2014.ISDS.Abstracts.Final.pdf


ISDS Annual Conference Proceedings 2014. This is an Open Access article distributed under the terms of the Creative Commons Attribution-
Noncommercial 3.0 Unported License (http://creativecommons.org/licenses/by-nc/3.0/), permitting all non-commercial use, distribution, 
and reproduction in any medium, provided the original work is properly cited.

ISDS 2014 Conference Abstracts

Ratio of Excess ED ILI Visits to Seroprevalence, 
Influenza A/H1N1 Infection, FL, 2009
Richard Hopkins*2, 1, Aaron Kite-Powell1, Kate Goodin1 and Janet J. Hamilton1

1Epidemiology Bureau, Florida Department of Health,, Tallahassee, FL, USA; 2University of Florida, Gainesville, FL, USA

Objective
To estimate the number of infections due to the novel 2009 

influenza A/H1N1 virus corresponding to each ED visit for ILI in a 
four-county area of Florida. Knowing such ratios, one could (in future 
similar situations) estimate the cumulative number of infections due 
to a novel influenza virus in a population.

Introduction
A seroprevalence survey carried out in four counties in the Tampa 

Bay area of Florida (Hillsborough, Pinellas, Manatee and Pasco) 
provided an estimate of cumulative incidence of infection due to 
the 2009 influenza A (H1N1) as of the end of that year’s pandemic 
(1). During the pandemic, high-level decison-makers wanted timely, 
credible forecasts as to the likely near-term course of the pandemic. 
The cumulative percentage of people who will be infected by the 
end of the epidemic can be estimated from the intrinsic reproductive 
number of the viral strain, its R0, which can be measured early in the 
epidemic (2). If the current cumulative number of infections can be 
estimated, then one can determine what fraction of the eventual total 
number of infected people have already been infected.

Methods
Excess emergency department (ED) visits for influenza-like 

illness (ILI) during the pandemic period (compared to the four non-
pandemic years 2007, 2008, 2010, and 2011) were estimated using 
the ESSENCE-FL syndromic surveillance system for the four-county 
area. For each week, the percentage of visits that were for ILI, by 
age group, was compared between 2009 and the average of the other 
four years. The difference, the excess in percent ILI, was then applied 
to the observed number of ILI visits for each week and age group 
to calculate the excess number of visits. These excess visits were 
summed for the epidemic period, and compared to the estimated 
number of infected residents of the four-county area obtained by 
applying the prevalence from the serosurvey to the census population 
of the four-county area, by age group.

Results
There were an estimated 44 infections for every ILI ED visit. Age-

specific ratios rose from 19.7 to 1 for children aged <5 years to 45.4:1 
for those aged 5 to 17, to 47.3:1 for those aged 18 to 24, 36.8 for 
those aged 25 to 49, 55.4 for those aged 50 to 64, and 143.8 to 1 for 
persons aged >64 years. The proportion of infections by age group 
is similar but not identical to the proportion of excess ED visits. For 
example, 13.1% of excess ED visits were for children under age 5, 
while 9.7% of estimated infections were in that age group. Persons 
aged 65 years and older, accounted for 2.3% of excess ED visits and 
6.5% of infections.

Conclusions
These ratios provide a way to estimate cumulative incidence 

of influenza in an epidemic, using excess numbers of ILI visits to 
emergency departments. These can be used in real time for planning 
and forecasting, when carrying out timely seroprevalence surveys is 
not practical. Syndromic surveillance data allow age and geographic 

breakdowns in estimates of ILI visits and cumulative infection rates. 
If R0 has been estimated independently, then the seroprevalence at 
which transmission will slow and stop can be compared to the current 
prevalence estimated from ED visits using these ratios. Similar ratios 
should be calculated elsewhere if both population-based ED visits 
for ILI and seroprevalence data for an epidemic strain of influenza 
are available.

Influenza-like illness visits to Florida emergency departments by week, 2007 
to 2011.

Keywords
Influenza; prevalence; emergency department visits

References

(1) Cox CM, Goodin K, Fisher E, et al. Prevalence of 2009 Pandemic 
Influenza A (H1N1) Virus Antibodies, Tampa Bay Florida — 
November–December, 2009. PLoS ONE 2012 6(12): e29301. 
oi:10.1371/journal.pone.0029301

(2) Nokes DJ, Anderson RM (1988) The use of mathematical models in 
the epidemiological study of infectious diseases and in the design of 
mass immunization programmes. Epidem. Inf 1988; 101, 1-20

*Richard Hopkins
E-mail: hopkinsrs@comcast.net    

Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org * 7(1):e29, 2015