Measuring Frailty in Systemic Lupus Erythematosus

Objective Recent research has explored frailty in systemic lupus erythematosus (SLE) using multiple measures. We examined the agreement among frailty measures and the association of each with cross‐sectional and longitudinal health outcomes. Methods We used data from the California Lupus Epidemiology Study (CLUES) to examine the following measures of frailty: Systemic Lupus International Collaborating Clinics (SLICC) Frailty Index (SLICC‐FI), Short Physical Performance Battery (SPPB), and Fatigue, Resistance, Ambulation, Illness, and Loss of Weight (FRAIL) scale questionnaire. PROMIS Physical Function 10a (PF) was tested as a proxy measure of frailty. Agreement between frailty classifications by each measure was assessed. Cross‐sectional associations of frailty classifications with hospitalization, valued life activities disability, cognitive impairment, 6‐minute walk test distance, self‐reported disease damage, fatigue, and depressive symptoms were assessed with logistic and linear regression analyses. Associations with hospitalization, disease damage increase, and disability increase over the subsequent 3 years were assessed Cox proportional hazards analyses. Results Percentages of participants identified as frail varied among the measures, from 10.8% to 45.9%. Agreement among classifications ranged from slight to substantial (κ from 0.17 to 0.63). Most of the frailty measures were associated with both cross‐sectional and longitudinal health outcomes, with the notable exception of the SPPB. SLICC‐FI had the most consistent association with outcomes, followed by FRAIL and PF. Conclusion Multiple measures of frailty appear to identify the risk of poor health outcomes. The intended use, as well as the simplicity and practicality of implementing the measure, may be the most important considerations in choosing a frailty measure.