Every pharmacovigilance system ever built rests on a quiet assumption: that adverse events arrive at a pace a periodic process can absorb. Collect, code, assess, report — on a cadence measured in days and weeks. For most of pharmacology, that assumption holds.

CAR-T breaks it. Cytokine release syndrome and immune effector cell-associated neurotoxicity (ICANS) can escalate over hours, in a patient population that is often already fragile. The clinically decisive window — where grading, intervention, and benefit-risk judgment actually change outcomes — opens and closes faster than a batch-oriented safety process can turn around.

The consequence is not merely a slower report. It is that the documented physician judgment regulators expect may not exist at the moment it mattered. A flag was raised; who adjudicated it, under what criteria, and when? For advanced cellular therapies, that question has to have an answer measured against the clock the biology sets — not the clock the process was designed around.

This is why we treat high-acuity oncology safety as its own discipline. The fix is not more headcount applied to the old model; it is a model in which signal detection runs at machine speed and a named, credentialed physician adjudicates and signs in the same window — with the trail generated as the work happens.