1 min read
Connect with Veristat at Precision Clinical Trials in Boston
April 30, 2026 | Boston
⚛️ Where Clinical Strategy Meets Execution That Holds Up Under Scrutiny
On April 30, we will...
2 min read
When an oncology drug is tested, the signal is often brutally clear. A tumor either shrinks on a scan or it does not. A patient either lives or does not. In psychiatric drug development, the endpoints are almost never so crisp. In PTSD trials in particular, the entire question of whether a drug works rests on structured conversations.
The instrument at the center of this world is the CAPS-5, the Clinician-Administered PTSD Scale for DSM-5. It is the near-universal primary endpoint for modern PTSD registration trials, and the FDA has explicitly acknowledged it as acceptable for this purpose.
A trained clinician spends forty-five to sixty minutes asking a patient detailed questions about each of the twenty symptoms that define the disorder, rating each one on a combined frequency-and-intensity scale from zero to four, with a total ranging from zero to eighty. A reduction of approximately ten to thirteen points from baseline is generally considered to represent reliable, clinically meaningful change, though thresholds vary by study and population.
Supporting the CAPS-5 is a family of other scales. The PCL-5 is a twenty-item patient self-report that tracks the same symptoms. The MADRS assesses co-occurring depression. The Sheehan Disability Scale captures functional impairment. The EQ-5D-5L measures quality of life for health-economic purposes. The CGI-S is a single-item global severity rating. The Columbia-Suicide Severity Rating Scale documents suicidality. The MINI confirms the diagnosis at screening and rules out exclusionary comorbidities. Each instrument has a specific role, and together they attempt to triangulate a multidimensional condition into analyzable numbers.
The challenge with any subjective endpoint is that it is vulnerable to influences that have nothing to do with the drug. Expectancy is probably the largest. A patient who believes the treatment will help will often report improvement whether or not the biology of the disorder has shifted.
This is the placebo response, and in PTSD trials it is substantial: meta-analyses of pharmacotherapy trials have repeatedly shown placebo response rates that often exceed 30 percent on standard outcome measures, varying considerably by trial design and population. A drug therefore needs to show improvement not merely against no treatment but against the powerful combination of hope and attention that accompanies being in a trial.
For psychedelic trials, this problem becomes particularly acute. A participant who has received a strong psychoactive drug generally knows it. In the MDMA trials that were recently rejected by the FDA, the overwhelming majority of participants correctly guessed their assignment, and functional unblinding was explicitly identified as a central concern by the advisory committee. Their expectancies then loaded onto the subjective ratings that defined the primary endpoint. The trial technically met statistical significance, but the FDA concluded it could not tell how much of the effect came from the drug and how much from knowing one had received the drug.
Central and remote raters, blinding integrity questionnaires, and active placebos are all strategies for mitigating these problems. None of them solves the fundamental fact that in PTSD, the outcome is an interior experience, and the science of measuring interior experiences remains imperfect, even now.
Contact us today to learn how we can help.
