Collaborative infectious disease surveillance using emergency medical services response (ambulance run) data

A collaborative data-sharing effort between Orange County Epidemiology & Assessment (E&A) and Orange County Fire Authority (OCFA) has been established to evaluate the usefulness of emergency medical services (EMS) response (ambulance run) data for infectious disease surveillance. This data is collected in a way that allows for grouping into syndromic categories. Because recent data can be delayed in electronic entry, data is provided weekly via auto-generated email for the one to eight weeks prior.

We conducted time-series cross-correlation analyses using SPSS 15.0 to compare recent (incomplete) OCFA data to lagged (complete) OCFA data, by CDC disease week. Data from the 2006-2007 winter season for the influenza-like illness (ILI) syndrome category was used for analysis.

Recent data (1 week prior) was on average 75% complete compared to complete data (8 weeks prior). Recent data had a high correlation coefficient (0.835) with complete data; this finding was statistically significant (p=0.003). Less recent data (2 and 3 weeks prior) had higher correlation coefficients (0.988 and 0.996, respectively) with complete data.

Although less recent weeks were more strongly correlated, correlation for the most recent data was sufficient as a predictor of complete data trends. Establishing the ability of incomplete data to predict trends in the complete data helps to determine its usefulness for timely monitoring of infectious disease activity during emergency periods and supports continued data-sharing efforts between E&A and OCFA. Direct collaboration between these agencies helped evaluate a potentially useful data source and strengthened relationships for potential coordinated response to future emergencies.