Diagnostic clinical features of atopic dermatitis

Introduction

Atopic dermatitis (AD) is a common chronic disease which starts early in life and varies widely in clinical presentation at different ages and places. There is a uniform set of diagnostic clinical features which can be used in large population studies. Au-thors working in western countries on white races, have suggested many sets of diagnostic features.[1] Therefore keeping in mind the differences in envi-ronment and ethnicity of population in this part of the world, this study was planned to find out a mini-mum set of clinical features which will be diagnostic of AD.

Subjects and Methods

Seventy-three patients of AD with mild to moderate severity, not taking any drugs systemically and 71 age-matched controls suffering from other inflammatory dermatoses were selected. One case each of diaper rash, seborrheic dermatitis, perioral dermatitis, Henoch- Schonlein purpura, intertrigo, erythema multiforme, 2 cases each of impetigo, molluscum contagiosum, pompholyx, urticaria, con-tact dermatitis, 3 cases each of pityriasis rosea, tinea, stasis dermatitis, 3 cases each of scabies, viral warts, acne vulgaris, lichen simplex chronicus, 8 cases of lichen planus and 14 cases of psoriasis were stud-ied. The clinical details of all the cases were recorded on a proforma. All the cases were examined for 34 potentially useful clinical features, selected from dif-ferent studies,[1],[2],[3],[4],[5],[6],[7] including features for evaluation of photosensitivity which was noticed by Keong et al.[8]

At the end of the study, sensitivity and speci-ficity for each clinical feature was determined. Chi-square and relative values were calculated. Clinical features with X2 values less than 10 were omitted and the remaining were entered in correlation re-gression analysis. Using multiple regression technique a[4] minimum set of diagnostic clinical features for AD was determined.[4]

Results

One hundred and forty- four subjects including 73 cases and 71 controls were studied. Demographic variables of cases and controls were statistically similar, p> 0.01, [Table - 1].

The sensitivity and specificity of the clinical features were deter-mined, and Chisquare and relative values were calculated [Table - 2]. Clinical features with X2 value less than 10 were calculated from further analysis. The remaining 17 clinical features were found to be signifi-cantly associated with AD on corre-lation analysis, p< 0.0001 in all [Table - 3]. They were entered in multiple logistic regression analysis [Table - 4]. Six most useful clinical features which emerged from the regression analy-sis were: 1. presence of itch, 2. his-tory of flexural dermatitis, 3. history of dry skin, 4. family history of atopy, 5. personal history of diagnosed asthma and 6. visible flexural dermatitis. Photosen-sitivity was not found to be a significant feature. If three more clinical features such as 1. wool aggra-vating itch, 2. extensor dermatitis and 3. periorbital dermatitis are also added, they can favourably help in the diagnosis of difficult cases.

Discussion

This study reveals that diagnosis of AD and its differentiation from other inflammatory diseases can be done just by a set of six clinical features in which there is only one physical sign and the re-maining are historical features. Important clinical discriminators suggested as major signs by Hanifin and Rajka are included in them.[1],[2] Onset below two years was found to be important by Williams et al because a large number of their subjects were chlidren,[4] but this history can only be obtained from patients parents so it is difficult to be obtained in adults, Plenty of exposure to sunlight in this part of the world has not affected our study results because photosensitivity was not found to be a significant feature as stated by others also [9]. In this study it was found that addition of three more clinical fea-tures such as wool aggravating itch, extensor and periorbital dermatitis, to the six diagnostic fea-tures can favourably help in discriminating diffi-cult cases of AD from other inflammatory dis-eases. Thus the diagnostic clinical features de-rived from this study will be very useful for clini-cal studies on large population even in chang-ing environment.