English Roseanne

Biosurveillance systems typically receive free- text chief complaint and coded diagnosis data, however this data has limited specificity for notifiable disease surveillance. The Biosense System receives chief complaint and/or diagnosis data from over 360 hospitals and laboratory results from 24 hospitals in 7 states using the Public Health Information Network Messaging System (PHINMS) and HL7 standards. BioSense also receives final diagnosis from Veterans’ Affairs and Department of Defense outpatient clinics, but these clinics do not currently report laboratory findings. Chief complaints and diagnoses are assigned, as appropriate, to 11 syndromes (e.g., Gastrointestinal [GI]) (1) and to 78 more granular categories termed sub-syndromes (e.g., abdominal pain, nausea and vomiting, diarrhea) Surveillance for Salmonella infection is important since this agent is both a commonly- reported notifiable disease and a Category B bioterrorist agent.

Objective

To describe visits reported from BioSense hospitals with non-typhoidal Salmonella infections.

Analysis of time series data requires accurate calculation of a predicted value. Non-regression methods such as the Early Aberration Reporting System CuSum are computationally simple, but most do not adjust for day of week or holiday. Alternately, regression methods require larger counts, more computer resources, and possibly longer baseline periods of data. As increasing volumes of data are reported and analyzed, the predictive accuracy of simpler methods should be assessed and optimized.

Objective

To compare the predictive accuracy of three non-regression methods in analysis of time series count data.

BioSense is a national system that receives, analyzes, and visualizes electronic health data and makes it available for public health use. In December 2007 CDC added the Influenza Module to the main BioSense application.

Objective

This presentation describes the new BioSense Influenza Module, its performance during the 2007-8 influenza season, and modifications for the 2008-9 influenza season.

The eleven syndrome classifications for clinical data records monitored by BioSense include rare events such as death or lymphadenitis and also common occurrences such as respiratory infections. BioSense currently uses two statistical methods for prediction and alerting with respect to the eleven syndromes. These are a modified CUSUM; and small area regression and testing (SMART), described by Ken Kleinman. At the inception of BioSense, these prediction methods were implemented as one-model-fits-all, and they remain largely unmodified. An evaluation of the predictive value of these methods is required. The SMART method, as used in BioSense, uses long-term data. As covariate predictors, day-of-week, a holiday indicator, day after holiday, and sine/cosine seasonality variables are used. Lengthy, stable historical data is not always available in BioSense data sources, and this obstacle is expected to grow as data sources are added. We wish to test regression methods of surveillance that use shorter time periods, and different sets of predictors.

Objective

This paper compares the prediction accuracy of regression models with different covariates and baseline periods, using a subset of data from CDC’s BioSense initiative. Accurate predictions are needed to achieve sensitivity at practical false alarm rates in anomaly detection for biosurveillance.

West Nile Virus (WNV) is a mosquito-borne virus that can cause meningitis and encephalitis. Since its discovery in New York City during an encephalitis outbreak in 1999, WNV has become endemic in North America. In the United States, 16,000 human WNV disease cases (including West Nile fever, meningitis, encephalitis, and unspecified clinical illness) and over 600 WNV-related deaths have been reported to the Centers for Disease Control from 46 states. Perennial WNV epidemics occur during summer months, peaking during late August. BioSense Early Event Detection and Situation Awareness System receives daily laboratory test order data feed in HL7 from Laboratory Corporation of America. In this study, test orders were studied for their correlation with WNV activity.

Objective

To determine the feasibility of using BioSense laboratory test order data for West Nile disease surveillance in the United States. 

In 2007, the CDC BioSense System received data from 450 non-federal hospitals. Hospitals provide data to Biosense based on their capability and willingness to supply electronic data. As of July 2008, Biosense is receiving data from 550 hospitals. The National Hospital Ambulatory Medical Care Survey (NHAMCS) is an annual national probability sample survey of hospitals that collects data on U.S. emergency department (ED) visits.

Objective

To assess the representativeness of BioSense ED data by comparing it with the NHAMCS results.

BioSense currently receives demographic and chief complaint data from more than 360 hospitals and text radiology reports from 36 hospitals. Detection of pneumonia is an important as several Category A bioterrorism diseases as well as avian influenza can manifest as pneumonia. Radiology text reports are often received within 1-2 days and may provide a faster way to identify pneumonia than coded diagnoses. Objective To study the performance of a simple keyword search of radiology reports for identifying pneumonia.