [cited 2025 Mar 30];385:e078378. Available from: https://www.bmj.com/content/385/bmj-2023-078378 Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis [Internet]. 2009 Jun 1 [cited 2025 Apr 15];53(6):982–92. Available from: https://pubmed.ncbi.nlm.nih.gov/19339088/ Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract [Internet]. 2012 Oct [cited 2025 Mar 30];120(4). Available from: https://pubmed.ncbi.nlm.nih.gov/22890468/ Mishra A, McClelland RL, Inoue LYT, Kerr KF. Recalibration methods for improved clinical utility of risk scores. Med Decis Mak [Internet]. 2021 May 1 [cited 2025 Mar 30];42(4):500. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC8977399/ Siregar S, Nieboer D, Versteegh MIM, Steyerberg EW, Takkenberg JJM. Methods for updating a risk prediction model for cardiac surgery: A statistical primer. Interact Cardiovasc Thorac Surg. 2019;28(3):333–8. Riley RD, Ensor J, Snell KIE, Debray TPA, Altman DG, Moons KGM, et al. External validation of clinical prediction models using big datasets from e-health records or IPD meta-analysis: opportunities and challenges. BMJ [Internet]. 2016 Jun 22 [cited 2025 Mar 30];353:i3140. Available from: https://www.bmj.com/content/353/bmj.i3140 Kang E, Park M, Park PG, Park N, Jung Y, Kang U, et al. Acute kidney injury predicts all‐cause mortality in patients with cancer. Cancer Med [Internet]. 2019 Jun 1 [cited 2025 Apr 20];8(6):2740. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC6558474/ Cho S, Kang E, Kim JE, Kang U, Kang HG, Park M, et al. Clinical significance of acute kidney injury in lung cancer patients. Cancer Res Treat [Internet]. 2021 Jan 18 [cited 2025 Apr 27];53(4):1015–23. Available from: http://www.e-crt.org/journal/view.php?number=3214 Ishitsuka R, Miyazaki J, Ichioka D, Inoue T, Kageyama S, Sugimoto M, et al. Impact of acute kidney injury defined by CTCAE v4.0 during first course of cisplatin-based chemotherapy on treatment outcomes in advanced urothelial cancer patients. Clin Exp Nephrol [Internet]. 2017 Aug 1 [cited 2025 Apr 27];21(4):732–40. Available from: https://link.springer.com/article/10.1007/s10157-016-1327-z Kurniawaty J, Setianto BY, Widyastuti Y, Supomo S, Boom CE, Ancilla C. Validation for EuroSCORE II in the Indonesian cardiac surgical population: a retrospective, multicenter study. Expert Rev Cardiovasc Ther [Internet]. 2022 [cited 2025 Apr 8];20(6):491–6. Available from: https://pubmed.ncbi.nlm.nih.gov/35579398/ De Jong VMT, Rousset RZ, Antonio-Villa NE, Buenen AG, Van Calster B, Bello-Chavolla OY, et al. Clinical prediction models for mortality in patients with Covid-19: external validation and individual participant data meta-analysis. BMJ [Internet]. 2022 [cited 2025 Apr 8];378:e069881. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC9273913/ Kasim SS, Ibrahim N, Malek S, Ibrahim KS, Aziz MF, Song C, et al. Validation of the general Framingham Risk Score (FRS), SCORE2, revised PCE and WHO CVD risk scores in an Asian population. Lancet Reg Heal West Pacific [Internet]. 2023 Jun 1 [cited 2025 Apr 8];35:100742. Available from: https://www.thelancet.com/action/showFullText?pii=S2666606523000603 Van Calster B, McLernon DJ, Van Smeden M, Wynants L, Steyerberg EW, Bossuyt P, et al. Calibration: the Achilles heel of predictive analytics. BMC Med [Internet]. 2019 Dec 16 [cited 2025 Mar 30];17(1):230. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC6912996/ Lokich J. What is the “best” platinum: cisplatin, carboplatin, or oxaliplatin? Cancer Invest [Internet]. 2001 [cited 2025 Mar 30];19(7):756–60. Available from: https://www.tandfonline.com/doi/abs/10.1081/CNV-100106152 Oka M, Yamamoto M, Mure K, Takeshita T, Arita M. Relationships between lifestyle, living environments, and incidence of hypertension in Japan (in men): based on participant’s data from the nationwide medical check-up. PLoS One [Internet]. 2016 Oct 27 [cited 2025 Apr 15];11(10):e0165313. Available from: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165313 Vergouwe Y, Steyerberg EW, Eijkemans MJC, Habbema JDF. Substantial effective sample sizes were required for external validation studies of predictive logistic regression models. J Clin Epidemiol [Internet]. 2005 May 1 [cited 2025 Apr 7];58(5):475–83. Available from: https://www.jclinepi.com/action/showFullText?pii=S0895435605000193 Riley RD, Snell KIE, Archer L, Ensor J, Debray TPA, van Calster B, et al. Evaluation of clinical prediction models (part 3): calculating the sample size required for an external validation study. BMJ [Internet]. 2024 Jan 22 [cited 2025 Apr 6];384:e074821. Available from: https://www.bmj.com/content/384/bmj-2023-074821 Tables Table 1. Clinical prediction models for cisplatin related acute kidney injury Motwani et al. Gupta et al. Definition of AKI Creatinine ≥ 0.3 mg/dL in 14 days Creatinine ≥ 2.0-fold or RRT in 14 days Predictors Age (years) ≤60: 0 points ≤45: 0 points 61–70: 1.5 points 46–60: 2.5 points >70: 2.5 points 61–70: 3.5 points >70: 4.5 points Hypertension 2 points 1 point Diabetes – 1 point Smoker – 1 point CDDP dosage (mg) ≤100: 0 points ≤50: 0 points 101–150: 1 point 51–75: 2 points >150: 3 points 76–100: 2.5 points 101–125: 3 points 126–150: 5 points 151–200: 7.5 points >200: 9.5 points Hemoglobin (mg/dL) – ≥12.0: 0 points 11.0–11.9: 1 point 12.0: 1.5 points Albumin (mg/dL) >3.5: 0 points >3.8: 0 points ≤3.5: 2 points 3.3–3.8: 1 point <3.3: 1.5 points Magnesium (mg/dL) – ≥2.0: 0 points <2.0: 1 point Acute kidney injury (AKI) was defined as the degree of increase in serum creatinine levels from baseline. RRT, renal replacement therapy; CDDP, cisplatin; WBC, white blood cell count. Table 2. Patients’ background Saito et al. (n = 1,684) Gupta et al. (n = 11,766) Motwani et al. (n = 2,128) Study duration 2014–2023 2006–2022 2005–2014 Age, years 66 (58–71) 59 (50–67) 56.8 (13.2) Male 1,124 (66.7) 6,935 (58.9) 1,173 (55.4) BMI, kg/m 2 21.9 (19.5–24.4) – 26.6 (5.7) CDDP dosage, mg 110 (90–130) 90 (60–160) 118.4 (77–151) Hypertension 595 (35.3) 2,900 (24.6) 1,046 (49.4) Diabetes 203 (12.1) 1,426 (12.1) 300 (14.2) Smoker 1033 (64.8) 6,762 (57.5) – Creatinine, mg/dL 0.75 (0.63–0.87) 0.9 (0.7–1.0) 0.9 (0.2) eGFR, mL/min/1.73 m 2 74.0 (63.2–87.3) 90.0 (75.0–101.0) 87.3 (19.4) WBC, ×10 3 /mm 3 5.99 (4.64–7.56) 7.1 (5.6–9.1) – PLT, ×10 3 /mm 3 256 (203–313) 255 (204–322) – Hemoglobin, mg/dL 12.5 (11.3–13.6) 12.8 (11.4–14.0) – Albumin, g/dL 3.9 (3.5–4.2) 4.1 (3.7–4.4) 4.0 (0.5) Magnesium, mg/dL 2.05 (2.00–2.10) 2.19 (1.94–2.19) – C-AKI 186 (11.0) – 289 (13.6) Stage 1 150 (8.9) – 234 (11.0) Stage > 2 36 (2.1) 608 (5.2) 55 (2.6) Patient background characteristics and outcomes in our cohort as well as those reported in previous studies have been summarised. Continuous variables are presented as medians (interquartile ranges), and categorical variables are presented as n (%). In a study by Motwani et al., continuous variables, except for CDDP dose, are reported as means (standard deviations). The proportions of missing data in our cohort were as follows: albumin (6.0%); magnesium (61.2%); platelets (0.1%); and white blood cells (0.7%). BMI, body mass index; CDDP, cisplatin; eGFR, estimated glomerular filtration rate; WBC, white blood cell count; PLT, platelet count; C-AKI, cisplatin related acute kidney injury. Table 3. Indication parameter of model validation Motwani model After recalibration Gupta model After recalibration C-AKI AUROC 0.613 (0.570–0.656) 0.613 (0.570–0.656) 0.616 (0.575–0.658) 0.616 (0.575–0.658) CITL −0.438 (–0.604–−0.283) −0.004 (−0.161–0.147) 0.346 (0.157–0.526) −0.003 (−0.150–0.138) Slope 0.550 (0.340–0.758) 1.006 (0.621–1.374) −0.075 (−0.155–0.035) 0.995 (0.664–1.375) Brier score 0.102 (0.092–0.112) 0.096 (0.085–0.108) 0.108 (0.095–0.119) 0.096 (0.086–0.107) Severe C–AKI AUROC 0.594 (0.482–0.697) 0.594 (0.482–0.697) 0.674 (0.584–0.768) 0.674 (0.584–0.768) CITL −2.242 (−2.602–−1.921) −0.023 (−0.403–0.283) −1.567 (−1.921–−1.237) −0.018 (−0.388–0.304) Slope 0.523 (0.090–1.008) 0.993 (0.057–1.972) −0.005 (−0.178–0.469) 0.996 (0.485–1.495) Brier score 0.047 (0.042–0.052) 0.021 (0.014–0.027) 0.032 (0.027–0.038) 0.021 (0.014–0.027) The mean values and 95% confidence intervals from the bootstrap samples are reported for each performance indicator. The parameters after recalibration are shown in the right-hand column, along with the corresponding original model indicators. AUROC, area under the receiver operating characteristic curve; CITL; calibration-in-the-large; Slope: Calibration slope. Additional Declarations No competing interests reported. Supplementary Files Supplymentalmaterial.docx Cite Share Download PDF Status: Published Journal Publication published 29 Sep, 2025 Read the published version in Journal of Pharmaceutical Health Care and Sciences → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6678420","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":463311742,"identity":"a5501bf3-00d8-4bba-9d30-d350bb8f6256","order_by":0,"name":"Kazuki