Atopic dermatitis affects 15 to 20% of preschool children in western countries [
For the current study, a questionnaire was based on the existing ISAAC (International Study of Asthma and Allergies in Childhood) questions for school children [
A case control design similar to that used by Williams et al. in his validation study on atopic dermatitis was chosen [
Study procedures.
The qualitative part of the investigation was carried out by means of a semistructured, one-to-one interview with the parents. The interview started with a general discussion regarding the understanding of the eczema questions and relevance of named issues on infants’ and preschool children’s eczema. Each interview was conducted by the same physician, the first author of this paper, to ensure consistency. The interview was transcribed and approved by the responders.
The regional ethical committee in Uppsala approved the protocol for the study, Dnr-C2007/41. Written consent was obtained 2 weeks prior to study start.
Inclusion criteria for index cases were eczema, treated either with emollients or topical glucocorticoids. Inclusion criteria for controls were an age of two month to six years and no eczema. Exclusion criteria were eczema treated with systemic treatment, other chronic diseases such as diabetes, acute infections, or if parents did not understand or speak Swedish fluently. A power analysis showed we would need 60 children to assess sensitivity and specificity with a power of 0.8 and alpha 0.05 [
Atopic dermatitis was defined by Hanifin and Rajka as
The three key questions are as follows. Does your child have or has your child had a If Yes, has this caused Has this red rash/eczema affected any of the
The children had to answer “yes” in each of the three questions to be classified as a questionnaire diagnosed eczema.
A questionnaire-diagnosed eczema was compared to a physician’s diagnosis which was the gold standard. Sensitivity and specificity as well as the positive predictive value (PPV) and negative predictive value (NPV) of the sample population and stratified age groups were assessed.
PPV was calculated as (prevalence × sensitivity)/(prevalence × sensitivity + (1 − prevalence) × (1 − specificity)) [
Data was analysed according to content [
Eczema occurred in 35 children, who reported dry skin twice as much as children without eczema (Table
Baseline characteristics of children with and without eczema due to specific criteria.
Patients’ diagnose status | |||
Eczemaa | Noneczema | ||
Number, | 35 (58.3) | 25 (41.7) | |
Age (month) | |||
Mean (SD) | 25.96 (±19.1) | 25.68 (±18.3) | |
Median (range) | 18.0 (4–66) | 19.0 (2–60) | |
Sex, | |||
Female | 19 (54.3) | 8 (32.0) | |
Male | 16 (45.7) | 17 (68.0) | |
Heredity, | |||
Yes | 27 (77.1) | 15 (60.0) | |
No | 8 (22.9) | 10 (40.0) | |
Dry skin, | |||
Yes | 27 (81.8)* | 6 (18.2)* | |
No | 8 (29.63)* | 19 (70.4)* | |
SCORAD Scoreb, | |||
Severe | 6 (17.1) | ||
Moderate | 15 (42.9) | ||
Mild | 14 (40.0) |
aPhysician diagnose, based on Hanifin and Rajka’s criteria.
bEczema severity assessed with SCORing Atopic Dermatitis.
*
A combination of affirmative answers to all three key questions, rash, itch, and location, predicted clinical eczema with good sensitivity 0.91 (95% CI 0.77 to 0.98) and specificity 1 (95% CI 0.86 to 1).
Predictive values were PPV = 1 and NPV = 0.976. Diagnostic accuracy by age groups was similar (Table
Validation of the used questionnaire by comparison of parental reported eczema with physician diagnose of the children by age group.
Patients’ age (month) | Psychometric measures | |||
Sensitivity | Specificity | Positive | Negative | |
0 to 23 | 0.91 (0.70–0.99) | 1 (0.81–1)b | 1 | 0.97 |
≥24 | 0.93 (0.66–1.00) | 1 (0.76–1)b | 1 | 0.98 |
0 to 66 (total) | 0.914 (0.77–0.98) | 1 (0.89–1)b | 1 | 0.98 |
bSingle sided 95% CI.
Most parents judged the questions to be understandable and suitable—“
Of 35 parents to children with eczema, 21 expressed feelings of stress and worry about their child’s health state and felt responsible for it. Four parents reported not wishing their children to be diagnosed with eczema at all because of fear of cortisone treatment. Itching and awakening due to itching were regarded as the most burdensome symptoms.
With three key questions, atopic dermatitis in preschool children can be identified via a parental questionnaire. The high sensitivity and specificity indicate that this questionnaire is a valuable diagnostic tool. Validated eczema questionnaires have been available for school children only. But the incidence of eczema is highest in preschool children. Our questionnaire provides a feasible diagnostic tool with high diagnostic precision, which can add to the field of epidemiological research in early childhood eczema.
The semistructured interview confirmed high face and content validity, which might be one reason for the high proportion of children with detectable eczema. The PPV of the questionnaire is the proportion of children with a positive result who actually had physician-diagnosed eczema. Predictive values are, however, related to the prevalence of eczema in the population, which could not be assessed with this study design. The high prevalence of eczema, 22%, in the population was derived from data from the same county [
It is important to estimate the severity of eczema, both for the description of the study population and for an assessment of generalizability. A score combining an assessment of disease extent with clinical features, duration plus intensity (SCORAD) was used [
The strength of the study is the population-based setting, as the questionnaire is intended to be used in a similar setting. Because all children were registered in a preventive care register and almost all invited children participated, it is unlikely that selection bias occurred, even though the sample was not randomly selected. Assessing the diagnostic accuracy of the diagnostic test on a sample of the general preschool child population allows generalizing results. The narrow confidence intervals suggest that the sample size was adequate and that estimates of the population value are within a reasonable range. Compared to a standardized skin examination protocol, the ISAAC questionnaire performed well in predicting eczema prevalence at the population level. However, on an individual level, a high proportion of flexural eczema was not confirmed by skin examination [
This study makes a contribution in solving diagnostic problems where there has been a lack of diagnostic questionnaires for small children. The questionnaire can be used in a population-based setting and across different severity groups of eczema. It is, however, important to take cultural aspects into account.
SCORing Atopic Dermatitis
Positive predictive value
Negative predictive value.