Predicting early mortality among elderly patients hospitalised in medical wards via emergency department: the SAFES cohort study.

Details

Serval ID
serval:BIB_65AF55A413BB
Type
Article: article from journal or magazin.
Collection
Publications
Title
Predicting early mortality among elderly patients hospitalised in medical wards via emergency department: the SAFES cohort study.
Journal
Journal of Nutrition, Health and Aging
Author(s)
Drame M., Jovenin N., Novella J.L., Lang P.O., Somme D., Laniece I., Voisin T., Blanc P., Couturier P., Gauvain J.B., Blanchard F., Jolly D.
ISSN
1279-7707 (Print)
ISSN-L
1279-7707
Publication state
Published
Issued date
2008
Peer-reviewed
Oui
Volume
12
Number
8
Pages
599-604
Language
english
Notes
Publication types: Journal Article ; Multicenter Study ; Research Support, Non-U.S. Gov't Publication Status: ppublish
Abstract
OBJECTIVES: The aim of the study was, by early identification of deleterious prognostic factors that are open to remediation, to be in a position to assign elderly patients to different mortality risk groups to improve management.
DESIGN: Prospective multicentre cohort.
SETTING: Nine French teaching hospitals.
PARTICIPANTS: One thousand three hundred and six (1 306) patients aged 75 and over, hospitalised after having passed through Emergency Department (ED).
MEASUREMENTS: Patients were assessed using Comprehensive Geriatric Assessment (CGA) tools. A Cox survival analysis was performed to identify prognostic variables for six-week mortality. Receiver Operating Characteristics analysis was used to study the discriminant power of the model. A mortality risk score is proposed to define three risk groups for six-week mortality.
RESULTS: Crude mortality rate after a six week follow-up was 10.6% (n=135). Prognostic factors identified were: malnutrition risk (HR=2.1; 95% CI: 1.1-3.8; p=.02), delirium (HR=1.7; 95% CI: 1.2-2.5; p=.006), and dependency: moderate dependency (HR=4.9; 95% CI: 1.5-16.5; p=.01) or severe dependency (HR=10.3; 95% CI: 3.2-33.1; p < .001). The discriminant power of the model was good: the c-statistic representing the area under the curve was 0.71 (95% IC: 0.67 - 0.75; p < .001). The six-week mortality rate increased significantly (p < .001) across the three risk groups: 1.1% (n=269; 95% CI=0.5-1.7) in the lowest risk group, 11.1% (n=854; 95% CI=9.4-12.9) in the intermediate risk group, and 22.4% (n=125; 95% CI=20.1-24.7) in the highest risk group.
CONCLUSIONS: A simple score has been calculated (using only three variables from the CGA) and a practical schedule proposed to characterise patients according to the degree of mortality risk. Each of these three variables (malnutrition risk, delirium, and dependency) identified as independent prognostic factors can lead to a targeted therapeutic option to prevent early mortality.
Keywords
Aged, Aged, 80 and over, Area Under Curve, Cohort Studies, Comorbidity, Delirium/epidemiology, Emergency Service, Hospital/statistics & numerical data, Female, France/epidemiology, Geriatric Assessment, Hospital Mortality, Humans, Male, Malnutrition/epidemiology, Predictive Value of Tests, Prognosis, Proportional Hazards Models, Prospective Studies, ROC Curve, Risk Assessment, Risk Factors
Pubmed
Web of science
Create date
15/04/2015 8:36
Last modification date
20/08/2019 14:21
Usage data