An Automated Syphilis Serology Record Search and Review Algorithm to Prioritize Investigations by Health Departments

Area of Science:

• Public Health
• Infectious Disease Surveillance
• Medical Informatics

Background:

• Reactive syphilis serologies require investigation to identify new infections, reinfections, or treatment failures.
• Current investigations rely on manual record searches and reviews, prioritized by nontreponemal test titer and age using a reactor grid.
• Manual review is time-consuming and potentially inefficient for public health departments.

Purpose of the Study:

• To develop and validate a computerized algorithm for automating the search and review of records for reactive syphilis serologies.
• To compare the efficiency and accuracy of the automated algorithm against the traditional reactor grid method.

Main Methods:

• Developed a computerized algorithm based on the syphilis case definition, including criteria for new infections and significant titer increases.
• Incorporated additional steps to enhance case detection and avoid missing potential syphilis cases.
• Tested the algorithm using a Florida Department of Health dataset and validated it with the New York City Department of Health and Mental Hygiene.

Main Results:

• The algorithm closed 49.9% of investigations compared to 27.0% by the reactor grid in Florida.
• The algorithm identified 99.4% of individuals classified as syphilis cases by health departments.
• In New York City, the algorithm closed 70.9% of investigations, demonstrating higher efficiency with more available historical data.

Conclusions:

• The automated algorithm effectively searches and reviews records, aiding in the identification of syphilis cases.
• The algorithm demonstrated significant potential for time savings, estimated at 590 workdays over 3 years in Florida.
• Automation of syphilis serology investigation can improve public health surveillance efficiency and resource allocation.