Table 5.

Multiple regression analysis

	an	Result	P-value	95% CI
bModel coefficients
HAQ-DI	947	0.349	<.001	0.247–0.451
Percent overall work impairment	323	11.819	.001	5.019–18.618
Percent activity impairment	912	14.728	<.001	11.341–18.115
cOdds ratios
Unemployment	966	1.737	.001	1.248–2.417
Patient-reported dissatisfaction	913	2.278	<.001	1.540–3.368
dRelative risk ratio
Both patient and physician agree	[Reference category in multinomial logistic regression]
Physician satisfied only	913	2.193	.002	1.325–3.629
Patient satisfied only	913	0.831	.443	0.517–1.334

All models controlled for age, region (the USA or Europe), patient sex, body mass index, current severity as stated by the physician, number of flares in the last 12 months and physician global assessment

CI confidence interval; HAQ-DI Health Assessment Questionnaire Disability Index; T2T treat-to-target

^aSample size for regression model

^bThe coefficient indicates the unit increase of parameter of interest for patients reporting anxiety/depression on the EQ-5D-3L domain compared with those not reporting any anxiety/depression

^cThe odds ratio indicates the odds of the parameter of interest occurring for patients reporting anxiety/depression on the EQ-5D-3L domain compared with those not reporting any anxiety/depression

^dThe relative risk ratio indicates the factor increase in the relative risk of only the physician being satisfied with current treatment (not the patient) for patients reporting anxiety or depression on the EQ-5D-3L domain compared with those not reporting any anxiety or depression