Finnish Foundation for Alcohol Studies, PO Box 220, FIN-00531 Helsinki, Finland
(Received 14 April 2003; accepted 14 April 2003)
I am happy that the authors (Berglund and Thelander, 2003) have found several of my comments (Poikolainen, 2002
) helpful and have improved their data and methods. It is the quality of those data and methods that determines the credibility of the findings and conclusions. Careful reading of the critical reviews on the SBU (2001)
report should suggest further improvements.
Encouraged by the authors constructive response to scientific criticism, I recommend some new improvements to be incorporated into the next edition. (1) Please make available all data abstracted from the original articles that were used in the calculations, at least in the internet edition. This would help possible future meta-analyses. (2) Please define the level of alcohol intake used in the calculations of the number needed to treat (NNT). (3) Also, whenever possible, please present the effect sizes based on NNT, odds ratios or mean changes. It is misleading to compare the effects of various treatments, an important aim of your project, with d-statistic. The d-statistic varies markedly for reasons other that the treatment effect from 1.0 to 2.3 in cases where NNT is always 10. To have unbiased estimates on the efficiency of treatments is also crucial to determine that there is real clinical significance, not only statistical significance. Statistical significance is easy to reach in large meta-analyses, but may be clinically irrelevant.
REFERENCES
Berglund, M. and Thelander, S. (2003) Reply to the commentary by Kari Poikolainen on evaluation of the effect of treatment of alcohol and drug problems by the Swedish Council on Technology Assessment in Health Care (SBU). Alcohol and Alcoholism 38, 443445.
Poikolainen, K. (2002) Invited Commentary: A nice try that fails: The Swedish Council on Technology Assessment in Health Care (SBU) evaluation of the effect of treatment of alcohol and drug problems: The epidemiologists view. Alcohol and Alcoholism 37, 416418.
SBU (2001) Behandling av alcohol-och narkotikaproblem. En evidensbaserad kunskapssamman ställning. Staten beredning för medicinsk utvärdering, rapport nr 156. Stockholm. ISBN: 9187890739. Available online at http://www.sb.se/admin/index.asp