EUVIMEDTutustu EUVIMEDiin
Multizentrische StudieValidierungsstudieOpen access

Complete Blood Count and Monocyte Distribution Width–Based Machine Learning Algorithms for Sepsis Detection: Multicentric Development and External Validation Study

Andrea Campagner iD, Luisa Agnello iD, Anna Carobene iD, Andrea Padoan iD, Fabio Del Ben iD, Massimo Locatelli iD, Mario Plebani iD, Agostino Ognibene iD, Maria Lorubbio iD, Elena De Vecchi iD, Andrea Cortegiani iD, Elisa Piva iD and 4 more

Journal of Medical Internet Research · 2025

DOI: 10.2196/55492 ↗

At a glance

Multizentrische Studie in Journal of Medical Internet Research, 2025. Written by Andrea Campagner and 15 co-authors. Institution: IRCCS Ospedale Galeazzi Sant'Ambrogio, Milan, Italy. Major MEDLINE topics: Sepsis, Machine Learning, Monocytes. Cited 13 times so far (Crossref); 67 references. The full text is freely available in PubMed Central.

TiivistelmäSubjectsBibliografiset tiedotCiteRelated studies

Abstract (original language)

Background

Sepsis is an organ dysfunction caused by a dysregulated host response to infection. Early detection is fundamental to improving the patient outcome. Laboratory medicine can play a crucial role by providing biomarkers whose alteration can be detected before the onset of clinical signs and symptoms. In particular, the relevance of monocyte distribution width (MDW) as a sepsis biomarker has emerged in the previous decade. However, despite encouraging results, MDW has poor sensitivity and positive predictive value when compared to other biomarkers.

Objective

This study aims to investigate the use of machine learning (ML) to overcome the limitations mentioned earlier by combining different parameters and therefore improving sepsis detection. However, making ML models function in clinical practice may be problematic, as their performance may suffer when deployed in contexts other than the research environment. In fact, even widely used commercially available models have been demonstrated to generalize poorly in out-of-distribution scenarios.

Methods

In this multicentric study, we developed ML models whose intended use is the early detection of sepsis on the basis of MDW and complete blood count parameters. In total, data from 6 patient cohorts (encompassing 5344 patients) collected at 5 different Italian hospitals were used to train and externally validate ML models. The models were trained on a patient cohort encompassing patients enrolled at the emergency department, and it was externally validated on 5 different cohorts encompassing patients enrolled at both the emergency department and the intensive care unit. The cohorts were selected to exhibit a variety of data distribution shifts compared to the training set, including label, covariate, and missing data shifts, enabling a conservative validation of the developed models. To improve generalizability and robustness to different types of distribution shifts, the developed ML models combine traditional methodologies with advanced techniques inspired by controllable artificial intelligence (AI), namely cautious classification, which gives the ML models the ability to abstain from making predictions, and explainable AI, which provides health operators with useful information about the models’ functioning.

Results

The developed models achieved good performance on the internal validation (area under the receiver operating characteristic curve between 0.91 and 0.98), as well as consistent generalization performance across the external validation datasets (area under the receiver operating characteristic curve between 0.75 and 0.95), outperforming baseline biomarkers and state-of-the-art ML models for sepsis detection. Controllable AI techniques were further able to improve performance and were used to derive an interpretable set of diagnostic rules.

Conclusions

Our findings demonstrate how controllable AI approaches based on complete blood count and MDW may be used for the early detection of sepsis while also demonstrating how the proposed methodology can be used to develop ML models that are more resistant to different types of data distribution shifts.

Subjects

MeSH terms (MEDLINE)

SepsisMachine LearningMonocytesHumansBlood Cell CountMaleAlgorithmsFemaleAgedMiddle AgedBiomarkers

Author keywords

artificial intelligencebiomarkerclinical signsclinical symptomscomplete blood countcontrollable AIdata distributiondetectiondevelopment studydiagnosticearly detectionexternal validationmachine learningmachine learning modelmedical machine learning

Substances

Biomarkers

Authors and institutions

  • IRCCS Ospedale Galeazzi Sant'Ambrogio, Milan, Italy.
  • University of Palermo, Palermo, Italy.
  • IRCCS San Raffaele Scientific Institute, Milano, Italy.
  • Department of Medicine, University of Padova, Padova, Italy.
  • Laboratory Medicine Unit, University-Hospital of Padova, Padova, Italy.
  • IRCCS Centro Di Riferimento Oncologico Aviano, Aviano, Italy.

ORCID profiles: Andrea Campagner, Luisa Agnello, Anna Carobene, Andrea Padoan, Fabio Del Ben, Massimo Locatelli, Mario Plebani, Agostino Ognibene, Maria Lorubbio, Elena De Vecchi, Andrea Cortegiani, Elisa Piva, Donatella Poz, Francesco Curcio, Federico Cabitza, Marcello Ciaccio

Cite

APA

Campagner, A., Agnello, L., Carobene, A., Padoan, A., Del Ben, F., Locatelli, M., Plebani, M., Ognibene, A., Lorubbio, M., De Vecchi, E., Cortegiani, A., Piva, E., Poz, D., Curcio, F., Cabitza, F., & Ciaccio, M. (2025). Complete Blood Count and Monocyte Distribution Width–Based Machine Learning Algorithms for Sepsis Detection: Multicentric Development and External Validation Study. Journal of Medical Internet Research. https://doi.org/10.2196/55492

Vancouver

Campagner A, Agnello L, Carobene A, Padoan A, Del Ben F, Locatelli M, et al. Complete Blood Count and Monocyte Distribution Width–Based Machine Learning Algorithms for Sepsis Detection: Multicentric Development and External Validation Study. Journal of Medical Internet Research. 2025. doi:10.2196/55492

BibTeX
@article{Campagner2025,
  author = {Campagner, Andrea and Agnello, Luisa and Carobene, Anna and Padoan, Andrea and Del Ben, Fabio and Locatelli, Massimo and Plebani, Mario and Ognibene, Agostino and Lorubbio, Maria and De Vecchi, Elena and Cortegiani, Andrea and Piva, Elisa and Poz, Donatella and Curcio, Francesco and Cabitza, Federico and Ciaccio, Marcello},
  title = {Complete Blood Count and Monocyte Distribution Width–Based Machine Learning Algorithms for Sepsis Detection: Multicentric Development and External Validation Study},
  journal = {Journal of Medical Internet Research},
  year = {2025},
  doi = {10.2196/55492}
}
Start for free →