The difficulty of risk assessment for hepatic injury associated with supra-therapeutic acetaminophen use

We examined historical and laboratory features to identify patients who do not experience hepatic injury when presenting with a history of repeated supra-therapeutic acetaminophen use by performing a retrospective, double cohort comparison over almost 5 y. One cohort included patients suffering severe hepatic injury after supra-therapeutic acetaminophen ingestion; the control cohort included all patients in a geographically limited population with repeated supra-therapeutic acetaminophen ingestion who did not experience severe hepatic injury. Demographics, baseline health, medications, ethanol consumption, acetaminophen dosage, reason for excessive acetaminophen dosing, acetaminophen concentration at presentation, presentation and peak hepatic enzymes, N-acetylcysteine (NAC) treatment, highest measured AST and outcome were abstracted from 1114 possible patient contacts. Twenty two of these met all inclusion criteria. Ten suffered severe hepatic injury, 12 did not. The actual acetaminophen dose consumed over a period of time was difficult to establish, but the historic quantity ingested exceeded 20 g in the majority of patients in both groups. High ethanol use was common in both groups. All patients who ultimately suffered severe liver injury presented with some injury. No patient who presented without some injury suffered severe injury. None of these patients were treated with NAC. Two patients who presented with minor injury (1 of whom received NAC), did not progress to severe injury. Future investigations of the cause of liver injury in patients exposed to supra-therapeutic doses of acetaminophen will require larger patient numbers. The lack of injury in similarly exposed patients underscores the peril of attempting to assess the role of risk factors without an appropriate control group. Our experience suggests history is unlikely to be clinically useful in predicting risk of injury. Future risk assessment studies should focus on objective presentation features like presence or absence of injury or serum acetaminophen levels.