Video explaining the efficacy of targeted indoor residual spraying during an Australian epidemic. | AAAS/Carla Schaffer
BOSTON – Scientists are testing methods ranging from selective insecticide spraying to intelligent robotic traps to curb disease-carrying mosquitoes by harnessing the power of data collection and targeted interventions, according to presentations at the AAAS annual meeting.
Spraying long-lasting insecticides indoors at identified “virus hotspots” reduced transmission of the dengue virus by up to 96% during an outbreak in Cairns, Australia, a new study in the 17 February issue of the journal Science Advances finds.
Aedes aegypti mosquitoes, the carrier of yellow fever, dengue, chikungunya and Zika viruses, predominantly rest indoors, hiding in dark spaces like closets or under beds. This behavior severely limits the usefulness of interventions such as truck-mounted insecticide sprays.
“I refer to it as the cockroach of mosquitoes because it really likes living around humans,” study author Scott Ritchie, a professor at James Cook University and former director of Medical Entomology at Queensland Health’s Tropical Public Health Unit in Cairns, Queensland said during a press briefing at the AAAS annual meeting.
Gonzalo Vazquez-Prokopec, an assistant professor in the department of Environmental Sciences at Emory University, likened the effort in an earlier interview to “fighting a difficult-to-control vector and a complicated-to-control pathogen.” He said interventions have to be more efficient in how they deploy our resources and identify the correct mix of responses.
Indoor spraying with insecticides that can last for months is more effective against the mosquitoes, but used less frequently because it is time-consuming and resource-intensive.
But epidemiology has entered the era of big data, Vazquez-Prokopec said, ushering in smarter, targeted use of public health resources.
Vazquez-Prokopec and the other scientists studied public health data from 902 confirmed cases of dengue during the 2008-2009 Cairns epidemic, the largest in the city’s history.
As part of the Queensland Health program, nurses called individuals with confirmed dengue cases to identify likely exposure sites – an information gathering process known as contact tracing. Health workers then applied insecticides at many of these sites.
Not all potential exposure locations were sprayed, however, allowing researchers to directly compare dengue transmission rates between the sprayed and unsprayed premises.
The results show the method’s impact during a real-world epidemic, Vazquez-Prokopec said.
In comparison to unsprayed areas, targeted indoor spraying reduced the probability of dengue virus transmission up to 96%. This percentage excludes cases where dengue symptoms appeared within 10 days of spraying, since transmission likely occurred prior to the intervention.
“That level of protection is really unmatched compared to other methods we know exist,” Vazquez-Prokopec said during the AAAS briefing.
Microsoft researcher Ethan Jackson and his colleagues are also developing new approaches to fight mosquito-borne diseases, using robot-enabled technologies.
Jackson leads Project Premonition, a collaboration between Microsoft Research and several universities, which aims to detect pathogens before they cause outbreaks.
Detection is notoriously difficult because the majority of emerging infectious diseases jump from animal to humans through mechanisms that are not well understood, Jackson said.
“Because animals are difficult to track, we borrow technology from nature, the mosquito, which has evolved to locate animals hidden in the dark and collect blood samples,” Jackson said during the news briefing. “What if we could collect mosquitoes at scale and deduce from their body content the animals’ bacteria and viruses they have encountered?”
Project Premonition is doing just that by developing intelligent robotic traps that can capture virus-carrying mosquitoes and collect data from them straightaway, including information such as the time, place, humidity and temperature when a mosquito enters a trap.
The on-demand data collected by the project, which is stored using cloud-based technology, is often what public health officials need to predict disease spread and plan responses accordingly.
The traps were tested in 2016 in Houston, Texas, a high-risk location where many local mosquitoes are prone to becoming hosts for the Zika virus. The test provided training data to improve the system’s recognition abilities; the trap decides which species of mosquito it is supposed to catch based on the flap of its wing.
The Houston traps were more than 90% accurate in identifying mosquito species. Evaluation of the system using synthetic mosquito metagenomes showed that it can also identify viruses and the species that a mosquito had bitten with more than 99.9% accuracy. Early results suggest the system may also be useful for studying the population genetics of insecticide resistance, Jackson said.
In the future, Project Premonition is working to scale the technology for wider use and to enlist drones to identify mosquito hotspots and deliver traps.
“The project, I really believe, is pushing the boundaries and the intersections of robotics, machine learning and metagenomics to try to catch the next epidemic before it begins,” Jackson said.
Eva Lee, a professor at Georgia Tech and director of the Center for Operations Research in Medicine and Healthcare, spoke at the meeting about a new model called ASSURE. The prediction model can use many types of data to help policymakers choose the best, and most cost-effective, intervention strategies to contain an infectious disease outbreak.
For example, the model demonstrates that in Brazil, Zika infections could be reduced by half if just 20% of the population followed recommendations like using insect repellents and window screens and wearing long-sleeved shirts and long pants.
“For policymakers, this is really important,” Lee said.
Mark Mulligan, a distinguished professor of medicine at Emory University conducting research at the Emory Vaccine Center, broached the subject of clinical research and practice in the era of global epidemics.
Mulligan was part of a team in 2016 that helped care for and study five individuals infected with Zika. Early education of clinicians, public health officials and the public is essential for immediate epidemic response, he explained.
“Often the response comes a little too late to implement,” Mulligan said. “Having systems in place ahead of time is a critical need.”
[Associated Image: Atlantic Photography]