Can LLMs detect negation reliably in safety narratives? A practical look
There is a sentence pattern that appears in safety narratives all the time and that most pharmacovigilance NLP system...
Archive
Blog
Notes on research, methods, papers, and technical work.
Type to search across blog posts.
Stimulated reporting, Weber effects, and what happens to disproportionality after media coverage
If you have spent any time working with spontaneous reporting data, you have probably had the experience of seeing a ...
Pediatric pharmacovigilance: why children are not small adults, statistically or clinically
There is a phrase that comes up often in discussions about pediatric drug safety: children are not small adults. It h...
Benchmarking signal detection algorithms: the reference set problem
If you want to know whether a signal detection algorithm works, you need something to test it against. That sounds ob...
Postmarketing Spontaneous Reporting Systems: the Backbone of Real-World Drug Safety Surveillance
Once a medicine reaches the market, its safety story is far from complete. Clinical trials are essential, but they ar...
How a diabetes drug affects the brain
When people search for “diabetes drug affects brain”, they are usually asking a surprisingly good scientific question...
Artificial intelligence and large language models in drug safety
One thing I keep coming back to in drug safety is how much of the field still depends on reading. Reading labels. Rea...
Why retrieval-augmented generation matters more than model size in pharmacovigilance
There is a recurring assumption in conversations about AI and drug safety that goes something like this: if we just u...
When drug safety LLMs stop being demos and start becoming infrastructure
The most interesting thing in recent drug-safety LLM papers is not that the models are “good.” It is that the field i...
Signal detection in pharmacovigilance
Pharmacovigilance databases are full of patterns, but not every pattern is a signal and not every signal is a risk. O...