Rainy With a Chance of Plague Gregory Glass, PhD Director, Global Biological Threat Reduction Program Southern Research Institute Birmingham, AL Professor, Departments of Molecular Microbiology & Immunology and Epidemiology The Johns Hopkins Bloomberg School of Public Health Baltimore, MD
Challenges & Solutions Zoonoses are often sporadic Explosive (short generation time) especially when they adapt to new hosts Leaves little time to react Where Ecology Plays a Role For something new What is the Source? Where will the next outbreak occur? When will it happen?
Vector-borne/Zoonoses Vector-borne diseases: Infectious agents transmitted to humans through action of another species Many vector-borne diseases are transmitted by arthropods; for example, Lyme disease, bubonic plague, dengue, WNV Human ENVIRONMENT Vector Pathogen Reservoir Nearly all are zoonoses
Solution: Measure Environment & Health Data, Analyze, Locate Red areas 17x more likely to have Lyme dz Glass et al 1992
Where Do We Get Environmental Data? Remote Instrumentation Global Earth Observation System of Systems (GEOSS). This emerging public infrastructure is interconnecting a diverse and growing array of instruments and systems for monitoring and forecasting changes in the global environment. NASA, NOAA, USGS, EPA, CDC. DoD, USDA, NSF, DHS, USAID Types of ecological data: land use/cover elevation vegetation moisture temperature
Where Do We Get Environmental Data: In situ Monitors Summer Thunderstorm Globally, estimated 2% increase; In N. America regional increases/decreases
Where Do We Get the Health Data? Waiting until people are sick is a bit late
Where Do We Get the Health Data? Survey for potential reservoir and presence of vector Flagging for ticks along transect Blood sample for pathogen detection
What Do We Get When We Put It All Together- Forecasting Mosquitoes That Spread West Nile Virus Shone et al, 2006 J Med Entomology
0 Mosquito Count 200 400 600 0 0 200 400 600 Mosquito Counts 1000 2000 0 500 1500 2500 0 0 200 600 1000 0 Mosquito Count 400 800 0 200 600 1000 Different Environment Predicts Different Risk 1965 Model with Interval Lagged Covariates 1965 A 1973 Model with Interval Lagged Covariates 1973 B 1981 Model with Interval Lagged Covariates 1981 C June July August Sept Environment Temperature Humidity Windspeed Tides Stream flow Moonlight Health Mosquito counts Agency NCDC Samson Samson NOAA USGS USN Agency MD Dept Agriculture
Year-to-Year Variation in HPS
A Year in Advance, Spatial Extent of Risk Predicts Outbreak Years 120,000 km 2 Utah Colorado Study Area Arizona New Mexico
Plague is an Introduced Vector-borne Zoonosis
Plague Cases in the Western U.S., 1970-2003 130 0'0"W 120 0'0"W 110 0'0"W 100 0'0"W 90 0'0"W 40 0'0"N 130 0'0"W 30 0'0"N Washington Oregon California Nevada Idaho Montana Utah Colorado Arizona New Mexico Wyoming North Dakota South Dakota Nebraska Texas Kansas Oklahoma 40 0'0"N 30 0'0"N 1 Dot = 1 Case randomly placed in county of exposure µ 0 135 270 540 810 1,080 Miles 120 0'0"W 110 0'0"W 100 0'0"W 90 0'0"W
Plague Outbreaks Can be Predicted if you Know Rainfall the Year Prior Winter Temperatures 4 Years Prior Summer Temperatures 3 Years Prior You Need Relatively Long Records of the Environment and Health Data Brown et al 2010 Ben Ari et al 2009
Zoonoses May be Natural or Human Driven: Forecasting Anthrax Affects both people and animals Many zoonoses directly impact people AND agriculture Blackburn 2007
Foreign Policy Applications: Cooperative Biological Engagement Program Anthrax outbreaks Tularemia strains
Forecasting Into the Future: Changing Environments Brownstein et al 2005
Summary Human cases Zoonotic disease ecology has been cobbled together by disparate groups There are lots of research questions that can be answered Basic public health Climate variability Security Agriculture Ecology of zoonoses can give you the what More importantly it can give you the where and when With enough lead time there is time to do something about it. Andreo 2011