An NHSR determination is made by a qualified reviewer (not the investigator). Common examples include quality improvement, case reports of three or fewer patients, and analyses of fully de-identified or coded data without access to identifiers.