Clinical features and Staphylococcus aureus colonization/infection in childhood atopic dermatitis.

INTRODUCTION: Staphylococcus aureus (SA) colonization/infection is important in the pathophysiology of childhood atopic dermatitis (AD). This study evaluated which clinical features may predict presence of SA colonization/infection and reviewed antimicrobial sensitivity of SA in patients with AD. METHODS: The associations between bacteriologic culture results of skin swabs (taken at the most severely affected area and at the antecubital fossa) and SCORing-Atopic-Dermatitis (SCORAD), skin hydration, transepidermal water loss (TEWL), and quality of life were evaluated. RESULTS: Moderate-to-heavy growth of SA was present in 31% of the swabs of the most severe area and in 16% of the flexural (antecubital fossae) areas of 95 AD patients (12.5 ± 4.8 years). Binomial logistic regression showed moderate-to-heavy growth of SA in the severe area were associated with objective SCORAD (p = 0.004) and lesion intensity [erythema (p = 0.022) and lichenification (p = 0.035)]; and excoriation (p = 0.024) and TEWL (p = 0.009) in the antecubital fossa. The relative risk of isolating moderate-to-heavy growth of SA in the most affected area in patients with severe disease (objective SCORAD >40) is 2.73 (1.43-5.21, p = 0.001). Any growth of SA in either swab sites was associated with objective SCORAD and lesion intensity (p = 0.001-0.019). SA had no association with quality of life and other clinical parameters. All specimens of methicillin-sensitive SA were sensitive to cloxacillin. All methicillin-resistant SA (MRSA) (5.7%) was sensitive to co-trimoxazole and fusidic acid. CONCLUSIONS: Clinical features, especially severity and lesion intensity, are useful in "predicting" moderate-to-heavy SA colonization/infection in AD patients. Cloxacillin has a favorable sensitivity profile for MSSA, and co-trimoxazole and fusidic acid for MRSA. As colonization and infection are ambiguous and potentially overlapping clinical states, we recommend to abandon these terms and propose to describe quantitatively/semi-quantitatively SA isolation as none, mild, scanty, moderate or heavy growth instead in clinical trials.