Saito","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABN0lEQVRIie2RsWrDMBBATwicReBVwkn6CyqG0oCpf8XG4C4Z+gmanMWQ1Z+RLqKjjaFZAl7TzcbgqUO6pdDSSklNS+3QjqX4wUknjsedJICBgT8IVYEEVytWsSNA2goSbf20ggElBwX/RjmkGDD5SD6VHthi3bDkBlxzYWa1M4YxXxdZiebORIzyEmZ3HcUiXshWHPwkx9ieq8H4JsAcydAWJOTANh1lCt49Kzl4kGPD0spFGhjWq8x9AXM1RdRVzCrSinumlUutFPVoj+SbL8zHXsWigaEHQyutgFa2gQFIpr6g/V1Y0uBZwql/m2P7PCaUuNvapkgGdkQbnvbchRbX1UP84rjTIqvK59hx2dKvdkheTZZmUNWs+2KtedxQ/OXrDBU5E6eUlv23M3r6URkYGBj4/7wDAN5dg2FeHzEAAAAASUVORK5CYII=","orcid":"","institution":"Iwate Medical University Hospital","correspondingAuthor":true,"prefix":"","firstName":"Kazuki","middleName":"","lastName":"Saito","suffix":""},{"id":463311743,"identity":"465ada6b-9933-4660-adfa-fb308d1c4c92","order_by":1,"name":"Satoru Nihei","email":"","orcid":"","institution":"Iwate Medical University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Satoru","middleName":"","lastName":"Nihei","suffix":""},{"id":463311744,"identity":"392a690d-8a52-43c3-bbbc-5a02a40c90bf","order_by":2,"name":"Junichi Asaka","email":"","orcid":"","institution":"Iwate Medical University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Junichi","middleName":"","lastName":"Asaka","suffix":""},{"id":463311745,"identity":"b98d4523-c053-436b-94db-216b98440932","order_by":3,"name":"Kenzo Kudo","email":"","orcid":"","institution":"Iwate Medical University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Kenzo","middleName":"","lastName":"Kudo","suffix":""}],"badges":[],"createdAt":"2025-05-16 07:53:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6678420/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6678420/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s40780-025-00471-0","type":"published","date":"2025-09-29T15:57:03+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":83752128,"identity":"d1a014f2-b662-416a-add1-226bfeb24651","added_by":"auto","created_at":"2025-06-02 07:11:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":23717,"visible":true,"origin":"","legend":"\u003cp\u003eFlow diagram of patient selection.\u003c/p\u003e\n\u003cp\u003eFrom an initial cohort of 2,184 patients who received cisplatin (CDDP), 493 were excluded based on age, prior CDDP use, or weekly or daily regimens. After further exclusion due to missing baseline creatinine or outcome data, 1,684 patients were included in the final analysis.\u003c/p\u003e","description":"","filename":"Slide1.png","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/c0114e2d51e0920fd8655adb.png"},{"id":83752129,"identity":"1052eb66-f29b-463b-8a7c-e242c9fd1e62","added_by":"auto","created_at":"2025-06-02 07:11:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":42237,"visible":true,"origin":"","legend":"\u003cp\u003eReceiver operating characteristic (ROC) curves for the prediction of cisplatin-associated acute kidney injury (C-AKI).\u003c/p\u003e\n\u003cp\u003eThe red and blue lines represent the ROC curves of the Gupta et al. and Motwani et al. models, respectively. Panels A and B represent patients with C-AKI and severe C-AKI, respectively.\u003c/p\u003e","description":"","filename":"Slide2.png","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/03361da5526e28a405392b94.png"},{"id":83752322,"identity":"61de7e44-876c-429b-88a9-e5eac6a5ac7f","added_by":"auto","created_at":"2025-06-02 07:20:00","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":25279,"visible":true,"origin":"","legend":"\u003cp\u003eCalibration curves of the prediction models.\u003c/p\u003e\n\u003cp\u003eThe horizontal axis represents the predicted probability, and the vertical axis shows the observed incidence of cisplatin-associated acute kidney injury (C-AKI). Panels A and B correspond to the Motwani and Gupta models, respectively. The cohort was divided into ten risk-based groups, and the observed incidence for each group was plotted as triangles. The curves represent the smoothed regression lines. The red diagonal lines indicate perfect agreement between the predicted and observed probabilities. The calibration curves for severe C-AKI are shown in Fig. S2.\u003c/p\u003e","description":"","filename":"Slide3.png","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/789a50bdc1f424ac9e3cd76b.png"},{"id":83752132,"identity":"c498b678-db06-4111-811f-801bfee68bee","added_by":"auto","created_at":"2025-06-02 07:12:00","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":42153,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of risk stratification profiles between the original Gupta model and our cohort\u003c/p\u003e\n\u003cp\u003eComparison of the distribution of patients and incidence of severe cisplatin-associated acute kidney injury (C-AKI) across risk groups defined by the Gupta simple model score. The blue bars and lines represent our cohort, whereas the red bars and lines represent the original developmental cohort. Patients were categorised into four risk groups (low [score 0–5.5], moderate [score 6–9.5], high [score 10–15.5], and very high [score ≥16]). The right axis shows the proportion of patients in each group, and the left axis shows the observed incidence of severe C-AKI. The numbers in the table indicate the number of patients with severe C-AKI compared to the total number of patients in each risk group.\u003c/p\u003e","description":"","filename":"Slide4.png","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/6932d19ba3c365a28269e4b3.png"},{"id":83752134,"identity":"cdf8d099-956d-4560-850d-2ab916cbcaca","added_by":"auto","created_at":"2025-06-02 07:12:00","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":50344,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analysis of the prediction models.\u003c/p\u003e\n\u003cp\u003eThe net benefits for the prediction models developed by Motwani et al. and Gupta et al. are plotted across a range of threshold probabilities. Panel A displays results for patients with cisplatin-associated acute kidney injury (C-AKI), and Panel B displays results for patients with severe C-AKI. In both panels, dashed lines represent the models before recalibration, while solid lines represent the recalibrated versions. The threshold probability represents the point at which the benefits of interventions to prevent C-AKI outweigh the risks of those interventions. In simpler terms, a lower threshold probability (towards the left of the graph) indicates a greater emphasis on preventing nephrotoxicity compared to the potential benefits of treatment. Once a clinically relevant threshold probability is established, a superior model will demonstrate higher net benefits (i.e., a higher position on the graph).\u003c/p\u003e","description":"","filename":"Slide5.png","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/75e2960481e9e5259cfbc73b.png"},{"id":92884455,"identity":"3104c659-a72f-42bb-aabf-01e10a24dbea","added_by":"auto","created_at":"2025-10-06 16:13:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":887796,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/e27bc27f-8136-4806-92e2-c33ca9c098aa.pdf"},{"id":83752131,"identity":"aef64577-2a1f-449d-a3d1-f9947e3f8d31","added_by":"auto","created_at":"2025-06-02 07:11:59","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":41645,"visible":true,"origin":"","legend":"","description":"","filename":"Supplymentalmaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-6678420/v1/e490c72e3d18c3e82a295226.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"External validation and comparison of clinical prediction models for cisplatin- associated acute kidney injury: A single-centre retrospective study","fulltext":[{"header":"Background","content":"\u003cp\u003eCisplatin (CDDP) is an important chemotherapeutic agent with proven efficacy against various solid tumours (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). One of its major dose-limiting toxicities is nephrotoxicity (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), with cisplatin-associated acute kidney injury (C-AKI) occurring in 20\u0026ndash;30% of patients and in approximately 10% even after the first dose (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). C-AKI is associated with poor renal outcomes (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e), treatment interruptions, poor prognosis (\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e), prolonged hospital stays, and increased healthcare costs (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). Preventive strategies, such as hydration, forced diuresis with mannitol or furosemide (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e), and magnesium supplementation (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e), are commonly employed to mitigate CDDP-induced nephrotoxicity. However, these measures did not completely eliminate the risk of C-AKI.