Age-Related Prevalence and Antibiotic Resistance of Pathogenic Staphylococci and Streptococci in Children With Infected Atopic Dermatitis at a Single-Specialty Center

Background  Skin staphylococci and streptococci are known to exacerbate atopic dermatitis, but the prevalence changes that occur with age are unknown. This study examined the age-related prevalence and antibiotic resistance of these pathogenic bacteria in children with atopic dermatitis and suspected skin infections.

Observations  Medical records of 150 children with atopic dermatitis referred to a regional center, who had skin swabs taken for suspected infection, were studied retrospectively. All patients carried Staphylococcus aureus. The prevalence of methicillin sodium–resistant (P = .05) and fusidic acid–resistant (P = .001) S aureus tripled from infancy to school age. Lancefield groups A and G streptococci were the other pathogens found. The prevalence of group A streptococci was highest in children aged 3 to 6 (53%), compared with 11% of infants and 21% of patients aged 9 to 16 (P = .002).

Conclusions  Significant differences in the age-related prevalence of group A streptococci skin carriage and antibiotic resistance of S aureus isolates occurred in this group of children with atopic dermatitis and suspected skin infections. Skin swabs to determine bacterial type and antibiotic sensitivities provide an important guide to antibiotic prescribing in these children.

STAPHYLOCOCCUS AUREUS and group A streptococci are commonly found on the skin of children and adults with atopic dermatitis (AD), and these organisms can exacerbate the disease.^1-7 The extent of their prevalence and their sensitivity to antibiotics with regard to age are unknown. Knowing the pattern of changes in bacterial flora during childhood could help in determining which antibiotics are effective for the treatment of secondary infections in patients with AD.⁸

One hundred fifty patients with AD (70 boys and 80 girls) whose eczema had been swabbed for bacterial pathogens were retrospectively studied by reviewing their medical records. Skin swabs were not routinely taken from every patient with AD attending our clinic, but were taken from skin areas of patients in whom there was a clinical suspicion of secondary impetigo. None of the children were taking oral antibiotics at the time of the skin swabs. Data on previous antibiotic use were not available. The diagnosis of AD was based on the U.K. Working Party's Diagnostic Criteria for Atopic Dermatitis.⁹ All patients were 16 or younger (median, 5 years; interquartile range, 2-8 years) and attended the Regional Paediatric AD Clinic of the Manchester University Department of Child Health at Booth Hall Children's Hospital, Manchester, UK, from September 1, 1999, to August 31, 2000.

Data were analyzed using commercially available software (Statistical Package for Social Sciences, version 9.0 for Windows; SPSS Inc, Chicago, Ill), and the frequency of bacterial species was determined for each age group. χ² Analysis was used to determine statistical significance between groups.

Staphylococcus aureus was present on the affected skin of all 150 patients studied. Pyogenic streptococci belonging to Lancefield groups A, B, C, or G were found in 80 patients (53%). The most common species of Streptococcus was group A (63 children, 42%), followed by group G (22 children, 15%), group B (12 children, 8%), and group C (11 children, 7%).

Age-related changes in the prevalence of group A streptococci on affected AD skin follow a bell-shaped distribution (Figure 1A). Group A streptococci were most prevalent in children aged 3 to 6 (23 [53%]of 43 children), with the lowest prevalence occurring in infants younger than 12 months (4 [11%]of 36 children) and in patients aged 9 to 16 (7 [21%] of 33 patients) (P = .002). In contrast, the prevalence of non–group A streptococci did not change significantly with age (P = .60) (Figure 1B).

Thirty-nine (51%) of 77 children older than 5 had fusidic acid–resistant S aureus, compared with 30 (38%) of 80 children aged 1 to 5 and 5 (14%) of 37 infants younger than 12 months (P<.001) (Figure 2A). A similar trend occurred with the prevalence of methicillin sodium–resistant S aureus; 15 (19%) of 78 children older than 5 had 1 or more isolates of methicillin-resistant S aureus, compared with 8 (10%) of 80 children aged 1 to 5 and 2 (6%) of 36 infants younger than 12 months (P = .05) (Figure 2B).

