Related Experiment Video
Updated: May 10, 2025
12:08
Hybrid De Novo Genome Assembly for the Generation of Complete Genomes of Urinary Bacteria using Short- and Long-read Sequencing Technologies
Published on: August 20, 2021
4.9K
Validation of an Algorithm to Classify Urine Cultures in Family Medicine
Jack Zhang1, Rachael Morkem2, Akshay Rajaram2,3
1Section of Anesthesia, Northern Ontario School of Medicine University, Sudbury, Canada.
Applied Clinical Informatics
|April 23, 2025
Summary
The MicrobEx algorithm accurately classifies urine culture reports, demonstrating high performance in primary care settings. This automation shows promise for reducing clinician workload in managing test follow-ups.
Area of Science:
- Medical Informatics
- Clinical Pathology
- Bioinformatics
Background:
- Automating the classification of urine culture reports can significantly reduce the workload for clinicians.
- Effective management of test follow-ups is crucial for patient care and efficient healthcare delivery.
Purpose of the Study:
- To evaluate the performance of MicrobEx, a regular expression-based algorithm, in classifying urine culture reports within a primary care context.
- To assess the diagnostic accuracy of MicrobEx compared to manual review.
Main Methods:
- A retrospective validation study was conducted using 1,999 urine culture reports from an academic family health team.
- MicrobEx classifications were compared against human reviewer labels to calculate diagnostic performance metrics.
Main Results:
- MicrobEx achieved an overall accuracy of 95.3%.
- The algorithm demonstrated high sensitivity (88.6%) and perfect specificity (100%) in classifying the reports.
Conclusions:
- The performance of MicrobEx in this primary care setting was comparable to its original validation.
- Further research is needed to enhance MicrobEx accuracy and evaluate its utility across diverse primary care environments and complex cases.
