Score Model Effectively Predicts Mortality Risk in Dementia

Close-up of hand of senior on hand of dying elderly person as sign of support during sickness
Eleven acute and chronic factors were identified for building a risk score model for predicting mortality in patients with dementia.

A new predictive score model may be the first tool to rely on the same clinical factors to determine both short- and long-term mortality risks among patients with dementia, according to research published in the Journal of Clinical Psychiatry. Previous prediction models have been insufficient in predicting both short- and long-term mortality rates, and have worked only in specific populations.

To develop an informative tool to predict mortality rates for people with dementia at 6 months and 1, 2, 3, and 5 years, researchers collected data from 6556 patients age 65 years or older who received an ICD-9-CM dementia diagnosis and were enrolled in Taiwan’s National Health Insurance Research Database. These patients were divided into a derivation group (n=4371) and a validation group (n=2185). Potential risk factors included patient characteristics and comorbidities. Researchers used a regression coefficient-based scoring approach to develop the mortality-predicting risk scores.

Related Articles

At the end of the follow up period, 1693 of 4371 people with dementia (38.7%) from the derivation cohort were deceased; the mean duration of follow-up was 6.26 years. Researchers identified the following11 acute and chronic factors for creating the predictive score model: age, sex, chronic obstructive pulmonary disease, chronic heart failure, chronic kidney disease, diabetes, myocardial infarction, cancer, recent lung infection, recent urinary tract infection, and recent nasogastric intubation. The model, which produced scores from 0 to 24 points (higher scores indicated higher mortality), showed good predictive power across various life expectancies.

The study was limited by a lack of information regarding smoking status, physical activity, cognitive ability, disease severity, family history, education, activities of daily living, genetic factors, or environmental factors.

Researchers stated, “Use of this model may enable physicians, specialists, other health-care providers, patients with dementia, and patients’ families to acquire early prognostic information and thereby make decisions regarding further intervention.” In addition, this model may allow medical staff and policy makers to better coordinate individualized medical and/or psychological services for those with dementia as well as to evaluate the efficacy of interventions and clinical trials among higher-risk patients.


Cheng CM, Chang WH, Chiu YC, et al. Risk score for predicting mortality in people with dementia: a nationwide, population-based cohort study in Taiwan with 11 years of follow-up. J Clin Psychiatry. 2019;80(4):18m12629.