To the Editor:
Although thought provoking and likely to lead to significant research in the future, the data presented by Nanji et al.1 are not sufficient to support a conclusion that medication errors (MEs) occurring during the course of anesthesia lead to meaningful harm. A primary concern is the tenuous link between MEs and adverse drug events (ADEs). Forty of 91 actual ADEs were not related to MEs and were considered “nonpreventable.” No tabulation of the harm caused by these events is given, and it is possible that most of the significant and life-threatening outcomes fell into this category. Of the remaining 51 ADEs, the Naranjo algorithm determined that only about half were “probably” related to the ME, and the other half were considered “possibly” or “doubtfully” due to the error. Thus, MEs may have caused or contributed to less than a third of the ADEs. The overall rate of 28 of 3,671 (0.8%) is considerably smaller than the undifferentiated error rate of 193 of 3,671 (5.3%) and ADE rate of 91 of 3,671 (2.5%) offered by the authors.
A more critical look at the data is necessary to avoid the impression that MEs during anesthesia are a source of significant patient harm and to provide a proper baseline rate for future studies that attempt to lower the rate of ADE during anesthesia by behavioral or technical means. (One can imagine the results of this study being used to promote new barcode/syringe labeling systems, for example.)
What is needed for clarity is a table that gives counts (not percentages) of ADE distributed across the categories of severity. Then the reader can see if the data support the implied conclusion that preventable MEs actually caused significant harm. The authors may wish also to explain why the severity scale starts at “serious,” rather than some lesser degree of harm. In table 1, we present a suggested format for a table to be published as part of the authors’ response.
The authors declare no competing interests.