METHODS
The methodology is described elsewhere (Ferrari et al 201913). Briefly, we included the HDRs nationwide of all adolescents hospitalized in the years 2002–2015. Among them, we selected cases with at least one hospital stay for a diagnosis of cutaneous melanoma (ICD-9-CM diagnostic code = 172* “Malignant melanoma of skin” and/or ICD-9-CM diagnostic code = V1082 “Personal history of malignant melanoma of skin”) found in all the six diagnostic fields (main diagnosis and up to 5 secondary diagnosis) of the HDR. The first hospitalization indicating a diagnosis of cutaneous melanoma was defined as the index hospitalization. To select only incident cases, we excluded patients who were hospitalized during the 5 years or more before their index hospitalization (i.e. prevalent cases) for the same diagnosis or for diagnosis referring to cancer or its treatment. The diagnostic procedures and main treatments for melanoma were established searching in all the six procedural fields (main procedure and up to 5 secondary procedures) melanoma-specific ICD-9-CM procedure codes: melanoma surgical interventions were divided between radical excision (ICD-9-CM procedure code=86.4 “Radical excision of skin lesion”) and local excision (ICD-9-CM procedure code=86.3 “Other local excision or destruction of lesion or tissue of skin and subcutaneous”). We defined as main treatments those begun within 12 months of the diagnosis. Thus, study cohort starts in 2007 rather than 2002 and ends in 2014 rather than 2015. The treatment of each adolescent with melanoma was associated with only one hospital. When patients were treated at different hospitals, we assigned them to the hospital performing the radical excision or, if a radical excision was not performed, we assigned them to the hospital performing the local excision. Patients without any hospitalization for neither radical or local excision were assigned to the hospital where they were diagnosed (this applied to 12% of the adolescents with melanoma). The hospitals were divided into those having a unit affiliated to the national pediatric oncology network Associazione Italiana Ematologia Oncologia Pediatrica (AIEOP) - Italian Association of Pediatric Hematology and Oncology – and those without it (non-AIEOP hospitals).
Ethical approval for this study was obtained from the Ethical Commitee of Fondazione IRCCS Istituto Nazionale dei Tumori (N. INT 132/17).