Km. Saarinen et al., Diagnostic value of skin-prick and patch tests and serum eosinophil cationic protein and cow's milk-specific IgE in infants with cow's milk allergy, CLIN EXP AL, 31(3), 2001, pp. 423-429
Background The diagnosis of cow's milk allergy is based on a clinical respo
nse to an elimination-challenge test with cow's milk.
Objective We studied the usefulness of the skin-prick and patch tests and m
easurement of cow's milk-specific IgE and eosinophil cationic protein in se
rum as diagnostic tools for cow's milk allergy in a cohort of 6209 unselect
ed infants followed from birth for the development of cow's milk allergy.
Results Of the 239 infants challenged with cow's milk, 118 showed a positiv
e and 121 a negative response at a mean age of 6.9 months. A positive react
ion to a skin-prick test with cow's milk (greater than or equal to 3 mm) wa
s seen in 72 (61%) and 29 (24%) infants with positive and negative challeng
es, elevated serum cow's milk-specific IgE (greater than or equal to 0.7 kU
/L) in 52 (45%) and 15 (13%) infants, a positive reaction to patch test wit
h cow's milk protein fractions in 26 (26%) and eight (8%) infants, and elev
ated serum eosinophil cationic protein (greater than or equal to 20 mug/L)
in 22 (21%) and seven (13%) infants, respectively. Parallel use of the four
tests with the above-mentioned cut-off values correctly classified 73% of
the infants with a sensitivity of 0.76 and a specificity of 0.67. An immedi
ate reaction to cow's milk challenge correlated with skin prick test positi
vity and elevated serum milk-specific IgE, and tended to correlate with pat
ch test positivity.
Conclusions No single test or parallel use of the four tests could predict
the challenge outcome acceptably in this prospectively followed, unselected
cohort of 6209 infants. A positive reaction to one or more tests needs to
be confirmed by a challenge test and a negative response to all four tests
does not rule out the possibility of cow's milk allergy.