\u003c/p\u003e \u003cp\u003eGiven the frequency and potential severity of C-AKI, assessing the associated risks in advance is important. Clinical prediction models, tools that estimate the probability of a specific outcome using multiple predictors, are widely used in medical practice. In 2018, Motwani et al. developed a prediction model for C-AKI using data from the US population (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). This scoring system estimates C-AKI risk based on readily available clinical variables, such as age, CDDP dose, serum albumin level, and history of hypertension. In 2024, Gupta et al. proposed a new model that incorporated additional predictors, including blood cell counts, haemoglobin levels, and serum magnesium concentration (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). In their development and validation cohorts, the Gupta model demonstrated superior discriminatory performance compared to the earlier model developed by Motwani et al.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eClinical prediction models for cisplatin related acute kidney injury\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMotwani et al.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGupta et al.\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDefinition of AKI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCreatinine\u0026thinsp;\u0026ge;\u0026thinsp;0.3 mg/dL\u003c/p\u003e \u003cp\u003ein 14 days\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCreatinine\u0026thinsp;\u0026ge;\u0026thinsp;2.0-fold\u003c/p\u003e \u003cp\u003eor RRT in 14 days\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePredictors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;60: 0 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;45: 0 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e61\u0026ndash;70: 1.5 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e46\u0026ndash;60: 2.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;70: 2.5 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e61\u0026ndash;70: 3.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;70: 4.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHypertension\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 point\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDiabetes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ndash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 point\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmoker\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ndash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 point\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCDDP dosage (mg)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;100: 0 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;50: 0 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e101\u0026ndash;150: 1 point\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e51\u0026ndash;75: 2 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;150: 3 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e76\u0026ndash;100: 2.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e101\u0026ndash;125: 3 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e126\u0026ndash;150: 5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e151\u0026ndash;200: 7.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;200: 9.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHemoglobin (mg/dL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ndash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;12.0: 0 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11.0\u0026ndash;11.9: 1 point\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;11.0: 1.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWBC (\u0026times;10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ndash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;12.0: 0 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;12.0: 1.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlbumin (mg/dL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;3.5: 0 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;3.8: 0 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;3.5: 2 points\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.3\u0026ndash;3.8: 1 point\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;3.3: 1.5 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMagnesium (mg/dL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026ndash;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;2.0: 0 points\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;2.0: 1 point\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"3\"\u003eAcute kidney injury (AKI) was defined as the degree of increase in serum creatinine levels from baseline. RRT, renal replacement therapy; CDDP, cisplatin; WBC, white blood cell count.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe ultimate goal of such clinical prediction models is to inform clinical decision-making in real-world practice. To enable this, external validation is essential to confirm that a model performs reliably in populations or settings different from those in which it was originally developed, whether geographically or temporally (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). However, to date, neither of these C-AKI prediction models have been evaluated outside Western populations (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e), including the Japanese population. The Motwani and Gupta models were developed using different outcome definitions. Severe C-AKI, the primary outcome targeted by the Gupta model, represents a more clinically significant event associated with a worse prognosis than mild C-AKI (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). Therefore, directly comparing various outcome thresholds may offer valuable insights into their suitability for clinical applications. This study aimed to evaluate the external validity of the Motwani and Gupta models in a single-centre Japanese cohort and compare their predictive performance.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThis study followed the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis + artificial intelligence guidelines (16). The complete checklist is presented in Table S1. The patients and members of the public were not involved in the design or conduct of this study. As this was a retrospective analysis of existing clinical data without any prospective intervention, no formal study protocol was prepared, and the study was not registered in a public trial registry. All statistical analyses were performed using the R version 4.3.1. OpenAI\u0026apos;s GPT-4o language model was utilized to assist in the development and refinement of the R code used for these analyses. The code used for the data analysis is provided in Table S2. Statistical significance was defined as p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData source\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study used data from patients who received CDDP at Iwate Medical University Hospital between April 2014 and December 2023. The exclusion criteria were as follows: (i) age \u0026lt; 18 years at the time of administration, (ii) CDDP outside the study period or at another institution, (iii) treatment with daily or weekly CDDP regimens, and (iv) missing baseline renal function or outcome data. All eligible cases were included, and no a priori sample size calculations were performed.\u003c/p\u003e\n\u003cp\u003eData on patient characteristics (age, sex, height, and weight), smoking history, concomitant medications, dates and doses of CDDP administration, and relevant laboratory values (creatinine, albumin, white blood cells, platelets, haemoglobin, and magnesium) were extracted from the electronic medical records. The estimated renal glomerular filtration rate was calculated using the Japanese Society of Nephrology (17). Diabetes and hypertension were defined based on the use of the corresponding medications. The baseline laboratory values were defined as the most recent measurements obtained within 30 days of CDDP administration. Missing values were supplemented using regression-based imputation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDefinition of outcomes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, C-AKI was defined as an increase in serum creatinine levels within 14 days of CDDP exposure. Based on the Kidney Disease: Improving Global Outcomes (KDIGO) criteria (18), C-AKI was defined as either an increase of\u0026nbsp;\u0026ge;0.3 mg/dL or a\u0026nbsp;\u0026ge;1.5-fold rise from baseline. Although an increase of\u0026nbsp;\u0026ge;0.3 mg/dL over 14 days is a slight deviation from the strict KDIGO definition, this criterion was included to align with the outcome definition used in the Motwani model\u0026nbsp;(12). Severe C-AKI was defined as a\u0026nbsp;\u0026ge;2.0-fold increase from baseline or the initiation of renal replacement therapy (KDIGO stage\u0026nbsp;\u0026ge;\u0026nbsp;2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCalculation of scores and predictive probabilities\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe scores were calculated according to the scoring criteria listed in Table 1. For the Motwani model, predictive probabilities were derived using the baseline risk and odds ratio reported in their development cohort\u0026mdash;specifically, an incidence rate of 0.04 at a score of 0, and an odds ratio of 1.49 per one-point increase in score. As the simple model by Gupta et al. does not provide predictive probabilities for individual scores, the predictive probabilities were estimated using the primary model, which incorporates a cubic spline. The codes used in the calculations are listed in Table S2.