Skip to main content

Syndrome classification

Biosurveillance systems commonly use emergency department (ED) patient chief complaint data (CC) for surveillance of influenza-like illness (ILI). Daily volumes are tracked using a computerized patient CC classifier for fever (CC Fever) to identify febrile patients. Limitations in this method have led to efforts to identify other sources of ED data. At many EDs the triage nurse measures the patient’s temperature on arrival and records it in the electronic medical record. This makes it possible to directly identify patients who meet the CDC temperature criteria for ILI: temperature greater than 100 degrees F (T>100F).

Objective

To evaluate whether a classifier based on temperature >100F would perform similarly to CC Fever and might identify additional patients.

Referenced File
Submitted by hparton on

Emerging event detection is the process of automatically identifying novel and emerging ideas from text with minimal human intervention. With the rise of social networks like Twitter, topic detection has begun leveraging measures of user influence to identify emerging events. Twitter's highly skewed follower/followee structure lends itself to an intuitive model of influence, yet in a context like the Emerging Infections Network (EIN), a sentinel surveillance listserv of over 1400 infectious disease experts, developing a useful model of authority becomes less clear. Who should we listen to on the EIN? To explore this, we annotated a body of important EIN discussions and tested how well 3 models of user authority performed in identifying those discussions. In previous work we proposed a process by which only posts that are based on specific "important" topics are read, thus drastically reducing the amount of posts that need to be read. The process works by finding a set of "bellwether" users that act as indicators for "important" topics and only posts relating to these topics are then read. This approach does not consider the text of messages, only the patterns of user participation. Our text analysis approach follows that of Cataldi et al.[1], using the idea of semantic "energy" to identify emerging topics within Twitter posts. Authority is calculated via PageRank and used to weight each author's contribution to the semantic energy of all terms occurring in within some interval ti. A decay parameter d defines the impact of prior time steps on the current interval.

Objective

To explore how different models of user influence or authority perform when detecting emerging events within a small-scale community of infectious disease experts.

Referenced File
Submitted by elamb on

In 2004, the BioDefend (BD) syndromic surveillance (SS) system was implemented in Duval County hospitals (Jacksonville, FL). Daily emergency department chief complaints are manually classified and entered into the BD system by triage personnel. As part of a statewide implementation, the Electronic Surveillance System for the Early Notification of Community-based Epidemics (ESSENCE) began collecting data in the Jacksonville area during the winter of 2007-08. ESSENCE uses an automated data collection, chief complaint parsing and analysis process for data management and analysis. The use of two systems during the same period of time in one area provided a unique opportunity to retrospectively analyze characteristics of the BD and ESSENCE systems.

 

Objective

To compare detection of a community outbreak of influenza-like illness using two SS systems, one using a clinician’s classification of reason for visit and the other using an automated chief complaint parsing algorithm.

Referenced File
Submitted by elamb on

One of the first county-wide syndromic surveillance systems in the nation, the Syndromic Tracking and Reporting System (STARS) has been in operation since 11/01/2001, and now covers Hillsborough, Pinellas and Collier counties. STARS uses hospital emergency department visit data to detect aberrations of non-specific syndromes and serves as an earlier warning system for public health threats. Patient’s syndrome is collected upon arrival, separately from routine collection of clinical and administrative data; but in some hospitals the process is being streamlined with routine data collection. Aberration detection is done twice daily using the statistical system EARS developed by the CDC. Upon flagging of an aberration, follow-up investigation is conducted to verify cases, and identify source of exposure following a sequence of decision procedure. After several years of operation and some instituted enhancements, a systematic evaluation was called to (1) assess if STARS has met the operation specifications and (2) characterize system efficacy and effectiveness.

 

Objective

To evaluate STARS with respect to quality of syndrome diagnoses, timeliness and completeness of data collection and processing, performance of aberration detection methods, and aberration investigation.

Referenced File
Submitted by elamb on

TOA identifies clusters of patients arriving to a hospital ED within a short temporal interval. Past implementations have been restricted to records of patients with a specific type of complaint. The Florida Department of Health uses TOA at the county level for multiple subsyndromes (1). In 2011, NC DPH, CCHI and CDC collaborated to enhance and evaluate this capability for NC DETECT, using NC DETECT data in BioSense 1.0 (2). After this successful evaluation based on exposure complaints, discussions were held to determine the best approach to implement this new algorithm into the production environment for NC DETECT. NC DPH was particularly interested in determining if TOA could be used for identifying clusters of ED visits not filtered by any syndrome or sub-syndrome. In other words, can TOA detect a cluster of ED visits relating to a public health event, even if symptoms from that event are not characterized by a predefined syndrome grouping? Syndromes are continuously added to NC DETECT but a syndrome cannot be created for every potential event of public health concern. This TOA approach is the first attempt to address this issue in NC DETECT. The initial goal is to identify clusters of related ED visits whose keywords, signs and/or symptoms are NOT all expressed by a traditional syndrome, e.g. rash, gastrointestinal, and flu-like illnesses. The goal instead is to identify clusters resulting from specific events or exposures regardless of how patients present – event concepts that are too numerous to pre-classify.

