A framework for infection control surveillance using association rules

Area of Science:

• Infection Control and Hospital Epidemiology
• Data Mining and Machine Learning in Healthcare
• Public Health Surveillance

Background:

• Effective surveillance of antibiotic resistance and hospital-acquired infections is crucial for hospital infection control programs.
• Traditional surveillance methods may not efficiently detect novel or low-prevalence outbreaks.
• Identifying unexpected patterns is key to proactive infection management.

Purpose of the Study:

• To develop and evaluate an automated method using association rule mining for detecting new and unexpected patterns in hospital infection control data.
• To hypothesize that mining for low-support, low-confidence rules can identify outbreaks involving a small number of cases.
• To assess the efficiency and effectiveness of the proposed framework in identifying significant patterns.

Main Methods:

• Employed association rule mining, a data mining technique, to analyze hospital culture data.
• Preprocessed data and developed new templates to filter out uninteresting patterns.
• Applied the method to 3 months of culture data from 10 hospitals within the UPMC Health System.

Main Results:

• The developed process and system demonstrated efficiency and effectiveness in identifying new, unexpected, and potentially interesting patterns within surveillance data.
• The association rule mining approach successfully highlighted patterns that might be missed by conventional surveillance.
• The method proved capable of handling large datasets from multiple hospital sites.

Conclusions:

• The novel data mining approach shows promise for enhancing hospital infection control surveillance by detecting subtle and unexpected patterns.
• The efficiency of the system in identifying potentially significant patterns was confirmed.
• Further prospective studies are needed to fully establish the clinical relevance and utility of this automated surveillance process.