Background/Purpose: Poor SLE outcomes have been associated with certain clinical and socio-demographic characteristics, but the determinants of patient reported outcomes (PROs) in SLE are unknown. PROMIS, recently validated in SLE, offers dynamic computer adaptive tests (CATs) that precisely and efficiently measure PROs in physical, mental, and social health domains. The aims of this study were to identify clinical and demographic characteristics independently associated with worse health status in SLE patients as measured by PROMIS CATs.

Methods: Adults meeting ACR SLE classification criteria were recruited from an SLE Center of Excellence. Subjects completed 4 PROMIS CATs: physical function, fatigue, pain interference, ability to participate in social roles. SLE disease activity, flare, and damage were evaluated with the SELENA-SLEDAI and SLICC-ACR damage index. Multivariable generalized linear models were used to assess the cross-sectional association between PROMIS CAT scores and clinical and demographic traits.  All variables with p < 0.2 in univariate analyses were included in the multivariable analyses.  2-Way interaction terms were considered and significant interaction terms (p < 0.05) were included in the final models if doing so provided models of better quality (indicated by lower Akaike and Bayesian Information Criterion values).

Results: A diverse group of 204 SLE patients completed PROMIS CATs (table 1).  Black race, Hispanic ethnicity, disability, Medicaid, active arthritis, SELENA-SLEDAI flare, history of cognitive impairment/psychosis, deforming/erosive arthritis, and avascular necrosis were associated with statistically and clinically significantly worse PROMIS scores in univariate analyses. In the final multivariable model only disability and active arthritis were independently associated with worse physical function, fatigue, and social participation, while arthritis and cognitive impairment/psychosis were independently associated with worse pain interference.

Conclusion: Surprisingly, neither flare status nor socio-demographic characteristics were independently associated with these four important PROs, whereas active arthritis, disability, and a history of cognitive impairment/psychosis were. These data suggest that: 1) pain and physical and mental dysfunction drive patient well-being regardless of race, ethnicity, and insurance type; and 2) the physician-derived SELENA-SLEDAI flare measure poorly captures the impact of disease activity on SLE patients.