The science on this and other mosquito control technologies is sorely lacking. Visual by Flickr

The Blind Pursuit of Mosquito Control

With Zika currently raging in Latin America, public health experts are scrambling to synthesize everything they know about the virus. The paywalls on scientific journals have been lifted with the hope that Zika data sharing will spur international collaboration to staunch the epidemic, which has so far infected one million people and could infect up to four million, according to the World Health Organization. And yet, the literature on controlling the mosquitoes that carry the virus is woefully lacking, according to a new meta-analysis published in PLOS Neglected Tropical Diseases.

Researchers at the Liverpool School of Tropical Medicine, looking at mosquito control studies that sought to reduce the transmission of dengue – a virus similar to Zika that has killed far more people – found that of the 960 studies that have been undertaken in the last 35 years, only 41 stand up to scientific scrutiny. And of those 41 studies, only 9 were randomized control trials, a basic tenet for careful clinical research. The studies, on the whole, presented very little reliable evidence about which mosquito control techniques actually worked.

Spraying for mosquitoes in Panama. Visual by PAHO.

Spraying for mosquitoes in Panama. Visual by PAHO.

“We do not have a clear understanding of which of the currently available interventions actually work, where or when they succeed or might work best, and the reasons why they succeed or fail,” the authors wrote.

The studies surveyed were generally weak in experimental design and show almost no standardization across trials, the authors concluded.

We may have the tools to control mosquito-borne diseases like Zika and dengue, said Liverpool School of Tropical Medicine’s Philip McCall, who co-authored the review, “but we don’t know how to use them optimally.”

Despite the many decades — and many billions of dollars — spent fighting mosquitoes, experts still don’t have reliable, statistically significant evidence that any single control method actually works, according to McCall. Surely bed nets do the trick? Not necessarily: “There was no evidence that the use of mosquito repellents, bed nets or mosquito traps significantly increased or reduced the odds of dengue infection,” the study concluded. Three studies showed window screens lowered dengue incidence but, paradoxically, the use of knockdown sprays and mosquito coils “was significantly associated with an increased odds of dengue incidence.”

This would seem to present a serious problem for public health officials tackling new epidemics like Zika virus. Without better assessments of the efficacy of the tools used to fight Aedes aegypti and Aedes albopictus – two mosquitoes known to carry dengue as well as chikungunya and Zika – how can public health officials expect to control the next disease outbreak? It’s a worrisome thought, though it comes as no surprise to researchers who study the spread of disease.

“We know that current vector control strategies don’t work, at least not very well, given what we are seeing with increasing transmission of dengue,” said Mary Wilson, a visiting professor of epidemiology and biostatistics at the University of California, San Francisco. Wilson pointed to more literature that came to the same conclusions. So did Dawn Wesson, a mosquito expert at Tulane University.

“This is not the first report to say this. The problem is a lack of studies of the control methods, not necessarily that the control can’t work if implemented properly,” Wesson said, pointing to conclusions that agencies like the World Health Organization have also drawn.

Mosquito control works if implemented in “an expedient, comprehensive and sustained way,” the W.H.O. wrote in a communiqué this month.

But without more stringent study design and standardization, the new study seems to suggest, mosquito control teams will continue to be in the dark as to which control methods are best suited to their individual circumstances.