What To Make Of New Positive NSI-189 Results?

Dissects the NSI-189 antidepressant saga. Neuralstem's neurogenesis drug failed its pre-registered primary endpoint (MADRS) in FDA trials (stock -61%); months later the company trumpeted new POSITIVE secondary endpoints (stock +67%). Scott explains the primary-vs-secondary-endpoint safeguard against p-hacking, then shows the positive secondaries 'show pretty much the random pattern you would expect from spurious findings' — scattered across scales, barely below p=.05, present at 40mg but not 80mg, appearing in one stage and vanishing in the pooled analysis. He's 'very unimpressed.' Part III puzzles over the stock jump (efficient markets vs his dad's lucky medical bets; looser FDA standards), framed by the broader antidepressant-research malaise (everything is 'just SSRIs').

Quality 67 · Strong. 67 — Strong. A clear, instructive real-world lesson in secondary-endpoint p-hacking plus a sharp portrait of antidepressant-research stagnation. A model explainer, bounded to one drug episode, which places it mid-Strong.

Claude’s paradigm shift 40 · Moderate. 40 — Moderate. A lucid exposition of known statistics (the primary/secondary-endpoint trap); the value is the clarity, not new ideas.

Real-world impact 2 · Minor. 2 — a good within-discourse teaching example; no material reach.