The present invention relates to diagnosis, prognosis, risk assessment, and/or risk stratification of an adverse event, particularly mortality, of a subject. The invention relates to a method that comprises determining a level of proadrenomedullin (proADM) in a sample of said subject, and wherein said level of proADM is indicative of said adverse event of said subject, wherein said level of proADM is compared to a reference level of proADM; and wherein said adverse event of said subject is identified based on the comparison. The invention further relates to kits for carrying out the methods of the invention.