Layout 1 ISDS Annual Conference Proceedings 2012. 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 2012 Conference Abstracts A Piece of the Public Health Surveillance Puzzle: Social Contacts among School-Aged Children Molly Leecaster*1, Warren Pettey1, Damon Toth1, Jeanette Rainey2, Amra Uzicanin2 and Matthew Samore1 1Internal Medicine, University of Utah, Salt Lake City, UT, USA; 2Centers for Disease Control and Prevention, Atlanta, GA, USA Objective To enhance public health surveillance and response for acute res- piratory infectious diseases by understanding social contacts among school-aged children Introduction Timely and effective public health decision-making for control and prevention of acute respiratory infectious diseases relies on early dis- ease detection, pathogen properties, and information on contact be- havior affecting transmission. However, data on contact behavior are currently limited, and when available are commonly obtained from traditional self-reported contact surveys [1, 2]. Information for con- tacts among school-aged children is especially limited, even though children frequently have higher attack rates than adults, and school- related transmission is commonly predictive of subsequent commu- nity-wide outbreaks, especially for pandemic influenza. Within this context, high-quality data are needed about social con- tacts. Precise contact estimates can be used in mathematical models to understand infectious disease transmission [3] and better target sur- veillance efforts. Here we report preliminary data from an ongoing 2- year study to collect social contact data on school-aged children and examine the transmission dynamics of an influenza pandemic. Methods Our aim is to capture mixing patterns and contact rates of school- aged children in 24 schools and other non-school-related venues. We used a stratified design to ensure coverage of urban, suburban, and rural school districts, as well as climatically different areas (moun- tains and desert) in Utah. Elementary, middle, and high schools were chosen in each stratum. We defined a self-reported contact as anyone with whom the participant talked to face-to-face, played with, or touched. Contact logs collected subjective information (age, location, and duration) on self-reported contacts during a 2-day period. Ob- jective contact data were collected by using proximity sensors [4] that recorded signals from other sensors within approximately 3-4 feet. Mixing patterns during school and non-school-related activities were summarized for participating school-aged children. We devel- oped contact networks using proximity sensor data, providing visu- alizations of contact patterns as well as numeric contact measures. Contact networks were characterized with respect to degree distribu- tion, and density. The degree for each person was calculated as the number of unique contacts. The density for a network was calculated as the number of observed contacts divided by the number of possi- ble contacts. Results Two elementary schools, four summer camps, and one club par- ticipated in the study between May and August, 2012. Data were processed for the two schools and one camp. The mean degrees for the two schools were 28 and 29, with network sizes 109 and 129, re- spectively. The mean degree from camp was 43, whose network size was 141. The density of contacts was 0.26 and 0.22 for the schools and 0.31 for the camp. The density within classrooms at the two schools ranged from 0.78 to 0.98. School-aged children typically un- derreported contacts using the contact log compared with objective proximity sensor data; this difference was statistically significant. Conclusions The variability in these and other contact network characteristics represent factors that could impact influenza transmission. Quantify- ing these factors improves our understanding of influenza transmis- sion dynamics, which in turn can be used to adapt surveillance methods and control and prevention strategies. Almost all contact among students in our two elementary schools occurs within the classroom and the contact patterns differ by classroom, due to desk arrangement or other characteristics. Thus, during an elementary school outbreak it may be beneficial to focus on classroom-specific surveillance and control strategies. The study is ongoing and we expect the variability in contact rates and mixing patterns will be even greater for middle and high schools where students switch classrooms and classmates each period. These schools could benefit from alternative surveillance and control strate- gies that account for the heightened overall mixing of the student body. Keywords children; respiratory infectious disease; social network; transmission model; proximity sensor Acknowledgments This study is funded by the Centers for Disease Control and Prevention 5U01CK000177. References 1. Mossong, J, et al 2008 PLOS Medicine 5(3): 381–391. 2. Glass L and R Glass 2008 BMC Public Health 8(61). 3. Keeling M and K Eames 2005 J.R.Soc.Interface 2:295-307. 4. Salathé M, et al 2010 PNAS 1009094108. *Molly Leecaster E-mail: molly.leecaster@hsc.utah.edu Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org * 5(1):e25, 2013