OBJECTIVES:

IH, the most common tumor of infancy, has the unique ability to involute spontaneously after proliferation. Unfortunately, a subset rapidly develops complications, resulting in pain, functional impairment, life threatening or permanent disfigurement. A scoring tool (12 Yes/No items) supporting primary clinicians (PCs) to identify patients requiring referral to expert centers, has been developed by IH experts from 8 European countries. The aim of the present study is to validate the IH Referral Score (IHReS).

METHODS:

A first board of experts has selected 42 relevant IH photos, corresponding to the most common cases seen by PCs, with a brief clinical description (gender, age, date of apparition of the IH, complications if any). An independent experts board has defined the suitable referral decision (Gold Standard) before having access to the IHReS. The tool has further been used by non-expert PCs (34 at this stage) from 3 countries (Italy, Spain, France) to classify the 42 clinical cases, with a re-test step 7 days later.

RESULTS:

IHReS enabled cases classification with 88.5% sensibility and 71.9% specificity (optimal cut-off score was determined using ROC curve and the Youden index). Cronbach's alpha correlation between items of 0.61 assessed the internal consistency. Fleiss' kappa, used to assess the inter-rater agreement, was 0.50 (Moderate agreement) and increased to 0.62 (Substantial agreement) in the re-test step for PCs (respectively 0.66 and 0.73 for experts). Kappa of Cohen, used to assess the intra-rater agreement between test and re-test steps, was 0.67 for PCs (0.77 for experts).

CONCLUSIONS:

IHReS has proven to be a simple and validated algorithm, with good intrinsic properties, reliability (inter-rater) and repeatability (intra-rater). Learning curve to use the tool is fast and IHReS can easily be used by primary care physicians to help them to refer, when necessary, the right infantile hemangiomas to expert centers.