Published February 17, 2026 | Version first version
Journal article Open

Machine Learning-Based Mortality Prediction in Hemodialysis Bloodstream Infection: A SHAP-Integrated Approach for Real-Time, Personalized Risk Stratification and Clinical Decision Support

Authors/Creators

  • 1. Associate Professor, Department of Microbiology, College of Medicine , National University
  • 2. Professor, Department of Microbiology, College of Medicine, National University, Oman
  • 3. Department of Physiology , Faculty of Medicine, AL Neelain University Sudan
  • 4. Department of Pathology and Blood bank, Sohar Hospital

Description

Abstract:

Background

Bloodstream infection (BSI) represents a leading cause of morbidity and mortality in hemodialysis populations, with in-hospital mortality rates reaching 50%. Conventional clinical severity scoring systems (qSOFA, APACHE II) demonstrate inadequate predictive accuracy in dialysis cohorts, failing to identify 20-40% of high-risk patients at presentation. Machine learning with explainable artificial intelligence (SHAP) offers potential to improve mortality prediction while maintaining clinical interpretability.

Methods

This retrospective cohort study analyzed 148 hemodialysis patients with confirmed BSI from a tertiary dialysis center (2018-2022). Five machine learning algorithms were developed and validated using 5-fold stratified cross-validation: Random Forest, XGBoost, LightGBM, Logistic Regression, and Support Vector Machine. SHAP-derived feature importance analysis identified patient-specific and population-level mortality drivers. Performance was compared against qSOFA and APACHE II using AUROC as the primary metric.

Results

The cohort demonstrated 50% in-hospital mortality (n=74 deaths). Random Forest achieved superior discriminative ability with AUROC 0.94 (95% CI 0.89–0.98), representing 23-percentage-point improvement over qSOFA (AUROC 0.71, p<0.001) and 18-percentage-point improvement over APACHE II (AUROC 0.76, p<0.001). Cross-validation demonstrated robust generalization (mean AUROC 0.945±0.013, coefficient of variation 1.4%). SHAP analysis identified septic shock (SHAP 0.487) and mechanical ventilation (SHAP 0.412) as dominant mortality predictors. Notably, chronic neurological disease independently amplified mortality risk 3.91-fold but is not captured by traditional scoring systems. Top 10 features ranked identically across all four algorithms, validating finding robustness.

Conclusions

Machine learning with SHAP-based interpretability substantially improves mortality prediction accuracy in hemodialysis BSI compared with conventional clinical scores. SHAP transparency enables identification of patient-specific risk drivers and novel mortality associations, facilitating personalized clinical decision-making. A clinical integration framework demonstrating real-time risk reassessment as clinical status evolves is proposed. Prospective validation studies should assess whether SHAP-guided risk stratification improves patient outcomes compared with conventional management.

Files

Files (2.7 MB)

Name Size Download all
md5:050e6227493cd531fdfd4579320f749d
405.1 kB Download
md5:43bb72de8243301fcbaea2a93c3c0d65
214.0 kB Download
md5:7fc54319c7c5468216b6f3c568c95dd5
319.0 kB Download
md5:1b675be21856c64c9bdca4a0cc588adc
479.7 kB Download
md5:1a919ee912b9b6a7b60f3647cf1df53c
366.9 kB Download
md5:d093f48e238d9b4a76dc8dc4e9a56cb9
172.7 kB Download
md5:9f4f47d8576f78cba513e23217bf05eb
148.0 kB Download
md5:8decd6084e857da7454ac40e08e88884
414.5 kB Download
md5:f65814458d4772047bfe141602425497
15.5 kB Download
md5:158ed1e71d0134e7e994d4e2f52cf086
15.1 kB Download
md5:2aaceb6570d7b3e492b0eeceb94d673c
18.4 kB Download
md5:cbf66f48f6990920ef97aee8177229b9
17.4 kB Download
md5:fd7720190ddcc54a5efa8a99d73afdc9
15.6 kB Download
md5:25a7c7a29a5d39d5932dfdc22af33db3
16.2 kB Download
md5:5e13647bf16040f9bc0d184d7aab6bab
14.4 kB Download
md5:560f6643bb62ef64142538a020131afa
14.6 kB Download
md5:b6e2e011d17a9b17c9bb19885d4ed313
14.9 kB Download