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical performance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe statistical performance of the clinical prediction models was evaluated from three key perspectives: discrimination, calibration, and overall model fit (14). For each metric, the mean and 95% confidence intervals (CI) were estimated using 1,000 bootstrap samples.\u003c/p\u003e\n\u003cp\u003eDiscriminatory ability, the capacity of the model to distinguish between individuals who do and do not experience an event, was assessed using the area under the receiver operating characteristic curve (AUROC). Differences between the two AUROCs were tested using the bootstrap method.\u003c/p\u003e\n\u003cp\u003eCalibration, which reflects the agreement between the predicted and observed risks, was evaluated using the following metric: calibration-in-the-large, which indicates systematic over- or underestimation across the model. The ideal value is 0; values \u0026lt; 0 suggest an overestimation, whereas values \u0026gt; 0 suggest an underestimation. The calibration slope represents the steepness of the calibration curve. A slope of 1 indicates perfect calibration, and values \u0026lt; 1 indicate extreme predictions in certain probability ranges.\u003c/p\u003e\n\u003cp\u003eThe model fit was assessed using the Brier score, which quantifies the mean squared difference between the predicted probabilities and observed outcomes. A lower Brier score indicates a better model performance, with 0 being ideal.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRisk stratification with the Gupta simple model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the original study by Gupta et al., a simple model score was evaluated to stratify patients into risk groups, with accompanying expert recommendations on clinical management. By contrast, the Motwani et al. model does not propose a risk stratification system or clinical action guidelines. Therefore, we conducted an additional risk stratification analysis based on the Gupta model to explore how the risk grouping and associated clinical recommendations proposed in the original study were applied to our cohort. Patients were categorised into four groups according to their total score following the definitions from the original study as follows: low (0\u0026ndash;5.5 points), moderate (6\u0026ndash;9.5 points), high (10\u0026ndash;15.5 points), and very high (\u0026ge;16 points). The incidence of severe C-AKI was calculated for each group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRecalibration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRecalibration is a model-updating technique that adjusts regression coefficients to reflect the target population better (19). Logistic recalibration was performed for each scoring system (20). Specifically, the intercept and slope were re-estimated using logistic regression with the outcome as the dependent variable and each prediction score as the independent variable. Recalibration did not enhance the discriminative ability because the model structure remained unchanged. Although structural updates could potentially lead to substantial improvements in model performance, model updating was limited to recalibration given that this study was based on a single-centre cohort.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical utility\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDecision curve analysis (DCA) was performed to evaluate the clinical utility of each model. In DCA, the net benefit, defined as the benefit of correct classification minus the harm of misclassification, is plotted on the vertical axis, whereas the threshold probability, the probability at which the benefits of treatment and non-treatment are considered equal, is shown on the horizontal axis. At any given threshold probability, a higher net benefit indicated greater clinical usefulness. Conversely, a negative net benefit suggests that model-guided decision-making may cause more harm than benefits.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eA total of 2,184 patients who received CDDP during the study period were identified. After applying the exclusion criteria, 1,684 patients were included in the final analysis (Fig. 1). Patient characteristics and outcome frequencies are summarised in Table 2. The primary outcome, C-AKI, was observed in 186 (11.0%) patients, and severe C-AKI occurred in 36 (2.1%) patients.\u003c/p\u003e\n\u003cp\u003eThe statistical performances of the clinical prediction models are summarised in Table 3, with a full comparison of all metrics provided in Table S3. Receiver operating characteristic (ROC) curves and calibration plots are presented in Figs. 2 and 3. The discriminatory performances of the Gupta and Motwani models for C-AKI showed no significant difference (AUROC, 0.616 [95% CI, 0.575\u0026ndash;0.658] vs. 0.613 [95% CI, 0.570\u0026ndash;0.656]; p = 0.84). By contrast, the Gupta model demonstrated significantly better discrimination of severe C-AKI (AUROC, 0.674 [95% CI, 0.584\u0026ndash;0.768] vs. 0.594 [95% CI, 0.482\u0026ndash;0.697]; p = 0.02). The calibration slopes and other indices revealed that both models were poorly calibrated.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn addition, risk stratification based on the Gupta simple model score is shown in Fig. 4. This risk stratification analysis was exploratory and intended to illustrate the distribution of risk categories in our cohort rather than serving as a primary evaluation of model performance. Although the model demonstrated reasonable discrimination across risk groups, the overall risk distribution was lower than that of the original development cohort, with most patients categorised into moderate- or high-risk groups.\u003c/p\u003e\n\u003cp\u003eLogistic recalibration was performed to adjust both models to the characteristics of our cohort; the updated coefficients are presented in Table S4. The clinical utility was evaluated using DCA for both the original and recalibrated models (Fig. 5). For C-AKI, the non-recalibrated Gupta model (dashed red line) demonstrated lower net benefit than the model-independent strategy (black line) across most threshold probabilities, whereas the non-recalibrated Motwani model (dashed blue line) yielded net harm at higher thresholds (approximately \u0026ge;12.5%). By contrast, both recalibrated models exhibited positive net benefits across a wide range of thresholds with comparable clinical utility. A similar pattern was observed for severe C-AKI; both uncalibrated models showed net harm, which improved after recalibration. The recalibrated Gupta model demonstrated the highest net benefit for predicting severe C-AKI.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we conducted the first external validation of two existing clinical prediction models for C-AKI developed by Motwani et al. and Gupta et al. in a Japanese patient population. The discriminative ability of the two models for C-AKI was comparable. However, the Gupta model demonstrated superior performance in predicting severe C-AKI, which was clinically significant. Both models initially exhibited poor calibration, and the DCA further indicated that using unadjusted models for clinical decision making did not produce a net benefit. After logistic recalibration, both models showed improved clinical utility, with the recalibrated Gupta model offering the greatest net benefit to patients with severe C-AKI. These results highlight the importance of recalibrating the prediction models before their application to new populations, particularly in different geographic and clinical contexts.\u003c/p\u003e\n\u003cp\u003eClinical prediction models are typically optimised using development cohorts and perform poorly when applied to different populations. Discriminatory performance declined when the Motwani and Gupta models were applied to the Japanese cohort. White individuals accounted for approximately 80% of participants in the original development cohorts (5,12), suggesting that these models were primarily trained on non-Asian populations. Demographic differences, along with variations in healthcare delivery systems and baseline risk profiles, may have contributed to the reduced discrimination. Additionally, compared to the original studies, our cohort was older, had a smaller body size, and exhibited less variability in CDDP dosing (Table 2). This relative homogeneity in patient characteristics and risk profiles may have constrained the ability of the models to distinguish between patients who did and did not develop C-AKI\u0026nbsp;(21).