By Kristen Pfau
It’s 1:15am, and I have a taxi coming for me in under 3 hours! My flight for Dar es Salaam is at 6. (yikes!) While I needed to go to bed about 3 hours ago, I’ve been busy with last-minute logistics for my trip. Including a pre-trip blog. I actually wrote most of this about a week ago, and planned to edit it/expand it, but at this hour it will have to do! I promise the next blog will be less academic, and include pictures of my new home!
Between my last blog an now, I’ve worked very hard to develop my Expert Elicitation instrument; which, as I explained in the last posting, is a bit of a mix between a survey and an interview protocol. It’s a pretty structured way of asking an “expert” in a certain field to tell us what they think the facts are in areas where research doesn’t exist or doesn’t have a clear answer. Expert Elicitations are a relatively new way to characterize uncertainty in data, and have been suggested as an approach for how expert opinions can be used to provide support for public policy. For my project, I am asking experts about insecticides used for malaria control.
First, I’ll ask a series of questions that ask what level of impact different types of insecticide interventions have on communities, and how confident the person is that a certain reduction in prevalence would result from a certain intervention. Then, the same type of questions but asking about an individual. Generally, the major malaria intervention funding sources and malaria prevention campaigns say that ITNs can reduce malaria prevalence in a community by 20% (important note: a 20% reduction in prevalence is not an absolute number: if one community has a prevalence of 100%, ITNs would bring that number down to only 80%.
But if the starting prevalence is only, say, 40%...well 20% of 40% is only 8%, so the resulting prevalence in the community is 32%.)—but, controlled studies have shown anywhere from 95% reductions to no effect. The same goes with IRS: the very few controlled studies I have found show a range of 100% reduction (effectively eliminating malaria with the use of insecticides) to only a 6% reduction. My questions will try to get a better picture of what a decision maker can expect to happen when an intervention is implemented in a , and which factors are important in making that intervention more or less effective.
The last section—which in my opinion is the most interesting—looks at different side effects that have been associated with insecticide exposure. These side effects (headaches, skin/eye irritation, nausea, birth defects, cancer, death…) have been documented in association with high exposures to insecticides in agricultural uses. Throughout history, the public and the regulatory agencies have often decided that the agricultural benefits of using these insecticides are not worth the negative side effects or potential harms (often times, human health concerns are “potential harms” that we determine because of known harm to wildlife or laboratory animals)—such as with DDT.
But, when the public and regulatory authorities are balancing those negative side effects and concerns against the direct benefits to public health (rather than agricultural benefits, which has a less direct impact on health), the scale tips differently. In this situation, the burden of malaria is huge--we have plenty of data to support the size of the weight on that side of the scale. But, we don’t have many facts to determine the size of the weight on the other side. The final section of my expert elicitation will help determine that missing piece. My hope is that this will serve as support for exactly what the I've heard/read from many people actually on the ground: Here is the concern. The dangers of insecticides as we use them, and the dangers of DDT as we would like to use it, are so little compared to the dangers of not reducing malaria.
