The toxic trio: what’s the number needed to test?

This study retrospectively identified cases from a regional poison center database involving patients with supra therapeutic levels of valproic acid (VPA), lithium (Li), or carbamazepine (CZN)  and “altered mental status and an unclear history of psychiatric diseases or seizure disorder and . . . an unknown medication history regarding the use of VPC, Li, or CZN.” They identified 26 cases over a 3 year period: 8 in the VPA group (113-247 μg/ml), 9 in the Li group (1.9-5.2 mEq/L), and 9 in the CZN group (13.4-38,8 μg/ml). Although it is completely unclear what, if any, clinical significance was indicated by these levels, the authors note that one patient with a lithium level of 5.2 mEq/L (reference range < 1.2 mEq/L) underwent hemodialysis. Since they do not describe this case in detail, it is impossible to know if the dialysis was likely to improve patient outcome. The authors conclude by stating that their toxicology group will continue the practice of drawing levels for VPA, Li, and CZN in patients with altered mental status of unknown etiology. This small study leaves many questions unanswered. How many empirical levels would have to be determined to find one that was clinically significant. What’s the NNT (number needed to test)? Would this practice improve clinical outcomes, or result in unnecessary therapeutic interventions? What would be the cost of all these screening tests? They call for a prospective study, but since relevant cases seem so...
Source: The Poison Review - Category: Toxicology Authors: Tags: Medical carbamazepine drug screening lithium tegretol valproate valproic acid Source Type: news