\u003c/p\u003e\n\u003cp\u003eAnother potential source of performance variation is the differences in the definitions of AKI between studies. The Motwani model used C-AKI, whereas the Gupta model targeted severe C-AKI as the primary outcome (Table 1). None of the models was originally evaluated for transportability across different outcome definitions. Predicting severe AKI is particularly relevant for guiding treatment decisions because it is associated with worse renal prognosis and overall survival (12,22,23). However, even modest elevations in creatinine levels may affect renal outcomes (24). Therefore, differences in outcome definitions must be carefully considered when interpreting the model performance across populations.\u003c/p\u003e\n\u003cp\u003eThe decline in calibration performance was more pronounced than the reduction in discriminatory ability.\u0026nbsp;Calibration, which reflects the agreement between the predicted and observed risks, is critical for the reliable clinical application of prediction models. Recent studies have consistently reported that models developed in Western populations show greater discrepancies in calibration than discrimination when applied to Asian settings (25\u0026ndash;27). Similar to discrimination issues, this may be attributed to demographic differences, healthcare system variations, baseline patient characteristics, and differences in the definitions of C-AKI.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePoor calibration can be particularly detrimental when clinical decisions are based on predicted risk estimates\u0026nbsp;(28). Underestimation may lead to missed cases of C-AKI, whereas overestimation may prompt unnecessary therapeutic interventions, such as switching from cisplatin to alternatives (e.g. carboplatin or oxaliplatin\u0026nbsp;(29)), reducing chemotherapy doses, or intensifying supportive care. These interventions can increase healthcare costs and compromise treatment efficacy. The DCA (Fig. 5) showed that both models, prior to recalibration, provided either no net benefit or net harm across a wide range of clinically relevant threshold probabilities. Furthermore, risk stratification based on the original simple Gupta model showed that the observed incidence of severe C-AKI was lower in our cohort than in the original development cohort (Fig. 4). Although Gupta et al. recommended reducing the cisplatin dose or switching to an alternative treatment for patients classified as high or very high risk, the lower incidence observed in our cohort suggests that the direct application of these recommendations could lead to overtreatment in Japanese clinical practice. Altogether, these findings suggest that directly applying the original models without adjustment may lead to inappropriate decision-making in Japanese clinical practice.\u003c/p\u003e\n\u003cp\u003eIn the present study, we attempted model recalibration to improve the performance. Although recalibration often improves calibration, it is not expected to enhance discrimination. In our analysis, recalibration improved both the calibration and the clinical utility of the Motwani and Gupta models. This improvement may be attributed to the fact that both models retained a certain level of discriminatory ability despite being applied to different populations. However, the cutoffs for continuous variables and the weighting of individual predictors in existing models were originally optimised for development cohorts using data-driven approaches. Consequently, these thresholds and score ratios may not be optimal for Japanese clinical settings. Furthermore, suboptimal discrimination may reflect the absence of important, unmeasured predictors relevant to the local population. As our study was based on a single-centre cohort, model updating was limited to recalibration. Future efforts should aim for more comprehensive model updates, such as incorporating additional predictors, refining cutoff values, and validating across multiple centres, to enhance the applicability and accuracy of prediction models in Japanese clinical practice.\u003c/p\u003e\n\u003cp\u003eThis study has some limitations. First, the representativeness of the study population is limited. Owing to data availability, external validation was conducted using a single-centre cohort from a university hospital located in a rural area, where the prevalence of hypertension is higher than that in urban settings (30). In addition, our cohort may be skewed towards older patients because of regional demographic characteristics. These factors could have influenced both baseline risk and model performance, underscoring the need for cautious generalisation of our findings. Second, the proportion of missing magnesium values (61.2%) was high. Serum magnesium is a predictor in the Gupta model, and the large amount of missing data may have led to an underestimation of the model\u0026rsquo;s performance. As magnesium measurement is not technically difficult, routine assessment of serum magnesium levels is recommended when applying the Gupta model in future clinical settings. Lastly, no a priori sample size calculations were conducted. Although the commonly cited criterion of 100 events for model validation has been met (31), the recent literature suggests more precise sample size recommendations for external validation studies (32). Based on these updated criteria (the parameters used in this calculation are shown in Table S5), the sample size in this study may have been insufficient to ensure precise calibration estimates. Therefore, conclusions regarding the calibration should be interpreted with caution. Future research should aim to validate and refine these prediction models using larger and more diverse cohorts, ideally through national databases or multicentre observational studies, to enhance their generalisability and clinical utility in Japanese practice.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eTo the best of our knowledge, this study is the first to externally validate the Motwani and Gupta models for predicting C-AKI in a Japanese population. Although both models demonstrated modest discriminatory performance, the Gupta model showed superior ability to identify patients at risk for severe C-AKI, a clinically important outcome. However, both models demonstrated poor initial calibration, limiting their direct clinical utility. Importantly, recalibration substantially improved both calibration and clinical utility, underscoring the need for population-specific model adaptation prior to clinical application. These findings highlight the potential utility of existing models in the Japanese setting, but only after appropriate recalibration. Future research should focus on model refinement through structural updates and validation in large multicentre Japanese cohorts to support reliable, individualised risk stratification and optimise the safe use of cisplatin in clinical practice.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cstrong\u003eAKI:\u003c/strong\u003e Acute Kidney Injury\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAUROC:\u003c/strong\u003e Area Under the Receiver Operating Characteristic curve\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBMI:\u003c/strong\u003e Body Mass Index\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eC-AKI:\u003c/strong\u003e Cisplatin-Associated Acute Kidney Injury\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCDDP:\u003c/strong\u003e Cisplatin\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCI:\u003c/strong\u003e Confidence Interval\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCITL:\u003c/strong\u003e Calibration-In-The-Large\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDCA:\u003c/strong\u003e Decision Curve Analysis\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eeGFR:\u003c/strong\u003e Estimated Glomerular Filtration Rate\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eKDIGO:\u003c/strong\u003e Kidney Disease: Improving Global Outcomes\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePLT:\u003c/strong\u003e Platelet Count\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eROC:\u003c/strong\u003e Receiver Operating Characteristic\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRRT:\u003c/strong\u003e Renal Replacement Therapy\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWBC:\u003c/strong\u003e White Blood Cell Count\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was conducted in compliance with the Ethical Guidelines for Medical and Biological Research Involving Human Subjects issued by the Ministry of Health, Labour, and Welfare of Japan, as well as the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the Iwate Medical University (MH2024-110). Due to the retrospective nature of the study, which involved the analysis of existing clinical data, the requirement for individual informed consent was waived by the ethics committee.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData supporting the findings of this study are available upon request from the corresponding author. The data are not publicly available because of privacy and ethical restrictions. