The Liver Data Says Formulation Limits Worked and Labels Did Not

Researchers at the University of Virginia reported Wednesday in Clinical Gastroenterology and Hepatology that lab-confirmed liver injuries from foreign substances rose from 10.9 exposures per million people in 2000 to 52.9 per million in 2024 ^[1][2][3]. The single largest contributor was not alcohol but acetaminophen taken by itself ^[4], which makes this a product problem for anyone shipping over-the-counter analgesics.

The study drew on roughly 220,000 reports to the National Poison Data System involving people over 15, collected between 2000 and 2024 ^[2]. The rate increase works out to about 4.9-fold ^[17]. These were not minor events: more than 80% of annual exposures led to a hospitalization ^[5]. Around 90% of the injuries involved medications ^[6], with acetaminophen alone accounting for 33% of reports in men and 45% in women ^[7], a gap of 12 percentage points ^[18]. Alcohol was second, rising during the pandemic years ^[8]. Other recreational drugs and dietary supplements also increased but stayed relatively small ^[9].

The useful part for operators is the divergence inside the same dataset. Injuries from medications that combine acetaminophen with other drugs fell sharply in the 2010s, by 60% to 85%, immediately after the Food and Drug Administration recommended new limits on how much acetaminophen those products could contain ^[10]. The authors read that as evidence regulatory action can work ^[11]. Meanwhile the single-ingredient category, where the control has been dose instructions on the package rather than a ceiling inside the pill, kept climbing across the study period ^[12].

That is a fairly clean natural experiment in interface design. When the maximum unit dose was changed in the formulation, harm dropped by most of its volume within a decade ^[10]. When the constraint lived in text that the user has to read, remember and apply across multiple products, it did not hold ^[12]. At least 90,000 Americans report toxic acetaminophen exposures every year under the current arrangement ^[13].

The label itself is part of the problem. Product labeling generally puts the daily maximum at 4,000 milligrams ^[14], while clinicians commonly advise staying under 3,000 if possible ^[15]. That is a printed ceiling roughly 33% above the number practitioners actually recommend ^[19], on a drug where the liver risk is concentrated in overdose ^[16]. Users take more than directed to get through bad pain or because they assume more works better ^[16]. None of this argues against the molecule: acetaminophen is easier on the stomach than NSAIDs and is considered the safest OTC analgesic in pregnancy ^[20], and taking it during pregnancy does not appear to raise autism risk, contrary to recent claims from Robert F. Kennedy Jr. ^[21].

Two things to watch. First, whether the plateau holds: poison center data from the past three years shows the acetaminophen-related injury rate has been steady ^[22], which is the difference between a fixed problem and a stalled one. Second, whether anyone applies the combination-product lesson to single-ingredient packs. The source data speaks to formulation limits and labeling, not to pack sizes or blister formats, so the effect of unit-dose packaging remains an open question rather than a finding.