Evaluation of the Performance of Biomarkers in Predicting Hemodialysis Patients Survival Using Time-Dependent ROC Curve
DOI:
https://doi.org/10.61882/ijkd.20.02.9299Keywords:
Time-dependent ROC curve, Survival, HemodialysisAbstract
Introduction. End-stage kidney disease (ESKD) is a growing global public health problem, and patients undergoing hemodialysis (HD) experience a high mortality rate despite advances in medical care. Identifying reliable prognostic biomarkers is therefore essential. Regression-based survival models are widely used in HD studies; however, while they estimate associations, they do not directly quantify the predictive accuracy of biomarkers. This study aimed to evaluate the predictive performance of selected baseline biomarkers for survival in HD patients using the time-dependent receiver operating characteristic (ROC) curve.
Methods. This retrospective cohort study included 2,192 ESKD patients undergoing maintenance HD in Fars province, Iran, between 2011 and 2020. Time-dependent area under the ROC curve (AUC) values were calculated to assess the prognostic performance of baseline biomarkers at 3, 12, 24, and 36 months after initiation of dialysis.
Results. Age, serum albumin, creatinine, and calcium showed relatively better predictive performance for survival. For age, the AUCs at 3, 12, 24, and 36 months were 64.2, 57.1, 57.1, and 58.8, respectively, while corresponding values for serum albumin were 67.5, 62.3, 61.4, and 59.8. Serum albumin and calcium demonstrated higher discrimination for early survival, whereas serum creatinine showed more stable predictive performance over the follow-up period. The combined risk score outperformed individual biomarkers, with AUCs of 71.2, 63.9, 63.5, and 64.1 at 3, 12, 24, and 36 months, respectively.
Conclusion. Time-dependent ROC analysis revealed time-varying prognostic performance of baseline biomarkers in HD patients and demonstrated improved discrimination when biomarkers were combined into a composite risk score.
Downloads
References
1. Thurlow J, Joshi M, Yan G, et al. Epidemiology of
End-Stage Kidney Disease and Disparities in Kidney Replacement Therapy. Am J Nephrol. 2021; 52(2):
98–107.
2. Shahbazi F, Doosti-Irani A, Soltanian A, Poorolajal J. National trends and projection of chronic kidney disease incidence according to etiology from 1990 to 2030 in Iran: a Bayesian age-period-cohort modeling study. Epidemiol Health. 2023; 45: e2023027.
3. Shahbazi F, Doosti-Irani A, Soltanian AR, Poorolajal J. Global forecasting of chronic kidney disease mortality rates and numbers with the generalized additive model. BMC Nephrology. 2024; 25: 286.
4. Hill NR, Fatoba ST, Oke JL, et al. Global Prevalence of Chronic Kidney Disease - A Systematic Review and Meta-Analysis. PloS one. 2016; 11(7): e0158765.
5. Shayan Z, Ebrahimi V, Ezzatzadegan J Sh. Factors Associated ith Short-Term and Long-Term Survival in Patients Undergoing Hemodialysis Using Semi-Parametric Mixture Cure Models. Iran J Kidney Dis. 2020; 14(1):44-51.
6. Bansal A, Heagerty PJ. A comparison of landmark methods and time-dependent ROC methods to evaluate the time-varying performance of prognostic markers for survival outcomes. Diagn Progn Res. 2019; 3:14.
7. Zhu Z, Wang X, Saha-Chaudhuri P, Kosinski AS, George SL. Time-dependent classification accuracy curve under marker-dependent sampling. Biom J. 2016; 58(4):974-92.
8. Shen W, Ning J, Yuan Y. A direct method to evaluate
the time-dependent predictive accuracy for biomarkers. Biometrics. 2015; 71(2):439-449.
9. Kamarudin AN, Cox T, Kolamunnage-Dona R. Time-dependent ROC curve analysis in medical research: current methods and applications. BMC Med Res Method. 2017; 17(1):53.
10. Heagerty PJ, Lumley T, Pepe MS. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics. 2000; 56(2):337-344.
11. Heagerty PJ, Zheng Y. Survival model predictive accuracy and ROC curves. Biometrics. 2005; 61(1):92-105.
12. Pepe MS, Zheng Y, Jin Y, Huang Y, Parikh CR, Levy WC. Evaluating the ROC performance of markers for future events. Lifetime data Anal. 2008; 14(1):86-113.
13. Cai T, Pepe MS, Zheng Y, Lumley T, Jenny NS. The sensitivity and specificity of markers for event times. Biostatistics. 2006; 7(2):182-197.