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo funds have been received from external agencies for this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was designed by KS and SN. KS performed the statistical analysis and wrote the manuscript, which was reviewed by SN; JA and KK supervised the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank Editage (www.editage.jp) for English language editing.\u003c/p\u003e\n\n"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eDasari S, Bernard Tchounwou P. Cisplatin in cancer therapy: molecular mechanisms of action. Eur J Pharmacol. 2014 Oct 5;740:364\u0026ndash;78. \u003c/li\u003e\n\u003cli\u003eMcsweeney KR, Gadanec LK, Zulli A. Mechanisms of cisplatin-induced acute kidney injury. Cancers (Basel). 2021;13(1572):1\u0026ndash;42. \u003c/li\u003e\n\u003cli\u003eTrevisani F, Di Marco F, Quattrini G, Lepori N, Floris M, Valsecchi D, et al. Acute kidney injury and acute kidney disease in high-dose cisplatin-treated head and neck cancer. Front Oncol. 2023;13(June):1\u0026ndash;10. \u003c/li\u003e\n\u003cli\u003ePatimarattananan T, Nongnuch A, Pattaranutaporn P, Unwanatham N, Jiarpinitnun C, Ngamphaiboon N. Risk and impact of delayed renal impairment in patients with locally advanced head and neck squamous cell carcinoma receiving chemoradiotherapy with cisplatin. Support Care Cancer [Internet]. 2021 Feb;29(2):877\u0026ndash;87. Available from: http://www.ncbi.nlm.nih.gov/pubmed/32524284\u003c/li\u003e\n\u003cli\u003eGupta S, Glezerman IG, Hirsch JS, Chen KL, Devaraj N, Wells SL, et al. Derivation and external validation of a simple risk score for predicting severe acute kidney injury after intravenous cisplatin: cohort study. BMJ [Internet]. 2024 Mar 27;384:e077169. Available from: http://www.ncbi.nlm.nih.gov/pubmed/38538012\u003c/li\u003e\n\u003cli\u003eKuo IC, Chu YC, Chen YH, Chan TC. Association between rapid renal function deterioration and cancer mortality in the elderly: a retrospective cohort study. Cancer Med. 2023;12(8):10008\u0026ndash;19. \u003c/li\u003e\n\u003cli\u003eCaddeo G, Williams ST, McIntyre CW, Selby NM. Acute kidney injury in urology patients: Incidence, causes and outcomes. Nephrourol Mon. 2013;5(5):955\u0026ndash;61. \u003c/li\u003e\n\u003cli\u003eVinagolu-Baur J, Javaid S, Frasier K, Li V, Ortega Yaguachi N, Rodriguez E, et al. Hospital related outcomes of acute kidney injury in patients with bladder cancer: A nationwide inpatient study. J Clin Oncol [Internet]. 2024 Jun 1 [cited 2025 Mar 30];42(16_suppl):e16624. Available from: https://ascopubs.org/doi/10.1200/JCO.2024.42.16_suppl.e16624\u003c/li\u003e\n\u003cli\u003eCheng Y, Nie S, Li L, Li Y, Liu D, Xiong M, et al. Epidemiology and outcomes of acute kidney injury in hospitalized cancer patients in China. Int J Cancer [Internet]. 2019 Jun 1 [cited 2025 Mar 30];144(11):2644\u0026ndash;50. Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/ijc.31993\u003c/li\u003e\n\u003cli\u003eCrona DJ, Faso A, Nishijima TF, McGraw KA, Galsky MD, Milowsky MI. A systematic review of strategies to prevent cisplatin‐induced nephrotoxicity. Oncologist [Internet]. 2017 May 1 [cited 2025 Mar 30];22(5):609\u0026ndash;19. Available from: https://dx.doi.org/10.1634/theoncologist.2016-0319\u003c/li\u003e\n\u003cli\u003eYoshida T, Niho S, Toda M, Goto K, Yoh K, Umemura S, et al. Protective effect of magnesium preloading on cisplatin-induced nephrotoxicity: a retrospective study. Jpn J Clin Oncol. 2014;44(4):346\u0026ndash;54. \u003c/li\u003e\n\u003cli\u003eMotwani SS, McMahon GM, Humphreys BD, Partridge AH, Waikar SS, Curhan GC. Development and validation of a risk prediction model for acute kidney injury after the first course of cisplatin. J Clin Oncol. 2018;36(7):682\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eRamspek CL, Jager KJ, Dekker FW, Zoccali C, Van DIepen M. External validation of prognostic models: What, why, how, when and where? Clin Kidney J. 2021;14(1):49\u0026ndash;58. \u003c/li\u003e\n\u003cli\u003eRiley RD, Archer L, Snell KIE, Ensor J, Dhiman P, Martin GP, et al. Evaluation of clinical prediction models (part 2): how to undertake an external validation study. BMJ [Internet]. 2024 Jan 15;384(part 2):e074820. Available from: https://www.bmj.com/lookup/doi/10.1136/bmj-2023-074820\u003c/li\u003e\n\u003cli\u003eBurns C V., Edwin SB, Szpunar S, Forman J. Cisplatin-induced nephrotoxicity in an outpatient setting. Pharmacotherapy. 2021;41(2):184\u0026ndash;90. \u003c/li\u003e\n\u003cli\u003eCollins GS, Moons KGM, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ [Internet]. 2024 Apr 16 [cited 2025 Mar 30];385:e078378. Available from: https://www.bmj.com/content/385/bmj-2023-078378\u003c/li\u003e\n\u003cli\u003eMatsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis [Internet]. 2009 Jun 1 [cited 2025 Apr 15];53(6):982\u0026ndash;92. Available from: https://pubmed.ncbi.nlm.nih.gov/19339088/\u003c/li\u003e\n\u003cli\u003eKhwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract [Internet]. 2012 Oct [cited 2025 Mar 30];120(4). Available from: https://pubmed.ncbi.nlm.nih.gov/22890468/\u003c/li\u003e\n\u003cli\u003eMishra A, McClelland RL, Inoue LYT, Kerr KF. Recalibration methods for improved clinical utility of risk scores. Med Decis Mak [Internet]. 2021 May 1 [cited 2025 Mar 30];42(4):500. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC8977399/\u003c/li\u003e\n\u003cli\u003eSiregar S, Nieboer D, Versteegh MIM, Steyerberg EW, Takkenberg JJM. Methods for updating a risk prediction model for cardiac surgery: A statistical primer. Interact Cardiovasc Thorac Surg. 2019;28(3):333\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eRiley RD, Ensor J, Snell KIE, Debray TPA, Altman DG, Moons KGM, et al. External validation of clinical prediction models using big datasets from e-health records or IPD meta-analysis: opportunities and challenges. BMJ [Internet]. 2016 Jun 22 [cited 2025 Mar 30];353:i3140. Available from: https://www.bmj.com/content/353/bmj.i3140\u003c/li\u003e\n\u003cli\u003eKang E, Park M, Park PG, Park N, Jung Y, Kang U, et al. Acute kidney injury predicts all‐cause mortality in patients with cancer. Cancer Med [Internet]. 2019 Jun 1 [cited 2025 Apr 20];8(6):2740. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC6558474/\u003c/li\u003e\n\u003cli\u003eCho S, Kang E, Kim JE, Kang U, Kang HG, Park M, et al. Clinical significance of acute kidney injury in lung cancer patients. Cancer Res Treat [Internet]. 2021 Jan 18 [cited 2025 Apr 27];53(4):1015\u0026ndash;23. Available from: http://www.e-crt.org/journal/view.php?number=3214\u003c/li\u003e\n\u003cli\u003eIshitsuka R, Miyazaki J, Ichioka D, Inoue T, Kageyama S, Sugimoto M, et al. Impact of acute kidney injury defined by CTCAE v4.0 during first course of cisplatin-based chemotherapy on treatment outcomes in advanced urothelial cancer patients. Clin Exp Nephrol [Internet]. 2017 Aug 1 [cited 2025 Apr 27];21(4):732\u0026ndash;40. Available from: https://link.springer.com/article/10.1007/s10157-016-1327-z\u003c/li\u003e\n\u003cli\u003eKurniawaty J, Setianto BY, Widyastuti Y, Supomo S, Boom CE, Ancilla C. Validation for EuroSCORE II in the Indonesian cardiac surgical population: a retrospective, multicenter study. Expert Rev Cardiovasc Ther [Internet]. 2022 [cited 2025 Apr 8];20(6):491\u0026ndash;6. Available from: https://pubmed.ncbi.nlm.nih.gov/35579398/\u003c/li\u003e\n\u003cli\u003eDe Jong VMT, Rousset RZ, Antonio-Villa NE, Buenen AG, Van Calster B, Bello-Chavolla OY, et al. Clinical prediction models for mortality in patients with Covid-19: external validation and individual participant data meta-analysis. BMJ [Internet]. 2022 [cited 2025 Apr 8];378:e069881. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC9273913/\u003c/li\u003e\n\u003cli\u003eKasim SS, Ibrahim N, Malek S, Ibrahim KS, Aziz MF, Song C, et al. Validation of the general Framingham Risk Score (FRS), SCORE2, revised PCE and WHO CVD risk scores in an Asian population. Lancet Reg Heal West Pacific [Internet]. 2023 Jun 1 [cited 2025 Apr 8];35:100742. Available from: https://www.thelancet.com/action/showFullText?pii=S2666606523000603\u003c/li\u003e\n\u003cli\u003eVan Calster B, McLernon DJ, Van Smeden M, Wynants L, Steyerberg EW, Bossuyt P, et al. Calibration: the Achilles heel of predictive analytics. BMC Med [Internet]. 2019 Dec 16 [cited 2025 Mar 30];17(1):230. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC6912996/\u003c/li\u003e\n\u003cli\u003eLokich J. What is the \u0026ldquo;best\u0026rdquo; platinum: cisplatin, carboplatin, or oxaliplatin? Cancer Invest [Internet]. 2001 [cited 2025 Mar 30];19(7):756\u0026ndash;60. Available from: https://www.tandfonline.com/doi/abs/10.1081/CNV-100106152\u003c/li\u003e\n\u003cli\u003eOka M, Yamamoto M, Mure K, Takeshita T, Arita M. Relationships between lifestyle, living environments, and incidence of hypertension in Japan (in men): based on participant\u0026rsquo;s data from the nationwide medical check-up. PLoS One [Internet]. 