Evaluating the Utility of Healthmap As a Supplementary Surveillance Tool

Monday, June 23, 2014: 10:30 AM
108, Nashville Convention Center
Melinda Thomas , Florida Department of Health, Tallahassee, FL
Aaron Kite-Powell , Massachusetts Institute for Technology (MIT) Lincoln Laboratories, Lexingtion, MA
David Atrubin , Florida Department of Health, Tampa, FL
Janet Hamilton , Florida Department of Health, Tallahassee, FL

BACKGROUND: The project assesses the outbreak detection utility of HealthMap, a publically available disease outbreak documentation system utilizing media resources to identify and track outbreaks, at the state and local level. This is accomplished using outbreak posts from EpiCom, the Florida Department of Health’s disease outbreak, health incident alerting, and health alert network. Results may help determine whether HealthMap should be monitored more closely as a supplementary surveillance tool and assist in the development of state guidance for utilization of EpiCom.

METHODS : This project compared EpiCom posts and HealthMap alerts in Florida for similarities in timing for outbreaks from January 1, 2009 to August 31, 2013. The project assessed sensitivity and timeliness of HealthMap, whether both sources captured the outbreak, and which source’s post was earlier. HealthMap post date was compared to EpiCom’s post date and date outbreak was reported to the county or state health department. Outbreaks of pertussis, legionellosis, salmonellosis, dengue, bacterial meningitis, measles, and influenza and influenza-like illness (ILI) were assessed. Results for measles are described below. The variation of EpiCom post timeliness by county size was also investigated.

RESULTS : During the study period, EpiCom reported on 13 confirmed measles outbreaks in Florida. Eight of these outbreaks were also found in HealthMap, giving HealthMap a sensitivity of 61.5% for measles. Two HealthMap measles posts based on non-U.S. media sources had no match to any EpiCom post. Based on post date, HealthMap was timelier than EpiCom for 2 of the outbreaks and just as timely for 3 of the outbreaks. However, based on the date the county health department was notified in the EpiCom posts, regular state and local surveillance detection practices are timelier for 100% of the outbreaks.

CONCLUSIONS : Preliminary analysis suggests that HealthMap is useful for surveillance activity but not for initial outbreak detection by the state and local health department due to its relatively low sensitivity and timeliness of detection. However, due to its worldwide focus, it may be useful in providing a better international view of disease activity, which could be especially valuable for Florida for situational awareness and surveillance activity due to the large amount of international travelers that visit. Additionally, there was evidence suggesting public health is being too conservative in its utilization of EpiCom. HealthMap can also be a means of informing the general public of health concerns in their area and assisting public health in recognizing what events are in the news.