Government officials won't be ready until fall to evaluate the effects of aerial spraying of mosquitoes in Sacramento County, and it may be impossible ever to draw firm scientific conclusions on how much the pesticide treatments helped to stem the West Nile virus.
That's partly because of weaknesses in the Sacramento-Yolo Mosquito and Vector Control District's surveillance system for the virus and partly because of circumstances beyond the agency's control, a Bee analysis shows.
For example, traps and tests the district uses to detect the virus in mosquitoes are inadequate for showing how well pesticide treatments have worked, according to independent scientists who reviewed the test results at The Bee's request.
"You really want more extensive data to be able to get a handle on the actual prevalence of the disease," said J. David Allan, an environmental scientist at the University of Michigan. "It's a matter of (doing) more thorough sampling."
Allan noted that the Sacramento agency probably does as good a job as its counterparts around the country, if not better.
"This is a challenging problem that most agencies are unable to address well because it requires extensive surveillance activities typically beyond the financial capabilities of most vector control districts," he said.
David Brown, manager of the local vector control agency, acknowledged that the data Allan reviewed have gaps, but said the district buttresses its mosquito infection findings with other ways of monitoring West Nile virus.
Those other ways include mapping reports of dead birds, human cases and horse cases.
Taken together, the tests and reports painted a picture in late July of a looming epidemic that the district sought to stop with a last-resort plan for aerial spraying.
It sprayed about 50,000 acres north of the American River on Aug. 8-10 and about 60,000 acres south of the river on Aug. 20 and 21. Portions of the zone south of the river also were treated on Aug. 11 and Aug. 22. (Efforts to spray the whole zone on those dates were stymied by wind.)
The decision to spray was highly contentious, raising questions from the public about whether it was wise to use pesticides so broadly against a disease that only rarely kills people and often causes no symptoms at all.
Public health officials were unwavering, noting the potential for the virus to disable or kill previously healthy people. As of Friday, Sacramento County continued to log the highest number of cases in the state - 112, nearly a third of which involve the serious neuroinvasive form of the disease.
Officials frequently defended spraying by talking about how many mosquitoes were found through the vector district's surveillance to be infected. The virus is transmitted by the bite of infected mosquitoes.
At the time, officials cited mosquito infection rates around 20 in 1,000 - well over the rate of five per 1,000 that public health officials consider epidemic.
Preliminary district data shows that the rate in the northern half of the county after spraying was about four infected mosquitoes per 1,000, suggesting that the treatment worked at least to suppress the prevalence of the virus in the insects.
But the trapping system upon which the rate is calculated has flaws.
The main problem is that the district moves most of those traps regularly, trying to pinpoint hot spots.
While chasing hot spots makes sense, it comes at the expense of being able to fairly compare one week's results with the next, said Walter Boyce, a veterinarian who heads the Wildlife Health Center at UC Davis.
"The more trap locations you have that are permanent, the better you can interpret changes and mosquito numbers and infection rates," Boyce said.
The district has another set of traps that don't move, but those measure only the abundance of mosquitoes in each place. They are not used to detect the virus.
Boyce said that in the absence of more data about infected mosquitoes, information from those fixed traps can contribute to an assessment of how well spraying worked.
"By reducing mosquito numbers and density, you reduce the probability or possibility that mosquitoes will acquire West Nile and be available to bite people," he said.
Brown said the district wants to boost the rigor of its surveillance, but it also needs to be practical, even if that means sacrificing scientific objectivity sometimes.
For example, he said, when looking for good locations for traps, district staff deliberately select spots where mosquitoes are known to lurk, such as creek sides or greenbelts, as opposed to parking lots.
The agency also has trouble distributing traps evenly across the district because traps get vandalized repeatedly in some areas.
Beyond mosquito traps, other components of the district virus surveillance are even less precise. For example, dead bird counts depend upon reports from the public, and only a small proportion of dead birds actually have been tested for the virus.
The number of human cases, too, is but a snapshot of infections in people, since doctors typically don't test for West Nile virus except in people with serious or lingering symptoms.
Boyce said a good way to gauge the effect of insecticide spraying in Sacramento would be to compare the county's experience with that of a similar community that opted not to spray.
Vicki Kramer, chief of the vector-borne disease section of the state Department of Health Services, agreed, but said it may not be possible to find the perfect comparison community because mosquito pesticides are being used widely in other parts of the Central Valley, as well.
Nevertheless, Kramer said health officials will be able to generally assess the value of spraying by looking at the dead bird reports, human and horse cases and mosquito surveillance data, gaps notwithstanding.
She said health officials particularly will scrutinize whether the rate of new human infections in Sacramento is lower after spraying compared with the rate of new human infections elsewhere in the state or in the state as a whole. That pattern won't emerge for several weeks yet.
In the end, it's possible the community will remain split on whether spraying was a good idea, Boyce said. "The effects of spraying may not be so dramatic that it's clear-cut to everyone that this was the right thing to do," he said. "It's ultimately a value judgment."
To spray or not to spray [270k PDF]
About the writer:
- The Bee's Edie Lau can be reached at (916) 321-1098 or firstname.lastname@example.org.