2016 Oct 27 [cited 2025 Apr 15];11(10):e0165313. Available from: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165313\u003c/li\u003e\n\u003cli\u003eVergouwe Y, Steyerberg EW, Eijkemans MJC, Habbema JDF. Substantial effective sample sizes were required for external validation studies of predictive logistic regression models. J Clin Epidemiol [Internet]. 2005 May 1 [cited 2025 Apr 7];58(5):475\u0026ndash;83. Available from: https://www.jclinepi.com/action/showFullText?pii=S0895435605000193\u003c/li\u003e\n\u003cli\u003eRiley RD, Snell KIE, Archer L, Ensor J, Debray TPA, van Calster B, et al. Evaluation of clinical prediction models (part 3): calculating the sample size required for an external validation study. BMJ [Internet]. 2024 Jan 22 [cited 2025 Apr 6];384:e074821. Available from: https://www.bmj.com/content/384/bmj-2023-074821\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1. Clinical prediction models for cisplatin related acute kidney injury\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eMotwani et al.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eGupta et al.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDefinition of AKI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eCreatinine \u0026ge; 0.3 mg/dL\u0026nbsp;\u003c/p\u003e\n \u003cp\u003ein 14 days\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eCreatinine \u0026ge; 2.0-fold\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eor RRT in 14 days\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003ePredictors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eAge (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026le;60: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026le;45: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e61\u0026ndash;70: 1.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e46\u0026ndash;60: 2.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;70: 2.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e61\u0026ndash;70: 3.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;70: 4.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eHypertension\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDiabetes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eSmoker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eCDDP dosage (mg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026le;100: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026le;50: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e101\u0026ndash;150: 1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e51\u0026ndash;75: 2 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;150: 3 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e76\u0026ndash;100: 2.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e101\u0026ndash;125: 3 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e126\u0026ndash;150: 5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e151\u0026ndash;200: 7.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;200: 9.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eHemoglobin (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ge;12.0: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e11.0\u0026ndash;11.9: 1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026lt;11.0: 1.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eWBC (\u0026times;10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026le;12.0: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;12.0: 1.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eAlbumin (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;3.5: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026gt;3.8: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026le;3.5: 2 points\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e3.3\u0026ndash;3.8: 1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026lt;3.3: 1.5 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eMagnesium (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026ge;2.0: 0 points\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026lt;2.0: 1 point\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eAcute kidney injury (AKI) was defined as the degree of increase in serum creatinine levels from baseline. RRT, renal replacement therapy; CDDP, cisplatin; WBC, white blood cell count.\u003cbr\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 2. Patients\u0026rsquo;\u0026nbsp;background\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"78%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003eSaito et al.\u003cbr\u003e\u0026nbsp;(n = 1,684)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003eGupta et al.\u003cbr\u003e\u0026nbsp;(n = 11,766)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003eMotwani et al.\u003cbr\u003e\u0026nbsp;(n = 2,128)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eStudy duration\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e2014\u0026ndash;2023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e2006\u0026ndash;2022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e2005\u0026ndash;2014\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eAge, years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e66 (58\u0026ndash;71)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e59 (50\u0026ndash;67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e56.8 (13.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e1,124 (66.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e6,935 (58.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e1,173 (55.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eBMI, kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e21.9 (19.5\u0026ndash;24.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e26.6 (5.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eCDDP dosage, mg\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e110 (90\u0026ndash;130)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e90 (60\u0026ndash;160)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e118.4 (77\u0026ndash;151)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eHypertension\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e595 (35.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e2,900 (24.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e1,046 (49.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eDiabetes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e203 (12.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e1,426 (12.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e300 (14.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eSmoker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e1033 (64.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e6,762 (57.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eCreatinine, mg/dL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e0.75 (0.63\u0026ndash;0.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e0.9 (0.7\u0026ndash;1.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e0.9 (0.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eeGFR,\u0026nbsp;\u003c/p\u003e\n \u003cp\u003emL/min/1.73 m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e74.0 (63.2\u0026ndash;87.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e90.0 (75.0\u0026ndash;101.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e87.3 (19.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eWBC,\u0026nbsp;\u0026times;10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e5.99 (4.64\u0026ndash;7.56)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e7.1 (5.6\u0026ndash;9.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003ePLT,\u0026nbsp;\u0026times;10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e256 (203\u0026ndash;313)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e255 (204\u0026ndash;322)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eHemoglobin, mg/dL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e12.5 (11.3\u0026ndash;13.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e12.8 (11.4\u0026ndash;14.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eAlbumin, g/dL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e3.9 (3.5\u0026ndash;4.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e4.1 (3.7\u0026ndash;4.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e4.0 (0.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eMagnesium, mg/dL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 23px;\"\u003e\n \u003cp\u003e2.05 (2.00\u0026ndash;2.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\n \u003cp\u003e2.19 (1.94\u0026ndash;2.19)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eC-AKI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e186 (11.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 22px;\"\u003e\n \u003cp\u003e289 (13.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003eStage 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e150 (8.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 22px;\"\u003e\n \u003cp\u003e234 (11.