Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Time series modeling for syndromic surveillance.

Ben Y Reis1, Kenneth D Mandl

  • 1Children's Hospital Informatics Program, Boston, Massachusetts, USA. reis@mit.edu

BMC Medical Informatics and Decision Making
|January 25, 2003
PubMed
Summary

This study presents a new methodology for predicting emergency department (ED) visit rates using time-series analysis. The developed models accurately forecast patient visits, aiding in early detection of public health threats through syndromic surveillance.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Design principles for integrated AI alignment.

Patterns (New York, N.Y.)·2026
Same author

Development and validation of a near-comprehensive RxNorm valueset of opioid medications.

JAMIA open·2026
Same author

What do LLMs value? An evaluation framework for revealing subjective trade-offs in assessment of glycemic control.

Proceedings of machine learning research·2026
Same author

Trends in Suicide Mortality by Method among US Individuals aged 10-24 Years from 1999 to 2024.

medRxiv : the preprint server for health sciences·2026
Same author

Embeddings of clinical codes enable knowledge-grounded AI in medicine.

NPJ digital medicine·2026
Same author

Learning Normal Representations for Blood Biomarkers.

ArXiv·2026

Area of Science:

  • Public Health
  • Epidemiology
  • Data Science

Background:

  • Emergency department (ED) syndromic surveillance systems are crucial for detecting bioterrorism events by identifying unusual visit rates.
  • Establishing expected ED visit rates is essential for reliable anomaly detection, but systematic methods are lacking.

Purpose of the Study:

  • To develop and validate a generalized methodology for creating models of expected ED visit rates.
  • To improve the accuracy of syndromic surveillance by understanding normal healthcare usage patterns.

Main Methods:

  • Utilized time-series methods, including trimmed-mean seasonal models and autoregressive integrated moving average (ARIMA) models, on a decade of historical pediatric ED data.
  • Developed models for both overall and respiratory-related ED visits based on chief complaints.

Related Experiment Videos

  • Tested model detection capabilities using simulated outbreaks.

    • Main Results:

    • ARIMA models achieved a mean absolute percentage error of 9.37% for overall visits and 27.54% for respiratory visits.
    • A detection system based on overall visit ARIMA models demonstrated high sensitivity (100%) and specificity (97%) for simulated outbreaks of 30 visits/day.
    • Sensitivity decreased with smaller simulated outbreaks (57% for 10 visits/day), while maintaining 97% specificity.

    Conclusions:

    • Time-series analysis of historical ED data is effective for syndromic surveillance and accurate forecasting of ED utilization.
    • The developed models account for long-term and recent trends, offering a comprehensive approach for public health monitoring.
    • The methodology is generalizable to other healthcare settings for automated anomaly detection in disease patterns and healthcare usage.