The natural history of allergy to fire ants is different from that to yellow jackets and honey bees. The epidemiologic questions in the second part of Dr. Jung's letter have been partly answered in retrospective studies involving field stings in large groups of patients [1-3]. We previously addressed some of those questions in a prospective study involving sting challenges [4]. Preliminary data of a new prospective study involving more than 300 sting-challenged patients indicate that none of the current epidemiologic indices unequivocally relates to the clinical reaction after a sting challenge. We agree that more research is necessary.

Dr. Imbeau is correct that the variability in injected venom may influence an insect-sting reaction. Because of this, we tried to standardize the challenge procedure. Insects were taken in a uniform method from the same colony each season; they were induced to sting within 1 minute after they had been captured; and they were killed after they had stung. The stinging period was standardized at 30 seconds, after which all patients developed a local wheal and flare reaction. Although some variability from the insect venom cannot be ruled out, we think that this variability was not important in our challenge procedure.

Drs. Hatch, Davis, and Engler raise several concerns about our study design. First, in most patients, insect-sting allergy is a self-limiting disease [1, 2]. The selection of persons who are at risk for a future insect-sting anaphylactic reaction is therefore difficult. Skin tests and levels of specific IgE and IgG4 do not relate to the clinical reaction after a sting challenge [4].

Many variables influence the reaction after an insect sting. Insect specificity is not a problem in The Netherlands, but it certainly is in the United States. After a field sting, the amount of injected venom may vary widely. In view of this, the amount was optimally standardized in our challenge procedure and was probably greater than is delivered in most field stings. Finally, in contrast to the combination of objective and subjective symptoms after a field sting, an in-hospital challenge provides the opportunity to observe anaphylactic symptoms objectively. We have not encountered possible resensitization by the challenge or the immunotherapy.

We know that our therapeutic regimen after an anaphylactic reaction in the intensive-care unit is not applicable to field stings. All patients in our study who did not receive epinephrine, however, recovered quickly and uneventfully. In contrast, all three patients receiving standard doses of epinephrine developed cardiac arrhythmias immediately after this therapy. In addition, it is known that epinephrine therapy in anaphylaxis may be accompanied by an increased risk for mortality [5].

The aim of our sting-challenge procedure is to predict the reaction after future field stings. We mailed two yearly questionnaires to all (untreated) patients with a negative sting-challenge test: The preliminary results show approximately a 4% recurrence of (mild) anaphylactic symptoms after a field sting up to 10 years after the challenge. This rate is similar to the prevalence of insect-sting anaphylaxis in the general population [1-3]. This finding means that approximately 80% of persons will be saved from unnecessary, costly, and time-consuming immunotherapy as compared with a 4% risk of a mild anaphylactic reaction in untreated persons. Although we are well aware of the potential legal consequences in the United States regarding the sting-challenge procedure, we think that this risk/benefit ratio is low enough to advocate the standardized, in-hospital sting-challenge procedure as the current best criterion for selection of persons for venom immunotherapy.