Morbid RANKIN score strongly predicts in-patient mortality within 90 days in older people in both acute and community care settings independently of FANGG (fracture, acquired neurological deficit or any geriatric giants) and acute illness markers (AIMs)

Understanding the relative impact of factors that are associated with poor outcome of older people admitted either to an acute or a rehabilitation setting is essential to further our knowledge in provision of appropriate care. We conducted a prospective study to examine whether FANGG and AIMs are important prognostic indicators of mortality outcome in hospitalised older people when patients' morbid functional status is considered. Participants were two consecutive series of 200 patients admitted to care of the elderly wards in an acute teaching hospital and a community hospital in Birmingham, UK during April to August 2004 and the same months in 2005. The association with the outcome of mortality was examined in a univariate analysis, and then multiple logistic regression models. A total of 400 patients (men 116, 29.2%) were included in this study (mean age=85.3 years, range 64-104 years). There were 72 in-hospital deaths (18.0%). The prevalence of FANGG is low in this series; the majority (89.3%) had none or only one factor. The adjusted analysis showed that only age (p=0.05), gender (p=0.01) and morbid Rankin score (p<0.001) were predictive of outcome. The morbid Rankin score of 5 had over threefold higher likelihood of dying as in-patient (OR=3.31; 95% CI=1.48-7.40, p=0.003) independently of age, gender, site, FANGG, Rankin×AIMs interaction and AIMs. The morbid Rankin score strongly predicts the in-patient mortality over and above of age, gender, FANGG and AIMs in this patient population.