Reliability of Measurements of Tricyclic Levels-Reply

In Reply.—  We agree with Dr Bank's contention that the reliability of a drug level needs to be questioned and that laboratory error is always a possibility. Use of a well-established laboratory with a standardized methodology for withdrawing and processing blood samples minimizes this potential for error.¹ In our case, a local medical school laboratory was used and the assay technique was well established with internal quality control assays done routinely. Furthermore, serial blood levels of our patient demonstrated one other "toxic range" level of 1,040 ng/mL and a subsequent graduated decline over the initial hospital course. We believe that the potential for laboratory error in this case was minimal. We believed that a comment on the possible lower cardiotoxicity of doxepin was beyond the scope of our report. Although one report has suggested that this is the case in animals,² we know of no controlled study definitively