Skip to main content
TSRC, Tob. Sci. Res. Conf., 2023, 76, abstr. 52

Horwitz-Thompson equation as a benchmark for CORESTA collaborative study results

MORTON M.
Altria Client Services, Richmond, VA, USA

The Horwitz Equation is intended to approximate the relationship between the variability of an analytical method and the analyte concentration. The equation was developed by William Horwitz and colleagues to provide a benchmark for the reproducibility standard deviation in collaborative studies. It is sometimes referred to as the Horwitz trumpet because it predicts that as the concentration of the analyte decreases, the relative variability of the analytical method tends to be larger. Michael Thompson has more recently suggested refinements to the equation to make it better fit low concentration analytical results and that refined version is referred to as the Horwitz-Thompson (HT) equation. In this presentation, I compare numerous CORESTA Recommended Method collaborative study results involving 17 tobacco and tobacco product analytes and 23 smoke analytes to the HT equation. Overall, there is great variation around the predicted variability, but for traditional unburned tobacco and tobacco products, the CORESTA results are generally in line with the HT equation. For mainstream cigarette smoke, the results were normalized to the approximate cigarette tobacco weight of one gram per cigarette. Interestingly, with that normalization, the smoke analytes in the µg or mg per cigarette range tend to have greater variability than predicted by the HT equation whereas the analytes in the ng per cigarette range tended to have less variability than predicted by the HT equation. These data support the use of the HT equation as a rough benchmark for what to expect with unburned tobacco and tobacco product analytical results but suggest that the use may be more limited for smoke.