Results

Participation

The panel of 14 members (Table 1) comprised 11 doctors (seven specialists and four GPs with an interest in allergy) and three paediatric dietitians; nine were female. The mean age was 46.7 years (SD 5.4, range 40 to 57). Most (11) worked in England, with one each from Jersey, Scotland, and the Republic of Ireland.
Response rates to the three survey rounds were 100%, 93% and 100% respectively. One partial round 2 response, saved and not submitted due to lack of time, was excluded after discussion with the panel member. Two panellists were unable to attend the workshop.

Clinical symptoms

The results for rounds one and two, and the final agreed list, around which symptoms/timescales are indicative of an immediate (IgE) or delayed (non-IgE) allergic reaction are shown in Table 2.
In round one, opinions were sought on the 14 skin, gut or airways symptoms. Of these, consensus emerged on 12 immediate and 12 delayed allergy symptoms/timescales.
For round two, “abdominal pain/bloating, including ‘colic”’ was reworded to “abdominal pain”; “Noisy or difficulty breathing” was replaced by “Hoarse voice or cry” and “Swelling of tongue or airway/stridor”; and for delayed only, “Rash, redness of skin” replaced by “Rough or bumpy, itchy skin” and “New eczema flare”. After round two, there was no consensus for two symptoms: “Nasal congestion or runny nose” for immediate and delayed allergy; and “Rough or bumpy, itchy skin” for delayed allergy.
Allergy symptoms were not discussed further at the workshop but it was agreed in this write-up to reword “Rash, redness of skin”, accepted at round one as an immediate symptom, to “Worsening of eczema (new flare)”, to be consistent with delayed, and more appropriate for non-white skin. The final list comprises 12 immediate and 7 delayed allergy symptoms (Table S3).

SPT allergens

In round one, the panel had a choice of specific foods, commercial reagents or other. Consensus was reached at round one (Table S2) for wheat (commercial wheat extract, 86%) and at round two for soya (commercial soya extract, 83%). There was consensus at round one for egg (commercial extract, 86%) but no further consensus between commercial whole egg extract and commercial egg white extract (Table S2) in round two. There was no consensus for milk.

SPT thresholds

For each food, the panel were asked to interpret wheal sizes ranging from 0 to 9 mm as negative, equivocal or positive. There was consensus for all study foods for 0-1 mm being negative and 5 mm and above as positive (Table 3). Uncertainties remained in the 2-4 mm range, with varying levels of consensus depending on the food.

Dietary advice

The dietary advice findings from round two were presented to participants pictorially, as a flowchart. This depicted the dietary advice considered most appropriate for the 12 different possible combinations (“scenarios”) of symptoms (ingested recently – yes/no; symptoms – none, immediate or delayed) and SPT results (negative or positive) (figure S1) for a child who had previously ingested the food. For each scenario, the advice could vary by study food. There was consensus across all foods in five scenarios (e.g. SPT – negative, ingested recently – yes, symptoms – none = not allergic, include); no consensus for any food in three scenarios (e.g. SPT – positive, ingested recently – no; symptoms – delayed); and consensus for some but not all foods in the remaining four scenarios.
The seven scenarios where there was disagreement for one or more foods were carried forward into round three. Two additional scenarios were introduced into round three, for a child who had never previously ingested the study food and whose SPT was negative or positive. The combined findings from rounds two and three are presented in figure S2. There was consensus on 39 (70%) of the 56 possible food scenarios: consensus across all foods in seven (50%) scenarios, partial consensus in five (36%) scenarios and no consensus for any food in two (14%) scenarios.
All scenarios were reviewed in the workshop but discussion focused on, and resolved, all the areas of prior disagreement, with the resulting flow chart shown in Figure 1. Based on feedback, “positive” SPTs were relabelled as “sensitised”.