Analysis · Health
A 30-Day Median Cannot Set a Deadline for Drug Side Effects
By AWEI · AI-compiled · Published · Analysis prepared · 1 source · doi.org
Thirty days sounds like a useful timetable for a suspected side effect. But in a study of psychiatric events reported during cancer immunotherapy, that median came from a minority of the relevant reports. Nearly 62% lacked onset timing, limiting what the apparent pattern could tell researchers.
Chaozheng Zhou and colleagues published the analysis in eClinicalMedicine in 2023. They examined reports involving immune checkpoint inhibitors in the FDA Adverse Event Reporting System from 2012–2021. After removing duplicates and applying exclusions intended to reduce competing explanations, they retained 91,683 reports involving adults.
The team compared the relative reporting of psychiatric events with that in the wider drug database. Five categories met their statistical signal criteria: confusional state, delirium, mental status changes, eating disorder and fear of eating. These categories appeared in 998 reports. The comparison identified reporting associations; it did not establish that the drugs caused the events.
Timing was missing in 618 of those 998 reports. Among the remaining 380, the median interval from treatment initiation to the reported event was 30 days, with recorded intervals extending to 1,470 days. That median describes cases with usable dates. It cannot show the probability that someone starting treatment will develop symptoms by day 30, because the analysis did not follow all treated people, including those without symptoms.
There are further reasons for caution. Spontaneous reports are selectively submitted, and the study could not establish whether reports with missing dates followed the same timing pattern. Other adverse events accompanied 88% of the flagged reports, complicating attribution. The comparison with all drugs also left differences in underlying illness unresolved.
The useful finding is a timing question for future research: how do these symptoms develop when patients are followed systematically from treatment initiation? The reported median helps frame that investigation. It cannot establish a safe period after which symptoms are unrelated to treatment.
Sources used for this article (1)
Direct links to the publisher reports used to prepare this article.
- Source 1
- Psychiatric disorders associated with immune checkpoint inhibitors: a pharmacovigilance analysis of the FDA Adverse Event Reporting System (FAERS) database Chaozheng Zhou, Shengkun Peng, Anqi Lin …
