- Jim Ferraro
The Importance to the Case of an Ecological Study
One of the last witnesses to testify in the Castillo-DuPont case was Dr. Brad Pollock, an associate professor and the director of the department of epidemiology at the University of Florida in Gainesville. His area of expertise was in pediatric epidemiology. He testified for us in a rebuttal case, as did Sonya Sipes.
Epidemiological studies are designed to determine the cause and distribution of disease in populations, with the goal of identifying risk factors for disease as well as potential protective means to reduce the occurrence of disease.
While there are several different types of epidemiological studies, the bulk of the studies conducted are observational studies. These are studies in which people who have been exposed to some sort of disease are observed so that researchers can learn more about the subsequent outcome of the disease.
In many ways, observational studies are randomized, because no one is actually being subjected to some sort of potentially dangerous disease or noxious agent; researchers are merely watching a segment of the population that has already been exposed to such a disease or agent in order to witness and better understand the subsequent effects.
There are a number of subcategories of observational studies that can be ranked in order of how much evidence they provide in explaining the causes of a particular type of disease or condition. Some examples of these subcategories include cohort studies (also called prospective studies), case control studies, ecological studies, and cross-sectional surveys.
Having the findings of an ecological study—one in which disease occurrence is measured by exposure within an aggregate grouping of people—was extremely relevant to our case.
The example that Dr. Pollock gave on the stand involved a study that looked at the rate of childhood leukemia in the state of Florida, by examining geographic units like Dade County, Alachua County, and Broward County, and comparing rates of leukemia occurrence. While such a study doesn’t provide information on what an individual’s actual exposure to a disease may be, it does allow scientists to make an assessment based on a geographic grouping of people.
Dr. Pollock read about an attempt to conduct this type of study with regard to the relationship between benomyl exposure and incidence rates of microphthalmia and anophthalmia in Italy known as the Spagnolo Study.
In the simplest terms, the researchers got information about how much benomyl was sold in the region under examination and then looked at the acreage therein to reach some conclusions about exposure levels.
For the study, the scientists also looked at data on the occupations of the parents and birth defect records that indicated one or both parents were involved in the agricultural industry, along with information about the rate of occurrence of an anophthalmia in various regions. Dr. Spagnolo’s study pointed out that there was no possible way (nor was any attempt made) to conduct an ecological study on pregnant women who were exposed to benomyl at critical points in their pregnancy.
In other words, this study was meaningless to both the Castillos and the defendants. It proved literally nothing.
This was an essential point to drive home one last time in the trial, as DuPont had consistently pointed out that there had been no epidemiological studies to support our case. There were none to support the DuPont defense, either.
In my next post, I recount the testimony of Dr. Pollock.
I’d love to hear from you. Have you ever used studies in your work to support what you were arguing? Have you seen the same study used in a different way to say something the opposite? How did this play out? Thank you for sharing.