Title: Validation of the Population Health Impact Model by using the historical Swedish data on tobacco product use

Abstract:

Philip Morris International has developed the Population Health Impact Model (PHIM) to estimate the reduction in smoking attributable mortality that could result from the marketing of a reduced-risk products (RRPs)* on the population (as a whole) of the market where the RRP is introduced. The PHIM predictions cover mortality from four smoking-related diseases — lung cancer, ischemic heart disease, stroke, and chronic pulmonary obstructive pulmonary disease (COPD). In the absence of long-term epidemiological data, PHIM predictions can only be verified in future studies. An exception is Sweden, where many Swedish smokers have already switched from cigarettes to a proven reduced risk, and non-combustible tobacco product (Swedish snus) for which long-term epidemiologic data is available in the period 1980-2010. Applying the PHIM to the Swedish data, a counterfactual scenario assumes that snus had not been available in Sweden and that the population of Swedish male snus users continued to smoke cigarettes instead. In this case, an increased number of deaths from the four smoking-related diseases has been estimated. Two approaches were used to calculate this counterfactual mortality: 1) indirect standardization considering the mortality rates from other European countries and 2) a “crude method” considering the relative risks of sub-populations with specific smoking and snus use behavior. The distribution of tobacco use habits — snus users (current/former/never) and smokers (current/former/never) — was calculated for 1980-2010 period. Finally, the counterfactual hypothetical mortalities were implemented into the PHIM as input parameters, and the model output was compared with the actual Swedish mortality data in males.

Biography:

Laszlo Pecze has completed his PhD in Epidemiology at the age of 25 from University of Szeged, Hungary. He works as Computational Scientist at Philip Morris International in Neuchatel, Switzerland. He is responsible for computational analysis and statistical programming related to population health impact modelling. He has published five papers in eminent journals in the field of mathematical biology.