CoChat
Sign Up
Cited by 17Sepsis Diagnosis and TreatmentArtificial Intelligence in HealthcareDialysis and Renal Disease Management

Early identification of bloodstream infection in hemodialysis patients by machine learning

Open access
Zhou, Tong, Ren, Zhouting, Ma, Yimei, He, Linqian, Liu, Jiali, Tang, Jincheng, Zhang, Heping·2023·Heliyon
DOI 10.1016/j.heliyon.2023.e18263PMC PMC10375788
Open this paper in CoChat — free account

Abstract

Background: Bloodstream infection (BSI) is a prevalent cause of admission in hemodialysis (HD) patients and is associated with increased morbidity and mortality. This study aimed to establish a diagnostic, predictive model for the early identification of BSI in HD patients. Methods: HD patients who underwent blood culture testing between August 2018 and March 2022 were enrolled in this study. Machine learning algorithms, including stepwise logistic regression (SLR), Lasso logistic regression (LLR), support vector machine (SVM), decision tree, random forest (RF), and gradient boosting machine (XGboost), were used to predict the risk of developing BSI from the patient's clinical data. The accuracy (ACC) and area under the subject working curve (AUC) were used to evaluate the performance of such models. The Shapley Additive Explanation (SHAP) values were used to explain each feature's predictive value on the models' output. Finally, a simplified nomogram for predicting BSI was devised. Results: A total of 391 HD patients were enrolled in this study, of whom 74 (18.9%) were diagnosed with BSI. The XGboost model achieved the highest AUC (0.914, 95% confidence interval [CI]: 0.861-0.964) and ACC (86.3%) for BSI prediction. The four most significant co-variables in both the significance matrix plot of the XGboost model variables and the SHAP summary plot were body temperature, dialysis access via a non-arteriovenous fistula (non-AVF), the procalcitonin levels (PCT), and neutrophil-lymphocyte ratio (NLR). Conclusions: This study created an effective machine-learning model for predicting BSI in HD patients. The model could be used to detect BSI at an early stage and hence guide antibiotic treatment in HD patients.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1016/j.heliyon.2023.e18263

Cites

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Most reported ML-based BSI prediction models integrate a combination of vital signs, clinical scores, and laboratory values.

    “However, most reported ML-based BSI prediction models integrate a combination of vital signs, clinical scores, and laboratory values.”

    From Early prediction of bloodstream infections in ICU patients using machine learning methods based on routine laboratory parameters (2025) · cites this paper as [11]

    The Methods section ('Predictors and outcome measures'; page unavailable) says this paper’s ML models used vital signs and laboratory measurements. It does not say that most reported ML-based BSI models use that combination, and its listed predictors do not include clinical scores. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

  • Recent studies have explored various ML algorithms, such as logistic regression, random forest, support vector machines, and gradient boosting, to assess the risk of BSIs in real time.

    “Recent studies have explored various ML algorithms, such as logistic regression, random forest, support vector machines, and gradient boosting, to assess the risk of BSIs in real time [10–14].”

    From Early prediction of bloodstream infections in ICU patients using machine learning methods based on routine laboratory parameters (2025) · cites this paper as [11]

    The Methods and Introduction describe using logistic regression, random forest, support vector machines, and XGboost to predict bloodstream infection in hemodialysis patients. The paper aims at early prediction, but it does not report real-time risk assessment; its models were evaluated on patient datasets. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

  • Machine learning models have demonstrated significant potential in predicting BSIs by analyzing readily available clinical and laboratory data.

    “Machine learning models have demonstrated significant potential in predicting BSIs by analyzing readily available clinical and laboratory data.”

    From Early prediction of bloodstream infections in ICU patients using machine learning methods based on routine laboratory parameters (2025) · cites this paper as [11]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

No references could be resolved yet.

