Friday, June 03, 2005

Mosquito spraying: safe and effective?

So much as happened since West Nile Virus (WNV) first made its appearance in New York City in 1999 it seems like ancient history. But back then, there was fierce controversy over spraying pesticides to kill adult mosquitoes (adulticiding) and the debate revealed how little we know about the benefits and hazards of this accepted public health practice. CDC has just published the results of three investigations of human exposure to pesticides from broadcast ultra low volume (ULV) adulticiding applications (Mortality and Morbidity Weekly Reports, MMWR). The three studies were done by health departments in North Carolina, Virginia and Mississippi, the first two comparing a pesticide metabolites in the pre- and post-spray urine of a random sample in the treated areas, the third, comparing urines in a sample of people in a sprayed area with those in a control area. Some attempt was made in each case to account for household and occupational exposures. The pesticides were permethrin, naled and d-phenothrin.

On the basis of these data, CDC concluded that these kinds of ULV applications "are safe for humans" because results showed no significant evidence of exposure (using metabolites in spot urines as outcome). The data presented did not suggest much exposure, but evidence of exposure to humans is only one aspect of pesticide use. Another is damage to the environment and the ecosystem. Clearly some organisms are severely affected (the mosquitoes). We know little about effects on non-target organisms and this is a concern.

But there is yet another aspect that is rarely touched upon and is perhaps the most important. West Nile and other mosquito arboviruses like Eastern Equine Encephalitis are rare but serious diseases. The reservoir for these viruses is usually birds and humans get sick when a mosquoto bites both an infected bird and a human. Mosquitoes that bite both birds and humans are called bridge vectors. What we need to know is whether mosquito adulticiding actually interrupts human transmission of these diseases. There is surprisingly little evidence that it does (in fact I know of none). Adulticiding only reduces the mosquito population temporarily, and the amount of the reduction, particularly in an urban setting, is difficult to determine. Repeated applications would be necessary if this control mechanism is chosen, while the impact on the ecosystem of all the effects of pesticides, both acute and lingering, is unknown.

A number of factors contribute to reduced effectiveness of reaching target mosquitoes with typical truck-based application, particularly in urban environments like New York City:
  • the most prevalent bridge-vector mosquitoes prefer birds, particularly birds at rest, of which there are few in the street and building-front areas when most sprays are applied;
  • roosting areas may be higher than the reach of the spray;
  • buildings close to the street restrict the lateral spread of the spray;
  • backyard roosting areas are not effectively reached because close spacing of buildings limits penetration beyond the buildings;
  • the period the spray is effective and airborne is of relatively short duration.
  • In any environment, ground and aerial applications produce skips and patchy coverage, especially where habitat produces barriers to spray dispersal and penetration.
Because of these and other factors, the effectiveness of spraying will likely be significantly less than the laboratory measured effectiveness of the pesticides. Nor is it known what proportion of mosquitoes must be affected to have a meaningful and positive effect on transmission of the disease to humans. Arguments can and have been made that adulticiding may even have the paradoxical effect of increasing risk by its effects on mosquito predators (like dragon flies) or the selection of healthier mosquitoes with longer lives. These uncertainties are a reflection of our substantial ignorance about important facts, an ignorance which should provoke further caution.

So it's good news that on the basis of this evidence the ULV applications in these three instances didn't contaminate humans much. The question is, what else did or didn't they do?