The terms ‘frailty’ and ‘vulnerability’ are widely used in the gerontology literature [
1]. Frailty was originally defined as a condition in older people meeting three or more of the following criteria: i) unintentional weight loss; ii) self-reported exhaustion; iii) slow walking speed; iv) weak grip strength and, v) low physical activity level [
2]. The term vulnerability refers to a wider range of older people who are at increased risk of functional decline or death [
3]. The Vulnerable Elder’s Survey-13 (VES-13) is a risk prediction tool designed in the United States (US) to predict functional decline and death in older community-dwelling (≥65 years) people over 2 years follow-up [
3]. It has good clinical utility, as it is easy to administer and can be used to identify older people at higher risk of poorer health outcomes who can be targeted for community-based interventions. The VES-13 was derived through a methodologically robust process, whereby variables with potential predictive power were identified from the US Medicare database and different models tested for relevant outcomes [
3]. The final VES-13 model includes items relating to patient age, self-rated health and the ability to perform specified physical and functional tasks [
3]. A score of ≥3 is considered high-risk of experiencing future functional decline or death. It has been successfully validated in several community-based US studies to predict functional decline and death [
3‐
7]. In one such US study (
n = 649, ≥75 years) for each additional increase in VES-13 point, the odds of functional decline or death increased by almost 40% (odds ratio (OR) 1.37 (95% confidence interval (CI) 1.25, 1.50) and the model’s c-statistic was 0.75 (95% CI 0.71, 0.80) over five-year follow-up [
4]. The VES-13 has also been extensively validated to predict various adverse health outcomes in older people with an index diagnosis of cancer [
8].
However, validation of the VES-13 outside North America in older people without cancer has been limited [
9,
10]. A prospective Dutch study (
n = 354, aged ≥70 years) with one-year follow-up reported that the VES-13 was significantly associated with functional decline in older persons without cancer (OR 2.83, 95% CI 1.35, 5.95) [
9]. One Irish study (
n = 2,033 aged ≥65 years) examined the cross-sectional association of the VES-13 with healthcare utilisation and reported that people categorised as vulnerable (32%) had higher healthcare use including primary care visits, emergency room (ER) visits and use of hospital services [
10]. Predicting emergency admission is of interest, both from a clinical and policy perspective internationally, and the use of risk prediction models to identify high-risk people is increasingly advocated [
11,
12]. Adopting the VES-13 to predict emergency admission could have both clinical and policy implications as with an ageing population, examining innovative ways of identifying older people at highest risk is important.
The aim of this study is to examine the VES-13 in predicting mortality and emergency admission in older people. The specific objectives were: 1) To externally validate the VES-13 in predicting mortality in a cohort of older community-dwelling people, and 2) To examine the predictive accuracy of the VES-13 in predicting all-cause emergency admissions and a subset of emergency admissions resulting from ambulatory care sensitive conditions.