Research works printed but not linked: 38

  1. [1]Lars S.D. Mette N. Bente J. Søren J.-F. Jørgen Ø.L. Carl S.H. Schmeltz S.O. Risk and prognosis of bloodstream infections among patients on chronic hemodialysis: a population-based cohort study PLoS One 10 2015 e0124547 PMC4409390 25910221
  2. [2]C.A J. F.R N. T G.D. Shu-Cheng C. United States Renal Data System public health surveillance of chronic kidney disease and end-stage renal disease Kidney Int. Suppl. 5 2015 2 7 PMC4455192 26097778
  3. [3]P.P R. K.A J. A.M J. Epidemiology, surveillance, and prevention of bloodstream infections in hemodialysis patients Am. J. Kidney Dis. 56 2010 566 577 20554361
  4. [4]Emil N.-K.K. Hellmund L.G. Goya H.J. Risk of death after first-time blood stream infection in incident dialysis patients with specific consideration on vascular access and comorbidity BMC Infect. Dis. 18 2018 1 12 30572826 PMC6302499
  5. [5]R.F B. R.C S. S D.M. K.F N. Ashit H. M.A P. J.V R. Prospective comparison of eubacterial PCR and measurement of procalcitonin levels with blood culture for diagnosing septicemia in intensive care unit patients J. Clin. Microbiol. 47 2009 2964 2969 19641068 PMC2738124
  6. [6]I S.N. W.R E. W.S B. Richard M. B.D W. Who needs a blood culture? A prospectively derived and validated prediction rule J. Emerg. Med. 35 2008 255 264 18486413
  7. [7]Erica H. Payal P. W.L L. Anna C. F.S A. Vineet C. A model to predict central-line–associated bloodstream infection among patients with peripherally inserted central catheters: the MPC score Infect. Control Hosp. Epidemiol. 38 2017 1155 1166 28807074
  8. [8]Keyvan R. Anurag G. Gina B. Jana H. Jacob C. Qingqing M. Ritankar D. Early prediction of central line associated bloodstream infection using machine learning Am. J. Infect. Control 50 2022 440 445 34428529
  9. [9]Pc1 T. Cm S. Cc G. St M. Ra M. Vascular access in haemodialysis patients: a modifiable risk factor for bacteraemia and death QJM: Int. J. Med. 100 2007 415 422 17526532
  10. [10]Fysaraki M. Samonis G. Valachis A. Daphnis E. Karageorgopoulos D.E. Falagas M.E. Stylianou K. Kofteridis D.P. Incidence, clinical, microbiological features and outcome of bloodstream infections in patients undergoing hemodialysis Int. J. Med. Sci. 10 2013 1632 1638 10.7150/ijms.6710 24151435 PMC3804789 · DOI
  11. [11]Fram D. Okuno M.F.P. Taminato M. Ponzio V. Manfredi S.R. Grothe C. Belasco A. Sesso R. Barbosa D. Risk factors for bloodstream infection in patients at a Brazilian hemodialysis center: a case–control study BMC Infect. Dis. 15 2015 10.1186/s12879-015-0907-y PMC4377039 25879516 · DOI
  12. [12]Sho S. Takeshi H. Hiroo K. Atsushi N. Daisuke U. Takahiro I. Masahide F. Hiroki N. Shingo F. Yugo S. Shunichi F. Development and validation of a clinical prediction rule for bacteremia among maintenance hemodialysis patients in outpatient settings PLoS One 12 2017 e0169975 10.1371/journal.pone.0169975 PMC5231279 28081211 · DOI
  13. [13]Fei J. Yong J. Hui Z. Yi D. Hao L. Sufeng M. Yilong W. Qiang D. Haipeng S. Yongjun W. Artificial intelligence in healthcare: past, present and future Stroke Vas. Neur. 2 2017 PMC5829945 29507784
  14. [14]Alvin R. Jeffrey D. Isaac K. Machine learning in medicine N. Engl. J. Med. 380 2019 1347 1358 30943338
  15. [16]Shu-Ling F. M.N S. John L. R.D G. Diagnosing sepsis–the role of laboratory medicine Clin. Chim. Acta 460 2016 203 210 27387712 PMC4980259
  16. [17]Matteo B. Alessandro R. Elda R. Elisabetta D. Maria M. Federica D. Assunta S. Francesco C. Role of procalcitonin in bacteremic patients and its potential use in predicting infection etiology Expert Rev. Anti-Inf. Ther. 17 2019 99 105 30582713
  17. [18]Jiawei J. Rui L. Xin Y. Rui Y. Hua X. Zhi M. Yongqiang W. The neutrophil-lymphocyte count ratio as a diagnostic marker for bacteraemia: a systematic review and meta-analysis Am. J. Emerg. Med. 37 2019 1482 1489 30413366
  18. [19]R.C D. Arun P. G.H J. B.N S. Philipp S. de J.C.P.C. L.A.J M. M.G I. Kenneth B.J. The utility of peripheral blood leucocyte ratios as biomarkers in infectious diseases: a systematic review and meta-analysis J. Infect. 78 2019 339 348 10.1016/j.jinf.2019.02.006 30802469 PMC7173077 · DOI
  19. [20]W.C J. Hypoalbuminemia as surrogate and culprit of infections Int. J. Mol. Sci. 22 2021 4496 33925831 PMC8123513
  20. [21]Lalathaksha K. Jerry Y. Current concepts in hemodialysis vascular access infections Adv. Chron. Kidney Dis. 26 2019 16 22 30876612
  21. [22]Sho S. Yoshihiko R. Minoru M. Shungo Y. Kentaro T. Takeshi H. Kiichiro F. Shunichi F. Added value of clinical prediction rules for bacteremia in hemodialysis patients: an external validation study PLoS One 16 2021 e0247624 PMC7899347 33617601
