[Atopy screening: prick multitest, total IgE or RAST? On the value of allergologic testing of the staff of an industrial bakery].

The predictive value of diagnostic tests for the recognition and differentiation of allergic diseases of the immediate type has been evaluated on the basis of the results of a non-selective allergologic-pneumologic examination of the staff of an industrial bakery. Out of the 314 subjects tested, 22 (7%) had a respiratory allergy against flour (flour rhinitis and/or flour asthma), and 21 (6.7%) a preexistent atopic disease without actual symptoms of flour allergy. Thus the predominance of atopy in the group examined was 13.7%. A prick-multitest was carried out in all 314 subjects on each forearm and allowed simultaneous and easy application of twice eight test substances ("atopy-screening" and "bakery-line"). 50% of the bakers with flour allergy also showed positive skin tests with non-occupational inhalative allergens (housedust, housedust mite, pollen) while only 9.2% of the subjects without respiratory allergy to flour did so. 17.1% of the heavily exposed, asymptomatic bakers showed a positive immediate reaction to the highly sensitive intracutaneous test with wheat flour. Here the problem of latent sensitization arises. With histamine the prick-multitest resulted - compared to the intracutaneous test - in 1.6% "false negative" and 3.2% only questionably positive results. Furthermore, 6.1% of the subjects showed a strong, and 3.2% a very strong, nonspecific reaction to the control solution glycerin. Despite these restrictions the prick-multitest proved far superior to measurement of total serum IgE for screening of atopy. Technical improvements have been made in the meantime by the manufacturer. Specific IgE was detected in RAST screening in 59 of the 306 subjects who underwent blood tests. The geometric mean of total IgE values was only 75 U/ml, and for the RAST-negative subjects 23.3 U/ml, while their 95% confidence level was 225 U/ml. This logarithmic-normal distribution of the measured values, together with the excessive normal-atopic overlap, makes it impossible to lay the test cutoff point in a way which results in an acceptable relation between sensitivity and specificity. For the population examined the sensitivity of the total IgE measurement was already below 50% at a cutoff point of 80 U/ml, while the positive predictive value did not yet reach 50% at 160 U/ml. This diagnostic test is at present certainly used too often and without sufficient awareness of its very limited predictive value. As a screening method the RAST is too expensive. Furthermore, the problem of a latent sensitization also exists for the RAST.