Lung resection in the elderly.

" had a higher prevalence than that in the general US population, and not to study whether " movies influence smoking habits of their viewers, " as other authors have done. 2,3 In order to answer our question, we had to exclude films that took place " outside the United States, in earlier times.. . and smoking by United States citizens " because they obviously do not portray the lives of US citizens in the United States. Equally as important, we introduced objective methodology to a difficult subject. In their zeal to demonstrate a high prevalence of smoking in the movies, the authors from the same institution have repeatedly used arbitrary definitions such as " no smoking signs, " implied or actual tobacco consumption, talking about tobacco, and tobacco paraphernalia. 4 –7 As such, they have demonstrated an astonishingly high prevalence of smoking in all movies, and an even higher prevalence in R-rated movies. These data must unfortunately be dismissed since they are neither scientific nor objective. Furthermore, the letter of Glantz and Polansky has several major errors. First, we did not study " a select group of films from the 1990s. " We studied every movie that met our inclusion criteria. Science fiction and animation (not " fantasy ") movies have an even lower smoking prevalence than live action and non-science fiction movies. If included, the prevalence of smoking in all movies in our study would have been even lower and therefore, not statistically significant, exactly the same as it is now. Similarly, since none of these movies are rated R, they would not have changed the findings for R-rated movies in our study. Obviously, the net effect would still be exactly the same. These movies would not have changed either our findings or our conclusions. Second, " To obtain [our] target sample size, [we] collected data in R-rated films (from January 1) 1996-(until December 31) 1998 " by watching these movies in the theaters, but also continued to watch R-rated movies on video dating back to 1990, even though we had reached the necessary sample size. We collected data from virtually every PG and PG13 movie made in the 1990s that met the inclusion criteria, because there are so few of these movies released each year. Therefore, " this sampling strategy [could] underestimate smoking in PG13 films " does not make any sense since our study includes virtually …