Post-hoc validation of the Conley Scale in predicting the risk of falling with older in-hospital medical patients: findings from a multicentre longitudinal study

Background: The Conley Scale is one of the most widespread fall-risk screening tools in medical unit settings, despite the lack of data regarding its validity in patients currently admitted to these units. Aims: Establishing the validity of the Conley Scale in identifying patients at risk of falling in an acute medical setting. Methods: A 6-months longitudinal study in 12 acute medical units from September 2012 to March 2013, a total of 1464 patients with ≥65 years of age were consecutively enrolled and evaluated with the Conley Scale within 24 h of admission. A construct validity, internal consistency, and a priori and a posteriori predictive validity study was performed. Results: The explorative factor analysis showed a two-factor structure explaining a total variance of 48.3 %: previous history (30.41 %), and physical and cognitive impairment (17.9 %). The scale reported a poor internal consistency (Cronbach’s α = 0.465) and the capability to correctly identify 18/649 patients as being at risk of falling, whereas the negative predictive value was 98.5 %. The sensitivity and specificity values were 60.0 and 55.9 %, respectively. No difference emerged between patients scored as at risk and those scored as not at risk in the time elapsed from admission to the first fall (HR = 0.600, 95 % CI 0.289–2.247 p = .166). Discussion: The Conley Scale is not able to predict falls in elderly acute medical patients, and has reported poor internal consistency and accuracy. Conclusions: More studies are needed to develop appropriate tools to predict the risk of falling in elderly individuals admitted to an acute medical setting.