The continuing case of the Florida dentist.

Abstract

The subject of the analysis of viral DNA fingerprinting associated with the alleged human immunodeficiency virus type 1 (HIV-1) infection of a number of dental patients in Florida is scrutinized. Reference is made to the recent findings of Ou et al. which are published in this same issue of Science. Scientific and ethical issues are examined as well as the issue of what determines an appropriate level of statistical and scientific reliability for public policy or judicial review. The significance for a resolution of reliability questions is that in the absence of a medical position others unfamiliar will draw conclusions. The public is amenable to changing life styles without necessarily strong statistical evidence. There is agreement among nations however that the HIV-1 evidence is strong enough to recommend use of condoms and medical gloves and invest in research. Legal action mandates the highest statistical significance in the determination of verdicts without a reasonable doubt. In the Florida dental case data a manuscript and the review process became entangled in legal machinations. Out et al. and other scientists involved in legal action have a responsibility to decide in advance appropriate statistical criteria. The difficulties that arise from studies of the transmission of the HIV 1 virus are represented either in Ou et al. and the C2-V3 domain of transmission from dentist to patient or from the Wolinsky study of transmission from infected women to unborn children. The Ou et al. study of the 7 infected dental patients and the 35 local controls (LC) is questioned on the randomness of the LC sample the process of polymerase chain reaction (PCR) and the statistical tests of significance used (Wilcoxon rand sum statistic the distinction between cloned sequences and direct sequences the appropriateness of the bootstrapping procedure). The phylogenetic relationships among variants varies with modes of transmission and tree building is a difficult art and science. The Ou et al. study is considered a pioneering effort in the use of DNA typing to resolve individual viral transmission events but the issue is complex and statistical flaws have already occurred in the Wilson et al. hypothesis of African Eve.