Seattle Fire Department Harold Scoggins, Fire Chief
About Us|Contact Us
Seattle Fire Department Web Site Header bottom graphic   Home > Statistics > Infectious Disease Monitoring > Situation Found Surveillance Program
Department Information
Public Education
Fire Prevention
Medic One
- Reports and Statistics
- Search for Incidents
- Permit Status Search
- Infectious Disease Reports
Photo Gallery
Site Map
Infectious Disease Monitoring
Last Updated: August 13, 2010


The Seattle Fire Department tracks symptoms of patients that firefighters come into contact with that may be related to flu: cough, fever, gastro, rash, respiratory.  Reports of these symptoms are entered into a computer system built by the Fire Department shortly after firefighters leave the scene of the incident.  The data is monitored by the First Watch surveillance application (

The Sit Found program runs on mobile computers in all fire apparatus and on all computers in the fire stations.  The program is a Microsoft C#.NET application that combines CAD data with flu symptom data.  By tying the flu symptom and incident data together the Department can easily match infectious disease information with dispatch protocols and medical conditions and patient profile information.  This allows the Department to analyze trends and support the Department's infectious disease response program.


First Watch captures real-time data from the Sit Found database and the Seattle Fire Department (SFD) CAD system.  SFD and First Watch established triggers to notify a group of the SFD members when flu symptoms exceeded pre-defined thresholds.  The thresholds are based on historical trends using the four analytical methods listed below.

1)      Actual Events (STD)- actual events over a 24 hour period compared to the 14-day moving average of events. In other words, the count of events for hour X are compared to the moving average count for that hour.

2)      Syndrome to All (STA)– ratio of qualified events as compared against all events in the system – measured by hour of day, and day of week.

3)      Cumulative Summary (CUSUM) – a moving 14 day average that alerts when the cumulated observed counts exceeded the expected count by the predetermined threshold.

4)      Geo Cluster Analysis – when 5 calls fell within a 1 mile radius within a 12 hour period.

The first three – STD, STA, CUSUM – are applied to a trigger for Fever and Any Combination of Symptoms (fever-cough, cough-gastro, etc.)  If any of those measures has an event exceeding 2 Standard Deviations form the threshold an alert will be sent.

The geo-cluster trigger is applied to any symptom.

The alerts are e-mail messages that contain charts showing when the event exceeded a threshold, a map of the location of the event or events, and details about each incident that caused the alert: incident number, company, symptoms found, address, age, sex of patient.

The alert also includes a hot link to the First Watch web site so the person receiving the alert could go to web site and get more details about the alert.  The hot link is only good for 2 hours after the alert is sent – for security reasons.  All people receiving alerts have accounts with First Watch and can access the Seattle specific First Watch site at any time.

Below are sample charts, with more descriptive information about the chart, and a map from a typical alert message:

The Actual Events chart analyzes the number of events over a period of time and includes a historical average line, actual events line, trigger threshold line (can include hi or low) and a rolling count of events that updates every few minutes.  Actual Events chart also monitors for variances in values as events are changed or modified.  Measured by day of week and hour of day.

The Syndrome-To-All graph presents a ratio of qualified events within the data set compared to all events in the system during the Analysis Period.  This approach assists in qualifying volumetric increases as being system-wide or specific to a trigger’s data set. Measured by hour of day and day of week.

CUSUM is a time-series analysis method utilizing a moving average for the detection of temporal clusters looking at day-to-day variability and is designed to detect sudden changes in the mean value of a quantity of interest.  Cumulative sum methods cumulate the deviations between observed and expected counts in a period, resulting in an alert when the cumulated observed counts have exceeded the expected counts by some predetermined threshold. The FirstWatch application of CUSUM methodology was implemented using input from state level epidemiologists and CDC resources



Last Modified:   August 13, 2010
Seattle Fire Department Home | About Us | Contact Us | Related Links | Employment | Site Map