Related Experiment Video
Updated: Aug 8, 2026
A Semi-automated Approach to Preparing Antibody Cocktails for Immunophenotypic Analysis of Human Peripheral Blood
Published on: February 8, 2016
Automated detection of intravenous fluid contamination using longitudinal routine laboratory data: development and
Jordi Tortosa-Carreres1, Nahúm Martí-Montoro1, Jonnathan Andrés Acevedo-Galvis1
1Laboratory Department, Hospital Universitari i Politècnic La Fe. Av. Fernando Abril Martorell, 106, 46026 València, Spain; Health Research Institute Hospital La Fe (IISLaFe), Valencia, Spain; Clinical Laboratory Medicine Research Group.
This study developed an automated framework for intravenous fluid contamination (IFC) screening using lab data. The tool accurately identifies contaminated samples, showing potential for improved patient safety.
Area of Science:
- Clinical Chemistry
- Medical Informatics
- Artificial Intelligence in Medicine
Background:
- Intravenous fluid contamination (IFC) poses a risk to patient safety.
- Current screening methods may be labor-intensive or lack automation.
- Longitudinal laboratory data offers a potential source for automated screening.
Purpose of the Study:
- To develop and externally validate an automated framework for IFC screening.
- Utilize routinely available longitudinal laboratory data for model development.
- Assess the framework's performance in identifying contamination probability, type, and severity.
Main Methods:
- A bicentric retrospective study using internal and external validation cohorts.
- Calculation of normalized relative change variables from paired laboratory measurements.
- Development of a multitask neural network for contamination detection and severity estimation.
Main Results:
- High discriminative performance with ROC-AUC 0.94-0.96 and sensitivities/specificities ≥85%.
- Accurate identification of contamination probability, type, and severity.
- The framework correctly identified 97% of real-world suspected IFC cases and reduced unnecessary alerts by 17%.
Conclusions:
- The developed automated framework demonstrates significant potential for supporting IFC screening.
- The tool can accurately identify contaminated intravenous fluids using existing laboratory data.
- This approach may enhance patient safety and optimize laboratory workflows.
