Syndromic Surveillance through Measuring Lexical Shift in Emergency Department Chief Complaint Texts

Syndromic Surveillance through Measuring Lexical Shift in Emergency Department Chief Complaint Texts Hafsah Aamer, Bahadorreza Ofoghi, Karin Verspoor Department of Computing and Information Systems The University of Melbourne Parkville Victoria 3010, Australia haamer@student.unimelb.edu.au bahadorreza.ofoghi@unimelb.edu.au karin.verspoor@dunimelb.edu.au Abstract Syndromic Surveillance has been per- formed using machine learning and other statistical methods to detect disease out- breaks. These methods are largely depen- dent on the availability of historical data to train the machine learning-based surveil- lance system. However, relevant train- ing data may differ from region to region due to geographical and seasonal trends, meaning that the syndromic surveillance designed for one area may not be effec- tive for another. We proposed and analyse a semi-supervised method for syndromic surveillance from emergency department chief complaint textual notes that avoids the need for large training data. Our new method is based on identification of lexi- cal shifts in the language of Chief Com- plaints of patients, as recorded by triage nurses, that we believe can be used to monitor disease distributions and possible outbreaks over time. The results we ob- tained demonstrate that effective lexical syndromic surveillance can be approached when distinctive lexical items are available to describe specific syndromes. 1 Introduction The increase in new emerging pathogenic diseases like SAARS, Ebola, and the Zika virus requires an ongoing effective syndromic surveillance sys- tem. A syndromic surveillance system keeps track of the frequency of patients experiencing specific syndromes over time. Any abnormality in the nor- mal trend of syndromes with respect to time may imply a proximal disease outbreak. There are many data sources that can be used to perform syndromic surveillance, such as data from hospital emergency departments. Chief com- plaints provide us with one such rich data source to perform syndromic surveillance. A chief com- plaint is the set of signs and symptoms that the triage nurse registers upon a patient’s arrival at the emergency department of a hospital. The symp- toms are usually described in a terse, ungrammat- ical, and abbreviated language. Once the chief complaint is registered at the initial assessment, the patient sees the doctor and receives a detailed check up. This may include sending the patient’s specimen for testing in the laboratory especially when dealing with a potentially dangerous infec- tious disease. The results from the laboratory tests may take days before the end result is available and by then, there is a large risk of spreading the disease to other people who may come in contact with the infected person. Chief complaints are readily available in a dig- ital format and can therefore be easily processed using natural language processing algorithms. In the past, various work has been done to perform syndromic surveillance using supervised machine learning and statistical algorithms (Tsui et al., 2003; Espino et al., 2007; Chapman et al., 2005; Bradley et al., 2005). The major drawback of ma- chine learning methods is the requirement of his- torical data that can be used to train the system, and a sensitivity to the characteristics of specific …

Content truncated to 500 words