Primary ocular infections are not regularly notifiable to the US Center for Disease Control. Infectious ocular disease epidemics are not regularly reported, tracked, or predicted in the US, despite the fact that tracking would be of clear benefit to the eye health community. We hypothesize that several novel digital surveillance approaches can improve monitoring and detection of ocular disease epidemics, starting with one or the more common model eye infection: conjunctivitis. Severe outbreaks of conjunctivitis caused by adenovirus (epidemic keratoconjunctivitis or EKC) or bacteria occur without warning and sporadically in location and time. Epidemics are highly contagious with a significant burden of cost in the US, although long-term vision impairment is unusual. Risk factors for these and other ocular disease epidemics may be on the rise, and budgetary cuts may further reduce funding for traditional surveillance reporting and prevention. Thus new solutions are needed for monitoring ocular disease epidemics. Recently web searches, social media, and other sourced digital big data has been used to track infectious diseases. Here, we propose (Aim1) to identify candidate ocular disease epidemics using big data through a number of statistical approaches, as well as to (Aim2) use medical records and also develop a digital sentinel network of specialists for participatory surveillance, in order to validate any candidate epidemics, and (Aim3) Since even the health records and other clinical reports in Aim2 are not a perfect gold standard, we will also use a hidden state model to estimate, for each methodology, the sensitivity and specificity (and other appropriate descriptors) of detection of a true conjunctivitis epidemic.
Conjunctivitis is common with a significant burden of cost in the US, often causes discomfort, can occasionally also cause long-term vision impairment, and severe and highly contagious outbreaks can occur without warning. Delayed epidemic detection can increase the impact and societal burden. Successful completion of the project aims to use Big Data (Google web search, Twitter, and website log data) to improve tracking and detection of these ocular disease epidemics and validate them using clinical EMR data can increase awareness and curtail spread, reducing the burden on the US economy and human ocular health.
|Berlinberg, Elyse J; Deiner, Michael S; Porco, Travis C et al. (2018) Monitoring Interest in Herpes Zoster Vaccination: Analysis of Google Search Data. JMIR Public Health Surveill 4:e10180|
|Deiner, Michael S; McLeod, Stephen D; Chodosh, James et al. (2018) Clinical Age-Specific Seasonal Conjunctivitis Patterns and Their Online Detection in Twitter, Blog, Forum, and Comment Social Media Posts. Invest Ophthalmol Vis Sci 59:910-920|
|Sié, Ali; Diarra, Abdramane; Millogo, Ourohiré et al. (2018) Seasonal and Temporal Trends in Childhood Conjunctivitis in Burkina Faso. Am J Trop Med Hyg 99:229-232|
|Ramirez, David A; Porco, Travis C; Lietman, Thomas M et al. (2018) Ocular Injury in United States Emergency Departments: Seasonality and Annual Trends Estimated from a Nationally Representative Dataset. Am J Ophthalmol 191:149-155|
|Ramirez, David A; Porco, Travis C; Lietman, Thomas M et al. (2017) Epidemiology of Conjunctivitis in US Emergency Departments. JAMA Ophthalmol 135:1119-1121|
|Deiner, Michael S; Fathy, Cherie; Kim, Jessica et al. (2017) Facebook and Twitter vaccine sentiment in response to measles outbreaks. Health Informatics J :1460458217740723|
|Deiner, Michael S; Lietman, Thomas M; Porco, Travis C (2017) Uncertainties in Big Data When Using Internet Surveillance Tools and Social Media for Determining Patterns in Disease Incidence-Reply. JAMA Ophthalmol 135:402-403|
|Deiner, Michael S; Lietman, Thomas M; McLeod, Stephen D et al. (2016) Surveillance Tools Emerging From Search Engines and Social Media Data for Determining Eye Disease Patterns. JAMA Ophthalmol 134:1024-30|