Critique of FDA’s latest SSRI Data Analysis

The ALLIANCE FOR HUMAN RESEARCH PROTECTION (AHRP)
Promoting Openness, Full Disclosure, and Accountability
www.ahrp.org

PRESS BRIEFING

Bullet points: Critique of FDA Report Relationship Between Antidepressant Drugs and Suicidality in Adults (dated 11/17/06)

• Report is a slipshod exercise: an in-house examination of partial data supplied and coded by drug companies.

• No data has been posted to allow for other independent verifications, analyses and conclusions.

• Members of Advisory Committee, who are not necessarily experts on the issue, have had less than one month to examine report.

• Although FDA has denied importance of drug induced suicidality for 15 years, none of the scientists who have warned for 15 years that
antidepressants regularly induce suicidality have been consulted.

• Scientifically vetted publications from these scientists were not shared with members of Advisory Committee.

• The design and reporting  of clinical trials have a single purpose: to present sponsor’s drug in the most favorable light in order to gain market
approval from the FDA.

• After years of denials, FDA first reluctantly admits of a causal link between drugs and suicidality in children and adolescents. Now, it admits of a link in young adults, but not in adults.   What’s next?

* Table 18 actually shows a greater than two-fold risk of suicidality for drug-treated adults aged 45-54 compared to those on placebo.

      Age <25: RR = 2.30  (Confidence Interval = 1.04 – 5.09)
Age 45 – 54: RR =  2.29  (Confidence Interval = 0.73 – 7.14)
Age 45 – 64: RR = 1.75   (Confidence Interval = 0.68 – 4.48)  

* FDA continues to hide the drugs’ hazards behind a statistical smokescreen.
How can one deny that a risk exists if the data shows a possible five-fold, even seven-fold increased relative risk?

*FDA puts statistical significance ahead of clinical importance and public health.

* This is what authoritative clinical trial experts have to say about statistical significance: 
"The difference between statistical significance and clinical importance should always be borne in mind. Authors should particularly avoid the common error of interpreting a nonsignificant result as indicating equivalence of interventions. The confidence interval (see item 17) provides valuable insight into whether the trial result is compatible with a clinically important effect, regardless of the P value (94)." [Ref. 1]

• FDA provides no evidence that any age group or any indication is immune from the risk of drug-induced suicidality.

• Drug companies have previously not complied with earlier FDA requests to provide data from all trials. What reason is there to believe that all
relevant trials are now in FDA’s possession?

• Inexplicably, FDA asked drug companies to exclude suicide-related data occurring more than one day after drug discontinuation. That is one of the periods of greatest risk–as is acknowledged in the 2004 FDA-approved SSRI label. So, FDA’s inexplicable cut-off ensures that any discontinuation-induced adverse effects are not counted.

• Subjects in placebo groups have often been “washed out” of psychotropic drugs whose withdrawal effects persist for weeks — suicide-related events in placebo groups may be drug induced.

• The Gunnell (2005) study cited in FDA report counted 16 completed suicides among 40,000+ clinical trial participants; but the FDA reports 8 completed suicides from a database of almost 100,000 participants.

• Precisely because (1) suicide-related events are rare, (2) clinical trials exclude suicidal participants, and (3) clinical trials are not designed to
detect suicidality, signals of suicidality in such trials must be taken extremely seriously, which the FDA fails and has continually failed to do.

• FDA’s analysis is a highly selective document that obscures the full extent of the risk these drugs pose.

Reference 1. See: Section 20 on "Discussion: Interpretation of Results":  The Revised CONSORT Statement for Reporting Randomized Trials: Explanation and Elaboration  Douglas G. Altman, DSc; Kenneth F. Schulz, PhD; David Moher, MSc;Matthias Egger, MD; Frank Davidoff, MD; Diana Elbourne, PhD; Peter C. Gøtzsche, MD; and Thomas Lang, MA, for the CONSORT Group. (Annals of Internal Medicine. 2001;134:663-694)                   http://www.consort-statement.org/newene.htm