In search of a characterisation of frailty: applying exploratory factor analysis to ageing preoperative patients :...

Frailty is a useful prognostic concept which has spread to many clinical settings, including perioperative medicine. However, there is no consensus on its definition. This situation could impair its screening and the correction of underlying disturbances that have an influence on the clinical course of ageing patients.

In order to achieve a more precise characterisation of frailty, an exploratory factor analysis (EFA) was performed on the variables of eight frailty scales: Clinical Frailty Scale (CFS), FRAIL scale, Edmonton scale, Fried criteria, Robinson scale, Risk Analysis Index, the Frailty Index and the Modified Frailty Index. Later, a concordance study between the factors found in EFA and frailty according to the CFS (≥ 4 points) was conducted. One-hundred nine preoperative patients aged 65 years or older (60% men) were included, and data were collected from medical history and physical and laboratory tests. Our aim was to explain the behaviour of well-known frailty-related variables by identifying the factors that influenced them and to investigate whether these factors were related to frailty.

Three factors were found, each relating to a different set of variables: F1 representing comorbidities; F2 being an aggregation of disturbances in physical activity, cognitive status and anaemia; and F3 portraying alterations of the emotional sphere. The concordance study showed a strong association of F2 with frailty: adjusted OR 3.65 (95% CI 1.57 to 8.53). F3 presented a milder relationship: OR 2.54 (95% CI 1.28 to 5.02). No association of F1 with frailty was found: OR 1.15 (95% CI 0.58 to 2.26).

In our quest to characterise frailty, we found that this is best described by an aggregate of reductions in physical activity, impairment in cognitive status and anaemia, while comorbidities are not associated with it. This could support a modified version of the phenotypic model against other paradigms.