Methodological characteristic of selected studies.

We documented the number of cases, study years, country, data source, and age criteria. We noted data sources, including “clinical data” (eg, EHR data), physician or family “surveys,” “other” data sources (eg, genetic labs, family support groups, education/rehab facilities, prospective study, registry, claims database), “multiple data sources,” and “unknown/unspecified” data source. We noted if diagnostic criteria were “described” (in text), “referenced” (cited another study), “none” (eg, used clinician or family’s diagnosis), or “unknown/unspecified.” We reported if studies had a process for removing duplicates, and coded as “yes,” “unknown/unspecified,” or “single data source” (ie, data from 1 hospital and therefore less likely to have duplicates). We reported when studies corrected for under-ascertainment.