Additional details

References

  • REFERENCES 1. Zhang X, Tian S, Zhang X, Guo F, Chen B, Zhang D, et al. Research and predictive analysis of the disease burden of bloodstream infectious diseases in China. BMC Infect Dis. 2025 Apr 22;25(1):578. doi: 10.1186/s12879-025-10989-1. 2. Rudd KE, Johnson SC, Agesa KM, Shackelford KA, Tsoi D, Kievlan DR, et al. Global, regional, and national sepsis incidence and mortality, 1990-2017: analysis for the Global Burden of Disease Study. Lancet. 2020 Jan 18;395(10219):200-211. doi: 10.1016/S0140-6736(19)32989-7. 3. Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47(11):1181-1247. 4. GBD 2019 Antimicrobial Resistance Collaborators. Global mortality associated with 33 bacterial pathogens in 2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2022;400(10369):2221-2248. 5. La Via L, Sangiorgio G, Stefani S, Marino A, Nunnari G, Cocuzza S, et al. The Global Burden of Sepsis and Septic Shock. Epidemiologia (Basel). 2024 Jul 25;5(3):456-478. doi: 10.3390/epidemiologia5030032. 6. Weldetensae MK, Weledegebriel MG, Nigusse AT, Berhe E, Gebrearegay H. Catheter-Related Blood Stream Infections and Associated Factors Among Hemodialysis Patients in a Tertiary Care Hospital. Infect Drug Resist. 2023 May 22;16:3145-3156. doi: 10.2147/IDR.S409400. 7. WHO Calls for Global Action on Sepsis—Cause of 1 in 5 Deaths Worldwide. Consultato: 11 Giugno 2024. [(accessed on 1 June 2024)]. Available online: https://www.who.int/news/item/08-09-2020-who-calls-for-global-action-on-sepsis---cause-of-1-in-5-deaths-worldwide. 8. Guarino M, Perna B, Cesaro AE, Maritati M, Spampinato MD, Contini C, et al. Update on Sepsis and Septic Shock in Adult Patients: Management in the Emergency Department. J Clin Med. 2023;12(9):3188. 9. Srzić I, Nesek Adam V, Tunjić Pejak D. SEPSIS DEFINITION: WHAT'S NEW IN THE TREATMENT GUIDELINES. Acta Clin Croat. 2022;61(Suppl 1):67-72. 10. Bou Chebl R, Tamim H, Abou Dagher G, Sadat M, Ghamdi G, Itani A, et al. Sepsis in end-stage renal disease patients: are they at an increased risk of mortality? Ann Med. 2021 Dec;53(1):1737-1743. doi: 10.1080/07853890.2021. 11. Schamroth Pravda M, Maor Y, Brodsky K, Katkov A, Cernes R, Schamroth Pravda N, et al. Blood stream Infections in chronic hemodialysis patients - characteristics and outcomes. BMC Nephrol. 2024 Jan 3;25(1):3. doi: 10.1186/s12882-023-03442-5. 12. Yang Z, Chen X, Pi M, Tang Y, Lam RK, Yang C, et al. Infection and Immunodeficiency in Patients With Chronic Kidney Disease: Current Evidence and Potential Preventive Strategies of Chinese Medicine. Integr Med Nephrol Androl. 2025 June;12(2):e24-00028. DOI: 10.1097/IMNA-D-24-00028. 13. Locham S, Naazie I, Canner J, Siracuse J, Al-Nouri O, Malas M. Incidence and risk factors of sepsis in hemodialysis patients in the United States. J Vasc Surg. 2021;73(3):1016-1021.e3. 14. Venegas-Ramírez J, Hernández-Fuentes GA, Palomares CS, Diaz-Martinez J, Navarro-Cuellar JI, Calvo-Soto P, et al. Vascular Access Type and Survival Outcomes in Hemodialysis Patients: A Seven-Year Cohort Study. Medicina (Kaunas). 2025 Mar 25;61(4):584. doi: 10.3390/medicina61040584. 