Editor,—I was very disappointed to see that the first contribution to the Controversy series was not written by a paediatrician. There are plenty of controversial topics in paediatrics, including the one cited. There are also plenty of paediatricians perfectly qualified to take part in informed debate about them, again including the topic cited. The absence of a contrasting viewpoint in the same issue suggested to me the feature should be called “Opinion” rather than “Controversy” because the article is not a balanced review of the current state of allergy practice.

The BPA and latterly RCPCH have championed for decades the holistic approach to the care of children. Paediatricians are best placed to assess the integrated needs of a child with medical problems. This principle is very relevant to developing areas of specialisation in which there is short supply of expert advice, such as in allergy. Paediatric allergists assess the impact of the diagnosis on many non-medical facets of a child's life, including family lifestyle, integration into schools and peer groups, and the facilitation of appropriate independence from parental supervision.

It is tiring to have to rehearse the arguments for the adequate protection of subjects at risk of anaphylaxis. Epinephrine (as all doctors should now be calling adrenaline) is not the only help given in clinic to families with an allergic child. It is part of the integrated management plan, which appears to be effective1 though difficult to measure.2

It is very hard to prove that epinephrine saves lives and I agree that the notional “number needed to treat” with epinephrine to prevent a death from anaphylaxis is very high. Unsworth's title suggests that this “very high number” (my phrase) is too high. How has he measured that? What is too many? He quotes a prevalence of about 1% of Americans having peanut allergy. That is approximately 3 million subjects. We do not restrict insulin syringes to just a few insulin dependent diabetics because diabetes is so common that we cannot adequately care for all of them. Every allergic child has the right to best available care, which is not restricted to the first 100 through the clinic door (if they can find an allergy clinic).

Laparotomy will not save every patient with a leaking aortic aneurysm and epinephrine will not save every person who has anaphylaxis. Anaphylaxis is a critical situation in which prompt administration of epinephrine may (but occasionally may not) save a life. I think it unarguable that it is better to self treat and probably survive than not self treat and possibly die. Unsworth quotes one early paper about anaphylaxis from the US3 and more recent British data.4 ,5 These papers all say to me more that epinephrine is underused due to unavailability or inappropriate training and patient confusion, rather than that epinephrine is useless or dangerous. Most subjects did not have epinephrine available. Several of the deaths reported by Pumphrey5 were due to incorrect use of available epinephrine. In addition, epinephrine appears to be more dangerous in the hands of doctors who give it IV than in the hands of allergic subjects who self treat IM. I recommend your readers look at the report on the latest series of food related deaths.6

In the absence of any perfect predictive test, allergists are confined to basing risk of future severe reactions on just a few variables. The first is a history of previous severe reactions.3 The majority of peanut allergics have had a severe reaction in the past7 ,8 and more than 60% have asthma, the second known association with severe reactions.3 ,7 According to current opinion, then, even after just one reaction to peanut most subjects are considered at risk of severe future reactions. Many minor reactors to peanut progress to more severe reactions7 and new data confirm this convincingly.9 I do not think there are adequate data to change my practice from needing a very good reason not to prescribe epinephrine to most (but not all) subjects who have reacted to peanut, a food known to be associated with a risk of a severe allergic reaction.

Doctors must remember epinephrine is prescribed to be available for response to infrequent exposure at an uncertain future date, not to be taken four times a day. I have referred to this in the past10 as analogous to wearing a seatbelt on every car trip, every day, even though a serious car accident is unlikely on any individual day.

Unsworth is not up to date in his comments about the diagnosis of IgE mediated allergy. There are strong data from huge series of challenges, about the positive and negative predictive values of the tests used in allergy clinics.11-13 Unsworth does not even mention formal challenges, the cornerstone of modern food allergy practice. No allergist would prescribe an epinephrine kit on the basis of a positive SPT in the absence of a significant history or formal challenge.14

Children and adults at risk of food related anaphylaxis have enough of life's pleasures denied to them. The provision of epinephrine kits allows normal life to go on, involving school, overnight stays at friends, camping, and other normal activities of childhood. Anecdotally, parents seem to me less stressed when they leave clinic with information (however awful the scenarios described) and response strategies than when they arrive. I have never met a parent who reported being more scared of the epinephrine kits than of the prospect of allergen exposure (with or without epinephrine available).

Families must be taught when to use epinephrine and how to use autoinjectors. Until doctors can tell families that anaphylaxis will never happen we should continue to empower families, ensuring they are ready to respond as best they can to the disaster that allergen exposure represents. When anyone develops a real treatment for food related anaphylaxis I can stop prescribing epinephrine kits to people who currently need them.