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 28px;\"\u003e\n \u003cp\u003e\u0026nbsp; Stage \u003cu\u003e\u0026gt;\u003c/u\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 23px;\"\u003e\n \u003cp\u003e36 (2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 25px;\"\u003e\n \u003cp\u003e608 (5.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 22px;\"\u003e\n \u003cp\u003e55 (2.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003ePatient background characteristics and outcomes in our cohort as well as those reported in previous studies have been summarised. Continuous variables are presented as medians (interquartile ranges), and categorical variables are presented as n (%). In a study by Motwani et al., continuous variables, except for CDDP dose, are reported as means (standard deviations). The proportions of missing data in our cohort were as follows: albumin (6.0%); magnesium (61.2%); platelets (0.1%); and white blood cells (0.7%). BMI, body mass index; CDDP, cisplatin; eGFR, estimated glomerular filtration rate; WBC, white blood cell count; PLT, platelet count; C-AKI, cisplatin related acute kidney injury.\u003c/p\u003e\n\u003cp\u003eTable 3. Indication parameter of model validation\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" class=\"fr-table-selection-hover\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eMotwani model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eAfter recalibration\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eGupta model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 129px;\"\u003e\n \u003cp\u003eAfter recalibration\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" style=\"width: 601px;\"\u003e\n \u003cp\u003eC-AKI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eAUROC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.613 (0.570\u0026ndash;0.656)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.613 (0.570\u0026ndash;0.656)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.616 (0.575\u0026ndash;0.658)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 129px;\"\u003e\n \u003cp\u003e0.616 (0.575\u0026ndash;0.658)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eCITL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u0026minus;0.438 (\u0026ndash;0.604\u0026ndash;\u0026minus;0.283)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026minus;0.004 (\u0026minus;0.161\u0026ndash;0.147)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.346 (0.157\u0026ndash;0.526)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 129px;\"\u003e\n \u003cp\u003e\u0026minus;0.003 (\u0026minus;0.150\u0026ndash;0.138)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eSlope\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.550 (0.340\u0026ndash;0.758)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.006 (0.621\u0026ndash;1.374)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026minus;0.075 (\u0026minus;0.155\u0026ndash;0.035)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 129px;\"\u003e\n \u003cp\u003e0.995 (0.664\u0026ndash;1.375)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eBrier score\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 104px;\"\u003e\n \u003cp\u003e0.102 (0.092\u0026ndash;0.112)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.096 (0.085\u0026ndash;0.108)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.108 (0.095\u0026ndash;0.119)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 129px;\"\u003e\n \u003cp\u003e0.096 (0.086\u0026ndash;0.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" style=\"width: 601px;\"\u003e\n \u003cp\u003eSevere C\u0026ndash;AKI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eAUROC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e0.594 (0.482\u0026ndash;0.697)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e0.594 (0.482\u0026ndash;0.697)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e0.674 (0.584\u0026ndash;0.768)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 129px;\"\u003e\n \u003cp\u003e0.674 (0.584\u0026ndash;0.768)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eCITL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u0026minus;2.242 (\u0026minus;2.602\u0026ndash;\u0026minus;1.921)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026minus;0.023 (\u0026minus;0.403\u0026ndash;0.283)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026minus;1.567 (\u0026minus;1.921\u0026ndash;\u0026minus;1.237)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 129px;\"\u003e\n \u003cp\u003e\u0026minus;0.018 (\u0026minus;0.388\u0026ndash;0.304)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eSlope\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e0.523 (0.090\u0026ndash;1.008)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e0.993 (0.057\u0026ndash;1.972)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026minus;0.005 (\u0026minus;0.178\u0026ndash;0.469)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 129px;\"\u003e\n \u003cp\u003e0.996 (0.485\u0026ndash;1.495)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eBrier score\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e0.047 (0.042\u0026ndash;0.052)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e0.021 (0.014\u0026ndash;0.027)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e0.032 (0.027\u0026ndash;0.038)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 129px;\"\u003e\n \u003cp\u003e0.021 (0.014\u0026ndash;0.027)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe mean values and 95% confidence intervals from the bootstrap samples are reported for each performance indicator. The parameters after recalibration are shown in the right-hand column, along with the corresponding original model indicators. AUROC, area under the receiver operating characteristic curve; CITL; calibration-in-the-large; Slope: Calibration slope.\u0026nbsp;\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"clinical decision rules, cisplatin, acute kidney injury, validation study","lastPublishedDoi":"10.21203/rs.3.rs-6678420/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6678420/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Cisplatin-associated acute kidney injury (C-AKI) is a major complication of cisplatin therapy. Although two clinical prediction models have been developed for the US population, their external validity in the Japanese population remains unclear. This study aimed to evaluate the external validity of these models and compare their predictive performances in a Japanese cohort.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e We assessed the performance of two C-AKI prediction models developed by Motwani et al. and Gupta et al. in a retrospective cohort of 1,684 patients treated with cisplatin at Iwate Medical University Hospital. C-AKI was defined as a ≥0.3 mg/dL increase in serum creatinine or a ≥1.5-fold rise from baseline. Severe C-AKI was defined as a ≥2.0-fold increase or renal replacement therapy initiation. Model performance was evaluated using discrimination (area under the receiver operating characteristic curve [AUROC]), calibration, and decision curve analysis (DCA). Logistic recalibration was applied to adapt the model to the local population.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e The discriminatory performance for C-AKI was similar between the Gupta and Motwani models (AUROC, 0.616 vs. 0.613; p = 0.84). However, the Gupta model showed better discrimination of severe C-AKI (AUROC, 0.674 vs. 0.594; p = 0.02). Both models exhibited poor initial calibrations, which improved after recalibration. The recalibrated models yielded a greater net benefit in the DCA, with the Gupta model demonstrating the highest clinical utility in severe C-AKI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e Both models demonstrated discriminatory ability, with the Gupta model showing particular utility in predicting severe C-AKI. Recalibration may be necessary before applying these models in Japanese clinical practice.\u003c/p\u003e","manuscriptTitle":"External validation and comparison of clinical prediction models for cisplatin- associated acute kidney injury: A single-centre retrospective study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-02 07:11:46","doi":"10.21203/rs.3.rs-6678420/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f2e791dd-6dec-4052-b8ae-cc1a757c7e82","owner":[],"postedDate":"June 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-06T16:09:26+00:00","versionOfRecord":{"articleIdentity":"rs-6678420","link":"https://doi.org/10.1186/s40780-025-00471-0","journal":{"identity":"journal-of-pharmaceutical-health-care-and-sciences","isVorOnly":false,"title":"Journal of Pharmaceutical Health Care and Sciences"},"publishedOn":"2025-09-29 15:57:03","publishedOnDateReadable":"September 29th, 2025"},"versionCreatedAt":"2025-06-02 07:11:46","video":"","vorDoi":"10.1186/s40780-025-00471-0","vorDoiUrl":"https://doi.org/10.1186/s40780-025-00471-0","workflowStages":[]},"version":"v1","identity":"rs-6678420","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6678420","identity":"rs-6678420","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}