Children older than 5 had a higher prevalence of erythromycin ethylsuccinate–resistant S aureus (27 [35%] of 77 children) than younger children (26 [26%] of 100 children), but this trend was not significant (P = .20) (Figure 2C). Eleven (18%) of 62 children had erythromycin-resistant group A Streptococcus isolates. There were no significant changes in the prevalence of erythromycin-resistant group A Streptococcus with age (P = 1.0) (Figure 2C, inset). Twenty-five (49%) of 51 children with erythromycin-resistant staphylococci also had group A streptococci, although only 5 (20%) of these 25 were erythromycin-resistant group A Streptococcus isolates.

This is the first study to demonstrate age-related changes in pathogenic staphylococci and streptococci on the skin of children with AD. The key findings of the study are 2-fold. First, group A Streptococcus was not commonly found on the skin of infants and older children or adolescents with AD, but it was present in more than half of children aged 3 to 6. Second, although S aureus is ubiquitous on the skin of children with AD, the prevalence of fusidic acid and methicillin resistance tripled from infancy to adolescence.

The limitations of the study are that it is a retrospective study, including only a select group of children with AD referred to a specialty center because of the severe nature of their disease. Furthermore, a nonatopic group of children with impetigo was not available for comparison. Because impetigo is more common in children with AD, they would be expected to have more problems with antibiotic resistance as they grow older and are exposed to more antibiotics. Unfortunately, data were not available on previous antibiotic use in the children in the study. Local antibiotic preference and use would also be expected to affect antibiotic sensitivities of pathogenic bacteria.

What is the clinical relevance of these findings? The bell-shaped distribution of group A streptococci on the skin during childhood is similar to group A streptococcal throat carriage.¹⁰ Infection in school-age children with AD may be secondary to group A streptococci, rather than staphylococci, presumably acquired from schoolmates.

Group G streptococci were the second most common species of Streptococcus on the skin of patients.³ Such streptococci may cause cellulitis, pharyngitis, and systemic infection, particularly in persons with diabetes mellitus, alcohol abuse, or cancer.^11-13 This study demonstrates a difference in the epidemiology of group G compared with group A streptococci, with little change in prevalence from infancy to adolescence. The lack of age-related change in group G streptococci may reflect its opportunistic nature, which can become more of a problem in patients with impaired local or systemic immunity.

Staphylococcus aureus was found on eczematous skin of all patients with moderately severe AD.¹⁴ This study noted age-related changes in staphylococcal resistance to antibiotics, a tripling of resistance to fusidic acid and methicillin between infancy and adolescence, and a trend for increasing resistance of S aureus to erythromycin with age. Enterotoxins produced by staphylococci and streptococci may exacerbate AD,^15,16 and future studies should investigate whether the prevalence of staphylococci producing toxins parallels the usual decline in AD with age.

The study provides new information about the properties of bacteria that often exacerbate AD during childhood. Exacerbations of AD at the time of entry into kindergarten or school may be due to group A streptococci, rather than staphylococci. If recurrent exacerbations from group A streptococci occur after a course of oral penicillin V potassium, prophylactic treatment with penicillin V and treatment of other colonized family members may be necessary. In older children and adolescents, antibiotic-resistant staphylococci may be common. By school age, more than half of the children in our cohort had fusidic acid–resistant staphylococci. It is, therefore, our practice to avoid topical antibiotics in children with AD. We use oral antibiotics (eg, floxacillin) and routinely perform skin swabs to determine antibiotic sensitivities and to guide our antibiotic prescribing.¹⁷

Accepted for publication July 30, 2001.

Corresponding author: Peter D. Arkwright, MBBS, DPhil, Academic Unit of Child Health, Manchester University, Booth Hall Children's Hospital, Ward 3, Charlestown Road, Blackley, Manchester M9 7AA, England (e-mail: peter_arkwright@lineone.net).