15. Harms JC, Rangarajan S, Young CJ, Barker-Finkel J, Allon M. Outcomes of arteriovenous fistulas and grafts with or without intervention before successful use. J Vasc Surg. 2016 Jul;64(1):155-62. doi: 10.1016/j.jvs.2016.02.033. 16. Hotchkiss RS, Moldawer LL, Opal SM, Reinhart K, Turnbull IR, Vincent JL. Sepsis and septic shock. Nat Rev Dis Primers. 2016;2:16045. 17. Vijayan AL, Vanimaya, Ravindran S, Saikant R, Lakshmi S, Kartik R, et al. Procalcitonin: a promising diagnostic marker for sepsis and antibiotic therapy. J Intensive Care. 2017;5:51. 18. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016 Feb 23;315(8):801-10. 19. AbuTaha SA, Al-Kharraz T, Belkebir S, Abu Taha A, Zyoud SH. Patterns of microbial resistance in bloodstream infections of hemodialysis patients: a cross-sectional study from Palestine. Sci Rep. 2022;12(1):18003. 20. Rteil A, Kazma JM, El Sawda J, Gharamti A, Koubar SH, Kanafani ZA. Clinical characteristics, risk factors and microbiology of infections in patients receiving chronic hemodialysis. J Infect Public Health. 2020;13(8):1166-1171. 21. Centers for Disease Control and Prevention. Vital Signs: Health Disparities in Hemodialysis-Associated Staphylococcus aureus Bloodstream Infections — United States, 2017–2020. MMWR Morb Mortal Wkly Rep. 2023;72(6):133-140. 22. Zilberman-Itskovich S, Elukbi Y, Weinberg Sibony R, Shapiro M, Zelnik Yovel D, Strulovici A, et al. The Epidemiology of Multidrug-Resistant Sepsis among Chronic Hemodialysis Patients. Antibiotics. 2022;11(9):1255. 23. Sakr Y, Jaschinski U, Wittebole X, Szakmany T, Lipman J, Namendys-Silva SA, et al. Sepsis in intensive care unit patients: worldwide data from the intensive care over nations audit. Open Forum Infect Dis. 2018;5(10):ofy313. 24. Hassoun-Kheir N, Buetti N, Jaques D, Damonti L, Giulieri S, Huttner B, et al. Secular trends of bloodstream infections in hemodialysis patients: insights from a longitudinal Swiss study. Antimicrob Resist Infect Control. 2025;14:100. 25. Seok H, Jeon JH, Park DW. Antimicrobial Therapy and Antimicrobial Stewardship in Sepsis. Infect Chemother. 2020;52(1):19-30. 26. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing; CLSI M100, edition 30. Wayne (PA): Clinical and Laboratory Standards Institute; 2020. 27. Qin Y, Liao Y, Zhou J, Liu W, Chen H, Chen X, et al. Comparative evaluation of BacT/ALERT VIRTUO and BACTEC FX400 blood culture systems for the detection of bloodstream infections. Microbiol Spectr. 2025 Jan 7;13(1):e0185024. doi: 10.1128/spectrum.01850-24. 28. Can Ö, Bilek G, Sahan S. Risk factors for infection and mortality among hemodialysis patients during COVID-19 pandemic. Int Urol Nephrol. 2022;54(3):661-669. 29. Vergara A, Molina-Van den Bosch M, Toapanta N, Villegas A, Sánchez-Cámara L, Sequera P, et al. The Impact of Age on Mortality in Chronic Haemodialysis Popu-Lation with COVID-19. J Clin Med. 2021 Jul 7;10(14):3022. doi: 10.3390/jcm101403022. 30. Kazakova SV, Baggs J, Apata IW, Yi SH, Jernigan JA, Nguyen D, et al. Vascular Access and Risk of Bloodstream Infection Among Older Incident Hemodialysis Patients. Kidney Med. 2020 Mar 16;2(3):276-285. doi: 10.1016/j.xkme.2019.12.010. 