Jane McCusker1, Ritsuko Kakuma, Michal Abrahamowicz. 1. Department of Clinical Epidemiology and Community Studies, St. Mary's Hospital, McGill University, Montreal, Quebec, Canada. jane.mccusker@mcgill.ca
Abstract
BACKGROUND: This article will systematically review the methodological characteristics and results of studies of variables and indices that predict functional decline in older hospitalized patients. METHODS: We restricted this review to original longitudinal studies of predictors of either physical functional decline or nursing home admission among patients aged 60 and older. Two reviewers independently abstracted information on methodological characteristics and substantive results. RESULTS: Thirty articles were identified, derived from 27 different studies, reporting on 33 substudies. Substantial variability was found with respect to study design, outcomes measured, period of follow-up, predictors investigated, and analytic methods. Multivariable predictive indices were significantly associated with adverse outcomes in the majority of studies that investigated them, as were the following variables: age, diagnosis, activities of daily living, cognitive impairment (including delirium), and residence. CONCLUSIONS: The methodological heterogeneity of the studies identified limits quantitative synthesis of the results. Predictive indices for hospitalized elders appear to have moderate short-term predictive ability.
BACKGROUND: This article will systematically review the methodological characteristics and results of studies of variables and indices that predict functional decline in older hospitalized patients. METHODS: We restricted this review to original longitudinal studies of predictors of either physical functional decline or nursing home admission among patients aged 60 and older. Two reviewers independently abstracted information on methodological characteristics and substantive results. RESULTS: Thirty articles were identified, derived from 27 different studies, reporting on 33 substudies. Substantial variability was found with respect to study design, outcomes measured, period of follow-up, predictors investigated, and analytic methods. Multivariable predictive indices were significantly associated with adverse outcomes in the majority of studies that investigated them, as were the following variables: age, diagnosis, activities of daily living, cognitive impairment (including delirium), and residence. CONCLUSIONS: The methodological heterogeneity of the studies identified limits quantitative synthesis of the results. Predictive indices for hospitalized elders appear to have moderate short-term predictive ability.
Authors: Rebecca T Brown; Edgar Pierluissi; David Guzman; Eric R Kessell; L Elizabeth Goldman; Urmimala Sarkar; Michelle Schneidermann; Jeffrey M Critchfield; Margot B Kushel Journal: J Am Geriatr Soc Date: 2014-11-03 Impact factor: 5.562
Authors: J J Aarden; M van der Esch; R H H Engelbert; M van der Schaaf; S E de Rooij; B M Buurman Journal: J Nutr Health Aging Date: 2017 Impact factor: 4.075
Authors: Stefano Volpato; Julia Daragjati; Matteo Simonato; Andrea Fontana; Luigi Ferrucci; Alberto Pilotto Journal: Rejuvenation Res Date: 2016-02-23 Impact factor: 4.663
Authors: Alexander Bennett; Danijela Gnjidic; Mark Gillett; Peter Carroll; Slade Matthews; Kristina Johnell; Johan Fastbom; Sarah Hilmer Journal: Drugs Aging Date: 2014-03 Impact factor: 3.923
Authors: Brian K Gehlbach; Victor R Salamanca; Joseph E Levitt; Greg A Sachs; Mary Kate Sweeney; Anne S Pohlman; Jeff Charbeneau; Jerry A Krishnan; Jesse B Hall Journal: Am J Crit Care Date: 2011-09 Impact factor: 2.228
Authors: Bianca M Buurman; Juliette L Parlevliet; Bob A J van Deelen; Rob J de Haan; Sophia E de Rooij Journal: BMC Health Serv Res Date: 2010-10-29 Impact factor: 2.655