External validation of the Vulnerable Elder's Survey for predicting mortality and emergency admission in older community-dwelling people: a prospective cohort study.

BACKGROUND: Prospective external validation of the Vulnerable Elder's Survey (VES-13) in primary care remains limited. The aim of this study is to externally validate the VES-13 in predicting mortality and emergency admission in older community-dwelling adults.

METHODS: Design: Prospective cohort study with 2 years follow-up (2010-2012).

SETTING: 15 General Practices (GPs) in the Republic of Ireland.

PARTICIPANTS: n = 862, aged ≥70 years, community-dwellers Exposure: VES-13 calculated at baseline, where a score of ≥3 denoted high risk.

OUTCOMES: i) Mortality; ii) ≥1 Emergency admission and ≥1 ambulatory care sensitive (ACS) admission over 2 years.

STATISTICAL ANALYSIS: Descriptive statistics, model discrimination (c-statistic) and sensitivity/specificity.

RESULTS: Of 862 study participants, a total of 246 (38%) were classified as vulnerable at baseline. Fifty-three (6%) died during follow-up and 246 (29%) had an emergency admission. At the VES-13 cut-point of ≥3 denoting high-risk model discrimination was poor for mortality (c-statistic: 0.61 (95% CI 0.54, 0.67), ≥1 emergency admission (c-statistic: 0.59 (95% CI 0.56, 0.63) and ≥1 ACS emergency admission (c-statistic: 0.63 (95% CI 0.60, 0.67).

CONCLUSIONS: In this study the VES-13 demonstrated relatively limited predictive accuracy in predicting mortality and emergency admission. External validation studies examining the tool in different health settings and healthier populations are needed and represent an interesting area for future research.