Severity and carbamazepine level at time of initial poison center contact correlate with outcome in carbamazepine poisoning.

We reviewed data from carbamazepine poisonings reported to the Kentucky Regional Poison Center from January 1986 through March 1992 to identify information available at the time of poison center contact which correlates to outcome. The Spearman rank correlation test was used to describe the relationship between two ordinal variables and interval-level variables. The Kruskal-Wallis test was used to determine the relationship between categorical and ordinal variables. Two way analysis of variance was used to test the effect of routine carbamazepine use on final severity and carbamazepine level of 345 reports involving carbamazepine poisoning; 263 (76%) involved only carbamazepine ingestion and formed the database. One hundred eighty four (70%) carbamazepine ingestions occurred in victims < or = 17 years old, 79 (30%) occurred in adults. Severity assigned at the time of initial poison center contact was significantly correlated with outcome severity for children and adults (r > or = 0.9, p < 0.00001). The amount reported ingested influenced the correlation between initial and final severity; whereas, time elapsed between ingestion and poison center contact did not alter the correlation between initial and final severity. The reason for ingestion was significantly correlated with outcome (p < 0.00001). A significant correlation between outcome and peak carbamazepine level for each age group was observed (pediatric r = 0.5, p < 0.00001, and adult r = 0.4, p = 0.008). Carbamazepine levels > 85 mumol/L (> 20 mcg/mL) were associated with more severe toxicity.