Cheng CM, et al. - Researchers sought to develop an informative tool that may aid in predicting 6-month, 1-year, 2-year, 3-year, and 5-year mortality rates for dementia patients. From Taiwan’s National Health Insurance Research Database, they identified 6,556 patients aged ≥ 65 years who received ICD-9-CM diagnoses of dementia between 2002 and 2009 and followed them up until the end of 2013. They assessed patient characteristics and comorbidities that were considered potential risk factors for mortality and developed mortality-predicting risk scores using a regression coefficient-based scoring approach. The patients were randomly divided into a derivation cohort (n = 4,371) and validation cohort (n = 2,185). Death was reported for 1,693 of the 4,371 dementia patients (38.7%) in the derivation cohort by the end of the study. The predictive score model was created using 11 acute and chronic factors and produced scores from 0 to 24 points (higher scores indicated higher mortality). The developed predictive score model seemed to be the first tool that determines both short- and long-term mortality risks in patients with dementia using the same clinical factors. It displayed good predictive power for various life expectancies (area under receiver operating characteristic curve: 6-month = 0.852, 1-year = 0.779, 2-year = 0.725, 3-year = 0.721, 5-year = 0.703) and good calibration in the validation cohort.