BACKGROUND:
The prevalence of latex-specific IgE computed from the results of serologic assays is commonly thought to reflect, to a greater or lesser extent, the prevalence of latex allergy and its implied risk.
OBJECTIVE:
The study examines how imperfect test specificity of in vitro assays influences the precision of latex allergy prevalence that it estimates.
METHODS:
Various models encompassing a range of hypothetical test sensitivity and specificity values are investigated to gauge their influence on the estimate of latex allergy prevalence. The models examine these interactions in situations of high or low allergy prevalence.
RESULTS:
Serologic latex diagnostic assays with test specificity within the range of those of commercially available assays can greatly overestimate prevalence where the true prevalence is low (eg, of the order of one in 100 or one in 1,000). A formula to correct for errors in prevalence estimates arising from imperfect test sensitivity and specificity of an in vitro assay is presented.
CONCLUSION:
While serologic assays for latex IgE pose few hazards to the patient and are useful for confirming the diagnosis of latex allergy, the test results may vastly overestimate the true prevalence of latex allergy and its associated risks in situations where latex allergy is actually rare.
Spina bifida children have a high prevalence of latex allergy in studies reported from Europe and the USA. This study investigated the prevalence of latex allergy in a cohort of 24 spina bifida children at the Red Cross Children's Hospital from Cape Town, South Africa. The children were investigated using a detailed questionnaire, skin prick tests (ALK-Abello), ImmunoCap RASTs, Western blotting and ELISA, using the purified latex proteins Hev b1 and Hev b3 and whole latex preparation. A low overall prevalence of latex sensitization of 16.7% was found in the children. Children who were sensitive reacted to water insoluble to Hev b1 and Hev b3 proteins. The low prevalence of latex sensitization in the South African children may not be entirely explained by stringent latex avoidance. The children were from a low socioeconomic social status and 'hygiene' and other factors should be considered.