31. Sannathimmappa MB, Nambiar V, Al Siyabi KHSH, Hussain AS, Shah YA, Marimuthu Y, et al. Antimicrobial Resistance Pattern, Predisposing Factors, and Outcome of Serratia Infection in Patients Treated at a Secondary-Care Hospital in Oman: A 5-Year Retrospective Study. Adv Biomed Res. 2024 ;13:101. 32. Sannathimmappa MB, Nambiar V, Aravindakshan R, Al-Risi E. Clinical profile and antibiotic susceptibility pattern of Enterococcus faecalis and Enterococcus faecium with an emphasis on vancomycin resistance. Biomed Biotechnol Res J. 2023; 7:283-7. 33. Sannathimmappa MB. Global escalation in carbapenem-resistant Enterobacterales (CRE) and carbapenem-resistant Acinetobacter baumannii (CRAB) infections: Serious threat to human health from the pink corner. Biomed Biotechnol Res J. 2023; 7:9-16. 34. Sannathimmappa MB, Nambiar V, Aravindakshan R, Al-Kasaby NM. Patterns of methicillin-resistant Staphylococcus aureus (MRSA) strains isolated at a tertiary care hospital. J Datta Meghe Inst Med Sci Univ 2022; 17:228-33. 35. Sannathimmappa MB, Nambiar V, Aravindakshan R. A 5-Year Retrospective Study of Antibiotic Resistance Pattern of Pseudomonas aeruginosa Isolated from Various Clinical Samples at a Tertiary Care Hospital. Medical Journal of Dr. D.Y. Patil Vidyapeeth. 2023; 16(5): 687-692. 36. Tanzarella ES, Cutuli SL, Lombardi G, Cammarota F, Caroli A, Franchini E, et al. Antimicrobial De-Escalation in Critically Ill Patients. Antibiotics. 2024;13(4):375. 37. Shimada Y, Ohbe H, Kutsuna S, Kosaka S, Matsui H, Yasunaga H. Effect of Empirical Antimicrobial Therapy of Piperacillin-tazobactam Versus Carbapenems on Mortality for Patients with Sepsis from Nursing Homes. Intern Med. 2025 Jun 15;64(12):1820-1827. doi: 10.2169/internalmedicine.4426-24. 38. Corona A, De Santis V, Agarossi A, Prete A, Cattaneo D, Tomasini G, et al. Antibiotic Therapy Strategies for Treating Gram-Negative Severe Infections in the Critically Ill: A Narrative Review. Antibiotics (Basel). 2023;12(8):1262. 39. Abdul Salim S, Masoud AT, Thongprayoon C, Cheungpasitporn W, Soliman KM, Garla V, et al. Systematic Review and Meta-Analysis of Antibiotic and Antimicrobial Lock Solutions for Prevention of Hemodialysis Catheter-Related Infections. ASAIO J. 2021 Oct 1;67(10):1079-1086. doi: 10.1097/MAT.0000000000001382. 40. Tylicki L, Puchalska-Reglińska E, Tylicki P, Och A, Polewska K, Biedunkiewicz B, et al. Predictors of Mortality in Hemodialyzed Patients after SARS-CoV-2 Infection. J Clin Med. 2022;11(2):285. 41. Wang F, Ao G, Wang Y, Liu F, Bao M, Gao M, et al. Risk factors for mortality in hemodialysis patients with COVID-19: a systematic review and meta-analysis. Ren Fail. 2021;43(1):1394-1407. 42. Kinnunen S, Helanterä I, Juutilainen A, Bitar W, Helve J, Finne P. Time Trends and Causes of Infection-Related Mortality Among Patients Starting Dialysis in Finland: A Nationwide Cohort Study. Kidney Med. 2025 Apr 18;7(6):101012. doi: 10.1016/j.xkme.2025.101012. 43. Li XL, Mi T, Liu C, Feng M. A Mortality Risk Prediction Model for Septic Shock in Patients Aged ≥50: Role of Norepinephrine Index and Procalcitonin. Int J Gen Med. 2025 Jun 10;18:3045-3062. doi: 10.2147/IJGM.S520290.