Total IgE thresholds for detection of a positive specific IgE

Introduction: Specific IgE testing is a valuable tool for assessing allergy where skin prick testing is not practical, however its use is limited by cost. There is no recognised lower threshold of total IgE below which specific IgE testing would not be deemed appropriate. Objectives: To determine the level of serum total IgE which is predictive of a positive specific IgE, and whether this varies by allergen. Methods: Data was collected on all IgE measurements collected during 2011 in our institution. Those that had at least one specific IgE measured were selected (n=1518). ROC curves were used to detect the level of total IgE at which 90% and 95% sensitivity for any positive specific IgE was obtained. The data was analysed by specific allergen to determine whether the threshold changed. Results: Total serum IgE was predictive of a positive specific IgE with an AUC of 0.759. A sensitivity of 95% was obtained at 21.5IU/mL with a specificity of 32.6%, whereas 90% sensitivity was achieved at 33.5IU/mL giving 43.4% specificity. The most frequently requested specific IgEs were Aspergillus (n=802), House Dust Mite (n=466), grass pollen mix (n=373), cat dander (n=329), dog dander (n=179), tree pollen mix (n=142), peanut (n=126) and milk (n=108). ROC curve analysis for Aspergillus and House Dust Mite specific IgEs showed 95% sensitivity for aspergillus IgE was achieved at a total IgE of 52IU/mL, whereas HDM required a lower threshold of 18.5IU/mL to achieve the same sensitivity. Conclusions: The total IgE threshold for detection of positive specific IgEs varies by allergen. House Dust Mite specific IgEs present at low total IgE, whereas a positive Aspergillus specific IgE correlates with higher total IgE.