Objective:

To describe a collaboration with the Johns Hopkins Applied Physics Laboratory (JHU APL), the North Carolina Division of Public Health (NC DPH), and the UNC Department of Emergency Medicine Carolina Center for Health Informatics (CCHI) to implement time-of-arrival analysis (TOA) for hospital emergency department (ED) data in NC DETECT to identify clusters of ED visits for which there is no pre-defined syndrome or sub-syndrome.

 

Submitted by Magou on

The CDC recently developed sub-syndromes for classifying disease to enhance syndromic surveillance of natural outbreaks and bioterrorism. They have developed ICD9 classifiers for six GI Illness subsyndromes: Abdominal Pain, Nausea and Vomiting, Diarrhea, Anorexia, Intestinal infections, and Food poisoning. If the number of visits for sub-syndromes varies significantly by age it may impact the design of outbreak detection methods.

 

Objective

We hypothesized that the percentage of visits for the GI sub-syndromes varied significantly with age.

Referenced File
Submitted by elamb on

Abbreviation, misspellings, and site specific terminology may misclassify chief complaints syndromes. The Emergency Medical Text Processor (EMT-P) is system that cleans emergency department chief complaints and returns standard terms. However, little information is available on the implementation of EMT-P in a syndromic surveillance system.

 

Objective

To describe the implementation and baseline evaluation of EMT-P developed by the University of North Carolina.

Referenced File
Submitted by elamb on

In February of 2007, the Bureau of Epidemiology (BOE) received a request from Houston Department of Public Works to investigate a possible rise in gastrointestinal (GI) illness associated with complaints about poor water quality in a Northeastern Houston neighborhood. To investigate this complaint, BOE combined case report data with syndromic data from our Real-Time Outbreak Disease Surveillance (RODS). The Houston RODS collects and synthesizes real-time chief complaint data from 34 area hospitals and health facilities, representing approximately 70% coverage of licensed ER beds in Harris County. The system uses a Naïve Bayes Classifier to categorize ER chief complaints into 7 different syndromes, including GI illness.

 

Objective

To investigate public concern over a possible increase in GI illness associated with water quality complaints in Northeast Houston.

Referenced File
Submitted by elamb on

The North Carolina Bioterrorism and Emerging Infection Prevention System (NC BEIPS) receives daily emergency department (ED) data from 33 (29%) of the 114 EDs in North Carolina. These data are available via a Web-based portal and the Early Aberration Reporting System to authorized NC public health users for the purpose of syndromic surveillance (SS). Users currently monitor several syndromes including: gastrointestinal severe, fever/rash illness and influenza-like illness. The syndrome definitions are based on the infection-related syndrome definitions of the CDC and search the chief complaint (CC) and, when available, triage note (TN) and initial temperature fields. Some EDs record a TN, which is a brief text passage that describes the CC in more detail. Most research on the utility of ED data for SS has focused on the use of CC. The goal of this study was to determine the sensitivity, specificity, and both positive and negative predictive value of including TN in the syndrome queries.

 

Objective

This study evaluates the addition of TN to syndrome queries used in the NC BEIPS.

Referenced File
Submitted by elamb on

Of critical importance to the success of syndromic surveillance systems is the ability to collect data in a timely manner and thus ensure rapid detection of disease outbreaks. Most emergency department-based syndromic surveillance systems use information rou-tinely collected in patient care including patient chief complaints and physician diagnostic coding. These sources of data have been shown to have only limited sensitivities for the identification of cer-tain syndromes. Another potential source of information, which has not been previously studied, is the patient. Studies have shown that patients as well as parents can accurately report information about their own or their child’s illness. The value of of patient and parent self-reported informa-tion for disease surveillance systems has not been measured.

 

Objective

To determine whether patients and their families can directly provide medical information that enables syndrome classification.

Referenced File
Submitted by elamb on
Subscribe to Syndrome classification