Do Courts Employ the Appropriate Biostatistical Measures? Examples from Recent Legal Cases

• Activity Number: 618
• Type: Invited
• Date/Time: Thursday, August 8, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Scientific and Public Affairs Advisory Committee
• Author(s): Joseph L. Gastwirth*+ and Qing Pan
• Companies: Statistics Department, George Washington University and George Washington University
• Keywords: bioequivalence ; infringement ; FDA requirements ; accuracy of dog sniffs ; probable cause ; data inadequacy

Abstract

Recently, a manufacturer of an existing drug sued a company producing a new drug, claiming the "new" drug infringed on the original. The producer of the new version admitted that its drug involved the same key chemical, however, it was not "bio-equivalent", in the sense the FDA requires of generic drugs. The court decided that another definition of bioequivalence was more appropriate for infringement cases. The statistical properties of the different meanings of "bio-equivalence" will be described and the results of a simulation study support the court's decision. The U.S. Supreme Court will decide a case concerning the data the state should provide to establish a narcotics dog is sufficiently accurate to establish "probable cause". The accuracy of a dog sniff is measured by its sensitivity and specificity. The courts often use the predictive value of a dog's alerts in the field to evaluate the reliability of an alert. It will be shown that it is mathematically impossible to accurately estimate the three relevant parameters from a dog's field performance alone. With the additional information from a dog's training records, the three parameters can be accurately estimated.