Trial discontinuation can be informative, but it is not a complete safety assessment. A 2014 adult ADHD meta-analysis found no statistically significant difference in discontinuation between lisdexamfetamine and placebo, without establishing equivalent safety.
What evidence was pooled?
Maneeton and colleagues searched through January 2014 and included five randomized placebo-controlled trials of oral lisdexamfetamine, 30–70 mg/day. The report described 806 final-study or safety participants across the included studies.
This was an evidence synthesis, not a new 806-person experiment. Participants and findings from the original trials should not be counted again as independent replication simply because they appear in a meta-analysis.
Efficacy and discontinuation answered different questions
The pooled symptom outcome favored lisdexamfetamine, with a standardized mean difference of −0.97 and a 95% confidence interval of −1.15 to −0.78. The pooled response risk ratio was 1.99, with an interval of 1.50–2.63.
A standardized symptom difference and a response ratio do not establish the same outcome. Neither quantifies the probability of an adverse event or the safety of a specific individual.
What did the dropout comparisons show?
For overall discontinuation, the pooled risk ratio was 0.82, with a 95% confidence interval of 0.59–1.14. For discontinuation attributed to adverse events, it was 1.77, with an interval of 0.71–4.40. Neither comparison was statistically significant.
The second interval is wide and includes a substantially higher rate of adverse-event discontinuation with active treatment. Calling the treatments equally safe would replace uncertainty with an equivalence conclusion that the analysis did not establish.
Why dropout is an incomplete safety proxy
Participants may stop for lack of benefit, adverse effects, personal circumstances or other reasons. They may also remain in a trial despite meaningful adverse effects. A discontinuation measure therefore does not capture all experienced harms.
The authors themselves noted that pooled discontinuation alone was insufficient to determine safety. Limited trials and short observation windows also constrain the assessment of rare or delayed events.
Evidence quality and practical interpretation
All included trials were funded by the company holding the lisdexamfetamine patent, and the review contained only five trials. These features do not automatically invalidate the findings, but they limit confidence in broad claims and the assessment of reporting bias.
The historical synthesis supports symptom benefit in selected adults with ADHD while leaving uncertainty about safety comparisons. It does not establish harmlessness, equivalence to placebo, current prescribing advice or a driving-safety threshold.
Scientific source
Maneeton N, Maneeton B, Suttajit S, et al. Exploratory meta-analysis on lisdexamfetamine versus placebo in adult ADHD. Drug Design, Development and Therapy. 2014;8:1685–1693. doi:10.2147/DDDT.S68393. Original commentary on the supplied paper.
Leave a Reply