Changing nomogram risk zone classification with serial testing after acute acetaminophen overdose: a retrospective database analysis.

Context: The Rumack-Matthew nomogram stratifies patients into discrete risk zones following acetaminophen (APAP) overdose. Treatment decisions have traditionally been based on the initial risk zone. "Line-crossing" between zones occurs and is poorly understood. The study objective was to characterize line-crossing behavior in acute APAP overdose patients, especially moving from below to above the nomogram treatment threshold.

Methods And Materials: The study was a secondary analysis of the Canadian Acetaminophen Overdose Study (CAOS) database, a large medical record review of patients hospitalized in eight large Canadian cities (1980-2005) following APAP poisoning. Population consisted of acute APAP overdose patients with at least two serum concentrations performed during hospitalization. Using ordinal logistic regression, we studied the effects of patient demographics, ingestion size/timing, APAP concentrations, time to N-acetylcysteine (NAC), and co-ingestants on a three-level dependent variable: patients whose risk increased two or more zones, those remaining in the same or adjacent zone, and those whose risk fell by two or more zones.

Results: Of the 3201 eligible hospitalizations with 7705 APAP concentrations, half (1679, 52.5%) crossed at least one zone (up or down) within 24 h of acute ingestion, including 190 (5.9%), who crossed at least two lines into a higher risk zone, and 516 (16.1%) at least two lines into a lower risk zone. Of the 1251 patients initially below the nomogram treatment line of 150 μg/mL, 131 (10.8%) patients crossed above this line. Being older, male, and co-ingesting opioids, antimuscarinics, or NSAIDs were independently associated with line-crossing.

Conclusions: Patients commonly crossed nomogram risk zones, including from below to above the current treatment threshold. These findings support recommendations for serial APAP testing until the individual risk of hepatic injury is clearly established.