Issues in Building a Nursing Home Syndromic Surveillance System with Textmining: Longitudinal Observational Study

Area of Science:

• Public Health Surveillance
• Geriatric Medicine
• Epidemiology

Background:

• Nursing home data warehouses enable daily health monitoring of the elderly population.
• Syndromic surveillance using professional data is effective for public health but lacks long-term cohort follow-up.
• Electronic health records offer a rich source for real-time health monitoring.

Purpose of the Study:

• To develop and evaluate a national ecological public health surveillance system for nursing homes.
• To establish a robust system for monitoring the health of senior citizens in nursing homes.
• To assess the feasibility of using syndromic data for long-term health surveillance in this population.

Main Methods:

• A national network of 126 nursing homes contributed resident data.
• Syndromic information, including sociodemographic and clinical data, was extracted from electronic health records.
• Syndromes were defined using pattern matching and Delphi techniques; surveillance algorithms (EARS, Bayes) were employed for analysis.

Main Results:

• A cohort of 41,061 senior citizens was established.
• The EARS_C3 algorithm demonstrated good sensitivity and specificity for detecting influenza and acute gastroenteritis.
• High precision rates for syndromic data were observed during both peak and low epidemic periods.

Conclusions:

• Syndromic information is a valuable tool for developing a national public health surveillance system for senior citizens.
• Real-time access to validated health data on influenza and acute gastroenteritis addresses a critical public health need for this vulnerable population.
• The established database facilitates future ecological research to enhance prevention, care, and rapid response to health threats.