REVIEW OF: Ahn-Horst RY and Turner EH, Psychol Med 2024;54(5):1026–1033
STUDY TYPE: Literature review and meta-analysis

American Psychiatric Association guidelines endorse benzodiazepines, particularly FDA-approved clonazepam and alprazolam, as first-line treatments for panic disorder. But benzodiazepines are controlled medications with inherent risks, and the authors of this study investigated whether publication bias might affect our perception of alprazolam XR and, by extension, benzodiazepines generally. 

Alprazolam XR was chosen because its RCTs are online, whereas studies of older benzodiazepines require Freedom of Information Act requests for access. Researchers looked for applicable RCTs in the FDA’s publicly accessible Drugs@FDA website. They found five RCTs of alprazolam XR versus placebo for panic disorder that were conducted in the 1980s and 1990s. Researchers compared the study data submitted to the FDA with the findings later published in the well-known and peer-reviewed American Journal of Psychiatry, Journal of Clinical Psychopharmacology, and Psychiatric Annals.

In total, 531 individuals were treated with alprazolam XR and 349 with placebo. By FDA standards, only one of the five studies had positive results, meaning all primary endpoints were statistically significant. Two of the negative studies were published as positive, including one where all five primary endpoints were rated negative by the FDA. The other negative studies were never published.

Researchers conducted two meta-analyses, one with the original FDA data and one with the published data. Alprazolam XR’s effect size for panic disorder was 0.33 (95% confidence interval [CI] 0.07–0.60) with the original data, but the published effect size was inflated by 42% to 0.47 (95% CI 0.30–0.65). For reference, the original data’s effect size was comparable to that of antidepressants for panic disorder. 

CARLAT TAKE
Alprazolam XR is not as effective for panic disorder as the selectively published studies have led us to believe—although it still does have moderate efficacy for the condition. It is unknown if more commonly prescribed benzodiazepines—or other medications generally—are similarly misrepresented in published literature.