In the paper of Zonda and Lester (2006-2007) the authors have made a basic error in their data analysis. With the appropriate reanalysis of their data, the new results and conclusions will be completely the opposite of the original ones.
I have only recently read the paper by Zonda and Lester (2006-2007) in which they tried to “repeat” the Swedish Gotland study (Rutz et al, 1989a, 1989b, 1992; Rihmer et al, 1995) in Budapest. The authors found that the postgraduate GP training on diagnosis and treatment of depression resulted in A significant increase in the sales of antidepressants in the district (66.000 inhabitants) where the GPs were educated compared to the control district (82.000 inhabitants). However, the suicide rate in the district where the intervention took place remained the same in the year following the training (1999) and increased in the year after (2000). In the control district the suicide rate dropped remarkably in 1999 but increased again in 2000. The authors have concluded that the education of GPs has no impact on the suicide rates (Zonda and Lester, 2006-2007).