14. Firouraghi N, Ezzatzadegan J Sh, Sami A, et al. Data Quality Assessment and Recommendations to Improve the Quality of Hemodialysis Database. J Health Man & Info. 2018; 5(1):22.
15. Payne RB, Little AJ, Williams RB, Milner JR. Interpretation of serum calcium in patients with abnormal serum proteins. Br Med J. 1973; 4(5893):643-646.
16. Buuren S, Groothuis-Oudshoorn C. MICE: Multivariate Imputation by Chained Equations in R. J Stat Softw. 2011; 45.
17. Barnard J, Rubin DB. Small-Sample Degrees of Freedom with Multiple Imputation. Biometrika. 1999; 86(4):948-955.
18. Blanche P, Dartigues JF, Jacqmin-Gadda H. Review and comparison of ROC curve estimators for a time-dependent outcome with marker-dependent censoring. Biom J. 2013; 55(5):687-704.
19. Blanche P, Dartigues J-F, Jacqmin-Gadda H. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Statist Med. 2013; 32(30):5381-5397.
20. Hung H, Chiang C-T. Estimation methods for time-dependent AUC models with survival data. Can J Stat. 2010; 38(1):8-26.
21. Blanche P. timeROC: Time-dependent ROC curve and AUC for censored survival data. R package version
0.2–2015
22. Collins AJ, Foley RN, Herzog C, et al. Excerpts from the US Renal Data System 2009 Annual Data Report. Am J kidney dis. 2010; 55(1 Suppl 1):S1-420, A6-7.
23. Mailloux LU, Bellucci AG, Napolitano B, Mossey T, Wilkes BM, Bluestone PA. Survival estimates for 683 patients starting dialysis from 1970 through 1989: identification of risk factors for survival. Clin Nephrol. 1994; 42(2):127-135.
24. Ukowsky LR, Kheifets L, Arah OA, Nissenson AR, Kalantar-Zadeh K. Patterns and predictors of early mortality in incident hemodialysis patients: new insights. Am J Nephrol. 2012; 35(6):548-558.
25. Walther CP, Carter CW, Low CL, et al. Interdialytic creatinine change versus predialysis creatinine
as indicators of nutritional status in maintenance hemodialysis. Nephrol Dial Transplant. 2012; 27(2):771-776.
26. Park J, Mehrotra R, Rhee CM, et al. Serum creatinine level, a surrogate of muscle mass, predicts mortality in peritoneal dialysis patients. Nephrol Dial Transplant. 2013; 28(8):2146-2155.
27. Kalantar-Zadeh K, Streja E, Molnar MZ, et al. Mortality prediction by surrogates of body composition: an examination of the obesity paradox in hemodialysis patients using composite ranking score analysis. Am J Epidemiol. 2012; 175(8):793-803.
28. Vodičar J, Pajek J, Hadžić V, Bučar Pajek M. Relation
of Lean Body Mass and Muscle Performance to Serum Creatinine Concentration in Hemodialysis Patients. Biomed Res Int. 2018; 2018:4816536.
29. Kalantar-Zadeh K, Kilpatrick RD, Kuwae N, et al. Revisiting mortality predictability of serum albumin in
the dialysis population: time dependency, longitudinal changes and population-attributable fraction. Nephrol Dial Transplant. 2005; 20(9):1880-1888.
30. Goldwasser P, Kaldas AI, Barth RH. Rise in serum albumin and creatinine in the first half year on hemodialysis. Kidney Int. 1999; 56(6):2260-2268.
31. Tentori F, Blayney MJ, Albert JM, et al. Mortality risk for dialysis patients with different levels of serum calcium, phosphorus, and PTH: the Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J kidney Dis. 2008; 52(3):519-530.
32. Block GA, Kilpatrick RD, Lowe KA, Wang W, Danese MD. CKD-mineral and bone disorder and risk of death and cardiovascular hospitalization in patients on hemodialysis. Clin J Am Soc Nephrol. 2013; 8(12):2132-2140.
33. Rivara MB, Ravel V, Kalantar-Zadeh K, et al. Uncorrected and Albumin-Corrected Calcium, Phosphorus, and Mortality in Patients Undergoing Maintenance Dialysis. J Am Soc Nephrol. 2015; 26(7):1671-1681.
34. Jain A, Bhayana S, Vlasschaert M, House A. A formula to predict corrected calcium in haemodialysis patients. Nephrol Dial Transplant. 2008; 23(9):2884-2888.
Downloads
Published
Data Availability Statement
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
