Our statement that “the overall incidence of anaphylaxis was similar to that reported in previous studies, but that of hypersensitivity reactions was nearly seven times higher” accurately reflects our results.1  That said, the results of any study need to be interpreted within the context and methodology of the study. As specified in our title and throughout the report, the investigation was restricted to intraoperative hypersensitivity reactions. Because delayed hypersensitivity does not present clinically during the perioperative period, such reactions were not included in our data set. Including delayed reactions would, presumably, increase the incidence we report.

The six search criteria we used to identify possible intraoperative hypersensitivity reactions were not chosen arbitrarily. Instead, as described in the article, they were selected by a Delphi process involving 10 highly experienced board-certified anesthesiologists. All six criteria are clearly related to possible intraoperative hypersensitivity, as recorded in clinical, laboratory, or administrative records. In an effort to capture all possible intraoperative hypersensitivity reactions, we deliberately used broad search criteria, recognizing well that most screened cases would be false positives.

Dewachter and Mouton-Faivre suggest that certain medical key words should not have been used to identify candidate cases for further review. Restricting our search may have eliminated candidate cases that were adjudicated as having hypersensitivity reactions based on other aspects of their records. Therefore, restricting the search could only weaken our study. Most anesthesiologists would agree that flushing, laryngospasm, and stridor are clinical symptoms potentially consequent to intraoperative hypersensitivity reactions and, therefore, should be included in a key word search.

The adjudication committee did not select 264 cases. Instead, members of the committee independently and rigorously adjudicated more than 4,000 individual candidates using all available information in the medical record. The strength of our study was our novel and robust approach to using a large clinical registry to identify rare intraoperative events.

Dewachter and Mouton-Faivre criticize our inability to identify elapsed time between administration of suspected triggers and onset of clinical features. Operating room reality precludes accurate identification of elapsed time. Induction of anesthesia involves simultaneous administration of multiple medications; designating any particular medication as the “suspected trigger” would be purely speculative, especially in retrospect. It is also often difficult to identify the onset of intraoperative hypersensitivity reactions, especially mild ones manifest as rashes, which may not be observed until surgical drapes are removed. We discuss these difficulties in the article.

Of course it would have been helpful if our patients had more laboratory and allergy testing, but our registry reflects clinical practice at our center. Because our analysis was retrospective, we obviously had no ability to refer patients to an allergy clinic for additional evaluation.

It is important to distinguish between results and interpretation. As reported in Results, “use of neuromuscular blocking agents was not significantly associated with experiencing hypersensitivity reactions.” We agree that the result is surprising, and in Discussion, thus, we carefully frame the finding in the context of previous studies, most of which identify muscle relaxants as the leading cause of anaphylaxis.

Similarly, we found that “high-molecular-weight hetastarch was significantly associated with hypersensitivity reactions.” We cannot comment on the study by Dewachter and Mouton-Faivre because it is not yet publicly available. But anaphylaxis to hetastarch is rare. It is important to recognize that hetastarches might lead to an intraoperative hypersensitivity reaction or they might be given to treat ensuing vasodilation and hypotension. Either will result in a statistical association; thus, interpretation of this finding was carefully framed in our discussion.

In summary, our novel and robust study identifies a high incidence of intraoperative hypersensitivity reactions. In contrast, the incidence of severe reactions was comparable with that previously reported.

Drs. Saager and Sessler received honorarium for participation in an MSD Pharmaceutical (Oss, The Netherlands) expert panel.

Incidence of intraoperative hypersensitivity reactions: A registry analysis.