Here is some of our research in digestible chunks.

Taking advantage of healthcare data to improve influenza surveillance

By Elizabeth Lee

March 20, 2018

Every flu season, the U.S. Centers for Disease Control and Prevention (CDC) recruits roughly 3,000 physicians across the United States to report how many of their patients appear to have flu-like symptoms. These physicians form the core of the country’s sentinel surveillance system, a data source which is used to determine the geographic spread, timing, and severity of the influenza season nation-wide. While everyone acknowledges the importance of sentinel reporting, physicians are given few incentives to participate due to limited time and resources.¬†Our newest paper, recently published in PLoS Computational Biology, tackles the challenging question of how to improve targeting for sentinel physician recruitment by leveraging the high volume of aggregated medical claims data.

How can we improve sentinel site recruitment?

Compared to traditional sentinel surveillance, our medical claims data has reports from over 120,000 physicians and represents roughly 20% of all visits to health care providers during our study period. We found that our estimates of influenza disease burden and our inference about what drives the variation in its spatial distribution were most robust when the same sentinel locations reported data every year. Yet even with the best sentinel recruitment design, we observed that 10-30% of county-level estimates of disease burden were poor at the level of coverage at which the CDC collects U.S. outpatient influenza surveillance data. This means that surveillance practitioners should strive to recruit the same health care providers each flu season in order to get the most information out of the reported data.

What did we learn about influenza epidemiology?

The statistical surveillance model that we used to evaluate sentinel surveillance design also provided valuable insights about influenza epidemiology in the United States. During our study period of flu seasons from 2002-2003 through 2008-2009, we found that mid-Atlantic states had greater relative risk for influenza disease burden, and that socio-environmental factors, local population interactions, state-level health policies, and sampling and reporting levels contributed to the spatial patterns of disease.


A map of surveilled seasonal flu intensity

Social or Solitary: Does your social network protect you from disease?

By Pratha Sah

March 5, 2018

This post was originally written for Animal Ecology in Focus for the Journal of Animal Ecology

Animal species ranging from mammals, birds, reptiles and fish to insects exhibit an impressive diversity in sociality. Sociality describes the tendency of animals to associate with their own kind, and varies from animals who are solitary (such as desert tortoises and raccoons) to complex hierarchical societies (such as carpenter ants, honey bees and spotted hyenas). Social networks are a valuable tool to understand sociality of a species because they allow us to represent how individuals are connected to others through social interactions. Although we know a lot about how human societies are organized, we know far less about the remaining 99.9% of species on our planet! For my PhD research, I tackled the challenge of exploring the link between animal sociality and their social network structure. Previous literature in epidemiology has established that disease risks are heavily influenced by the structure of social networks. If networks are organized differently in species with different social behavior, can sociality be then linked to vulnerability towards infectious disease outbreaks?

Untapped potential of social network data

The path to answering this question turned out to be serendipitous. Initially, our plan was to focus on a few species where we could obtain social network and disease data by collaborating with field biologists. To plan for the experiments, I began a systematic exploration of the literature on animal social networks. To our surprise, we found an untapped wealth of data on social networks released alongside published studies. Towards the end of my third year, I was able to collect over 600 published social networks of 47 non-human species. We quickly realized that this dataset would be invaluable to answer questions about animal sociality and their disease risks at a much broader scale than initially anticipated.

The next challenge was to come up with a scientifically rigorous method of performing a meta-analysis on this dataset. The animal social networks in our dataset were collected in a variety of different research settings, so comparing the networks prima facie did not make sense. We decided to adapt a robust statistical approach that allowed us to account for differences in data collection methods and sampling effort, as well as control for taxonomy. We were therefore able to tease apart the biological differences between networks that were driven by social behavior of species as opposed to other non-biological explanations.

What did we learn?

Our recently published work in Journal of Animal Ecology, explains the dataset and the methods. We found that raccoons, desert tortoises and other solitary species have the largest variation in their numbers of friends, compared to other social species. In contrast, socially complex species, such as ants and spotted hyenas, have very similar number of friends. Finally, we found that gregarious species like bottlenose dolphins form many subgroups of friends; their social networks therefore comprise of a large number of smaller sized groups. Next, we performed computer simulations of disease outbreaks to find that the structure of social networks protect socially complex species from outbreaks of highly contagious pathogens. Social networks of gregarious species, however, make them more vulnerable to frequent outbreaks of such pathogens.

Why are these results important?

First, our results suggest that beyond taxonomic classification, social behavior of the species can predict the structure of their social networks. Second, contrary to the popular hypothesis that social-living is associated with a higher risk of disease transmission, we found that the organization of social networks can act as a first line of defense, alleviating the higher disease costs of group-living for social species. The study also highlights how meta-analysis of social networks can be performed to answer basic and applied questions in animal ecology.


Key network differences across social systems.