This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

Letters to the Editor

Odd Odds

See the reply

Odd Odds

To the Editor: In their Journal article published online ahead of print,1 Carpenter et al described a meta-analysis of the relationship between paroxetine treatment and emergent suicidality using data from 61 randomized controlled trials in adults. I am puzzled that, in their article, they chose to report risk estimates using odds ratios (ORs) rather than relative risks (RRs). The latter statistic is more easily understood and is generally preferred when randomized controlled trials data are available.2 One wonders whether the preference was to allow for adjustment for confounders in downstream analyses, given that, when the outcome is rare, RRs and ORs differ little.

The abstract of the article reported that, in the primary dataset, suicidal behavior was identified in 50 of 8,958 paroxetine patients (0.56%) versus 40 of 5,953 placebo patients (0.67%). Whereas this translates into an OR of 0.8, the OR actually reported was 1.2. The 50% difference in value is quite a substantial difference arising from statistical adjustment for confounders, if indeed adjustment is an explanation for the difference and not reporting error. In this context, the other ORs reported in the abstract showed no or negligible variation from their crude values. In any case, if a reported value is arrived at after adjustment, it should be specified as such.

References

1. Carpenter DJ, Fong R, Kraus JE, et al. Meta-analysis of efficacy and treatment-emergent suicidality in adults by psychiatric indication and age subgroup following initiation of paroxetine therapy: a complete set of randomized placebo-controlled trials. J Clin Psychiatry. 2011;72(11):1503-1514. PubMed doi:10.4088/JCP.08m04927blu

2. Streiner DL. Risky business: making sense of estimates of risk. Can J Psychiatry. 1998;43(4):411-415. PubMed

Chittaranjan Andrade, MD

andradec@gmail.com

Author affiliation: Department of Psychopharmacology, National Institute of Mental Health and Neurosciences, Bangalore, India. Potential conflicts of interest: None reported. Funding/support: None reported.

Related Articles

Volume: 72

Quick Links: Assessment Methods , Research Methods Statistics