Abstract: Objective To evaluate the diagnostic criteria of atopic dermatitis(AD). Methods The Hanifin and Rajka's criteria (gold standard), Williams'and Kang's criteria were compared by controlled investigation. Results The Williams' criteria showed the higher diagnostic efficiency in outpatient settings in comparison with the "gold standard", the sensitivity, specificity and Kappa value reaching 95.50%,97.52% and 0.93, respectively. The index" onset under the age of 2" in Williams'criteria could be applicable in any agegroup. In the population-based investigation, the sensitivity, specificity and Kappa value of Williams'criteria were 80.00%, 99.78% and 0.79, respectively. The criteria were suitable for point prevalence survey of the group aged 7～18 years. It was shown that an improved questionnaire based investigation had higher sensitivity(88.89%), specificity(99.78%) and Kappa value (0.82). The questionnaire consisted of pruritus plus two or more of the following four indexes:① the history of involvement of the skin creases;② a personal history of asthma or hay fever;③ a history of a dry skin;④ onset under the age of 2. Therefore, it could replace the conventional questionnaire for survey of the disease. Conclusions The Williams' criteria are suitable for clinical diagnosis in outpatient settings and epidemiological survey in populations with small sample size in our country. The modified criteria with questionnaire could also be used for mass survey of large sample.

Key words: Dermatitis,atopic, Diagnosis