  22. [25]Bin Y. Wenying Q. Cheng C. Anjun M. Jing J. Hongyan Z. Qin M. SubMito-XGBoost: predicting protein submitochondrial localization by fusing multiple feature information and eXtreme gradient boosting Bioinformatics 36 2020 1074 1081 31603468
  23. [26]Yingchang L. Mingyang L. Chao L. Zhenzhen L. Forest aboveground biomass estimation using Landsat 8 and Sentinel-1A data with machine learning algorithms Sci. Rep. 10 2020 1 12 32561836 PMC7305324
  24. [27]L.S M. Su-In L. A unified approach to interpreting model predictions Adv. Neural Inf. Process. Syst. 2017 30
  25. [28]J.M K. Julie M. Sw H.A.M. Svend E.-E. Simon S. Hans K. H. C S. S.N I. Prediction of bacteremia in the emergency department: an external validation of a clinical decision rule Eur. J. Emerg. Med. 23 2016 44 49 25222426
  26. [29]Eliot H.L. Nicholas D. Richard V. D.B D. F.L G. An external validation study of a clinical prediction rule for medical patients with suspected bacteraemia Emerg. Med. J. 33 2016 124 129 26246024
  27. [30]C. Tianqi, G. Carlos, Xgboost: a scalable tree boosting system, in: Proceedings of the 22nd Acm Sigkdd International Conference on Knowledge Discovery and Data Mining, n.d.: pp. 785–794.
  28. [31]L.C E. H.T S. Timmy L. Surendra S. S Y.A. Kenneth A. Michael A. Arif A. A.B C. G.M H. KDOQI clinical practice guideline for vascular access: 2019 update Am. J. Kidney Dis. 75 2020 S1–S164 32778223
  29. [32]Feldman H.I. Kobrin S. Wasserstein A. Hemodialysis vascular access morbidity J. Am. Soc. Nephrol. 7 1996 523 535 10.1681/asn.v74523 8724885 · DOI
  30. [33]S.C A. A.R C. W.A W. D.J J. B.E J. J.B M. M.J T. N.K A. Outcomes of arteriovenous fistula creation after the fistula first initiative Clin. J. Am. Soc. Nephrol. 6 2011 1996 2002 21737851 PMC3156429
  31. [34]Young E.W. Goodkin D.A. Mapes D.L. Port F.K. Keen M.L. Chen K. Maroni B.L. Wolfe R.A. Held P.J. The dialysis outcomes and practice patterns study (DOPPS): an international hemodialysis study Kidney Int. 57 2000 S74 10.1046/j.1523-1755.2000.07413.x –S81 · DOI
  32. [35]Fissell R.B. Fuller D.S. Morgenstern H. Gillespie B.W. Mendelssohn D.C. Rayner H.C. Robinson B.M. Schatell D. Kawanishi H. Pisoni R.L. Hemodialysis patient preference for type of vascular access: variation and predictors across countries in the DOPPS J. Vasc. Access 14 2013 264 272 10.5301/jva.5000140 23599135 · DOI
  33. [36]Marcel A. Josette B.C.G.D.R. H C. G J. High serum procalcitonin concentrations in patients with sepsis and infection Lancet 341 1993 515 518 8094770 PMC7141580
  34. [37]Mitra B. Faranak A. Ali B.M. Satarzadeh T.M. Reza S.A. Hamid K. Comparison of WBC, ESR, CRP and PCT serum levels in septic and non-septic burn cases Burns 34 2008 770 774 18513877
  35. [38]Franz R. Michael S. Katherina E. Irene T. Marlene B. Helmuth H. Dieter M. Michael B. Heinz B. Utility of sepsis biomarkers and the infection probability score to discriminate sepsis and systemic inflammatory response syndrome in standard care patients PLoS One 8 2013 e82946 PMC3859603 24349403
  36. [39]Raluca D. Dimitrie S. Simona H. Irina M. Adrian C. Procalcitonin: diagnostic value in systemic infections in chronic kidney disease or renal transplant patients Int. Urol. Nephrol. 46 2014 461 468 23990496
  37. [40]Falcão G.P. Menezes F.L. Duque P.I. Procalcitonin as biomarker of infection: implications for evaluation and treatment Am. J. Therapeut. 24 2017 e243–e249 25803230
  38. [41]for Disease C.C. Prevention, Reducing bloodstream infections in an outpatient hemodialysis center--New Jersey, 2008-2011 MMWR. Morbid. Mort. Weekly Report 61 2012 169 173 22419049

Cited by

Sorted by citations

Authors
Zhou, T., Ren, Z., Ma, Y., He, L., Liu, J., Tang, J., & Zhang, H.
Affiliations
  • Affiliated Hospital of North Sichuan Medical College (CN)
  • North Sichuan Medical University (CN)
Corresponding author
Zhang, Heping
ORCID
Venue
Heliyon
Publisher
Elsevier BV
Published
2023-07
Volume
9
Issue
7
Pages
e18263
ISSN
2405-8440 (ISSN-L)
DOI
10.1016/j.heliyon.2023.e18263
PMC
PMC10375788
OpenAlex
W4384156370
Cited by
17
References
0
Citations by year
2026: 1 · 2025: 3 · 2024: 12 · 2023: 1
FWCI
3.39
Type
journal article
Language
en
Open access
Licence
Locations
  • Heliyon (publishedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Health Sciences › Medicine › Epidemiology › Sepsis Diagnosis and Treatment
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
  • Health Sciences › Medicine › Nephrology › Dialysis and Renal Disease Management
Keywords
bloodstream infection, hemodialysis, early infection detection, infection risk prediction, XGBoost, SHAP, nomogram, procalcitonin, neutrophil-to-lymphocyte ratio, body temperature, antibiotic therapy

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review