In our study, a significant difference in the occurrence of reported food allergy emerged across the different AD severity categories. This seems to confirm previous studies in which severe vs mild to moderate AD in children was associated with a higher one-year prevalence of food allergy (27.0% vs 14.1%) . Skin barrier defects play a pivotal role in the pathogenesis of AD, and our findings mirror previously-published studies suggesting that skin barrier impairment, coupled with cutaneous allergen exposure, may be crucial in the development of food allergy and allergen sensitization through the skin (i.e., “dual-allergen exposure hypothesis”) [15, 16]. Little evidence is currently available on the association between AD severity and passive smoking in children. In a study on 100 Greek children, parental passive smoking was associated with severe AD (aOR: 4.6; 95% CI: 1.0–22.1; p = 0.050). However, a weaker association with severe AD was found, if compared to the other risk factors evaluated (excessive cleanliness p < 0.001, aOR: 59.4; 95% CI: 10.9–322.6; Radioallergosorbent Test, RAST > 0.7 kU/l p = 0.014, aOR: 7.9; 95% CI: 1.5–41.0) . In our study, no differences in the occurrence of passive smoking exposure and reported inhalant allergy were found across the AD severity categories (p = 0.92 and 0.15, respectively).
As for the treatment patterns, 9.8% of patients with severe AD had not been treated with TCS, TCI or any systemic steroids in the previous month. Such a figure underlines the magnitude of undertreatment among patients with severe AD, at least in those attending the Comano Thermal Spring Water Center. In a country like Italy, where the National Health System guarantees free unlimited access to pediatric primary care, AD undertreatment might largely be explained by limited access to medical care due to personal beliefs (e.g., seeking alternative treatments) or by low adherence (due to “corticophobia”, forgetfulness or other reasons). Additionally, 36.6% of patients with severe AD had been treated with systemic steroids in the previous month. Interestingly, although short courses of systemic steroids are an effective and inexpensive treatment, the latest European and Italian guidelines on the management of AD recommend a limited use because of their unfavorable risk-benefit ratio. Moreover, it is stated there that the indication for oral steroids in children should be handled even more cautiously than in adults [3, 10, 18,19,20]. It is also worth noting the limited use of emollients in our study population: only 55.1% of patients resorted to them in the previous month, their occurrence being slightly higher in patients with severe disease. Few studies have reported the prevalence in the use of emollients in pediatric patients with AD. A Polish study  has showed a prevalence of emollients use of 82% in AD patients. Still, the small study population (22 patients, adults and children) makes it difficult to generalize these findings. It is important to point out that emollients are an essential element in the treatment of AD. In fact, in several studies, regular use of emollients has proved to achieve a short-term steroid sparing effect in mild to moderate AD  and long-term use of emollients is recommended for the maintenance of stable disease in the latest European and Italian guidelines [3, 18,19,20]. Health economic analyses showed that the use of emollients is cost-effective compared with a no-treatment strategy . Several factors may explain our findings. First, as the Italian National Health System does not cover emollients, their cost could represent a hindrance to their use. Secondly, parents’ poor knowledge on the long-term efficacy of emollients in preventing AD flare-ups may also lead to underuse. Indeed, a recent interview study  showed that parents had mixed views on long-term emollient use to prevent exacerbations. The authors concluded that providing a rationale for long-term emollient use could help improve adherence. Educational programs and adequate time to allow for patient education during routine visits may also help raise awareness about the utility of emollients in managing AD.
Finally, it is interesting to point out in our study how surprisingly only 27.2% of patients on elimination diets had a reported food allergy. It is also interesting to highlight how 5.0% of patients who were not on elimination diets in our study had a reported food allergy. While elimination diets are indeed critical in AD patients with food allergy, it seems fundamental for pediatricians and dermatologists to promote an evidence-based management of suspected adverse reactions to food [25,26,27] based on a rigorous diagnostic allergy work-up carried out according to the current guidelines . Indeed, an incorrect management of food allergy can lead to inappropriate and dangerous dietary restrictions - especially in children, who are very vulnerable to nutritional deficiencies  or to the risk of reactions [30,31,32].
The main limitations of our study are mostly due to the clinical setting in which it was carried out, i.e., admission visits for balneotherapy. Indeed, the characteristics and treatment patterns in pediatric AD cases recorded at the Comano Thermal Spring Water Center may not be entirely representative of the general Italian pediatric population. Moreover, food and inhalant allergy was reported, but no rigorous diagnostic allergy work-up (based on the use of prick test, specific IgE or food challenge) nor other laboratory investigations were performed at the Comano Thermal Spring Water Center, which is not the appropriate clinical setting to carry them out. Lastly, data on antihistamine use were not collected. However, the large population size is undoubtedly the strongest point of our study, as it enabled us to describe the characteristics and treatment patterns in pediatric AD cases.