Specific immunotherapy is able to protect insect allergic patients aga
inst life-threatening sting reactions. Initiation of immunotherapy can
be decided on the basis of challenge tests. However, at least for chi
ldhood no data were available about the longterm predictive value of s
ting challenge tests. Between 1988 and 1991, in 92 bee venom allergic
children two sequential bee sting challenges one to three weeks apart
from each other were performed using living bees or Reless bee-venom.
The predictive ability of the initial test results regarding the natur
al follow-up history was now investigated. In August 1996 all 92 famil
ies could be re-contacted. For 48 children (52%) at least one natural
sting exposure was reported, of whom twelve patients were treated with
specific immunotherapy. One of them experienced a severe and one a mi
ld systemic reaction. Among the 36 re-stung patients without immunothe
rapy, two children experienced mild systemic reactions whereas in 34 c
hildren only a local reaction was reported. Thus, the negative predict
ive value of sequential bee sting challenge tests without systemic rea
ctions remains at a level of 94.4% (95%-confidence interval: 81.3-99.0
%) even over a period of more than five years.