Table. 2.

Characteristics indicated by severity of change during DAS28 Flares

	All DAS28 Flares (n = 323)	Moderate DAS28 Flare (n = 171)	Self-report DAS28 Flare (n = 31)	Observed DAS28 Flare (n = 34)	p	Pairwise comparison between DAS28 Flare subgroups (p value)
						M vs SR	M vs O	SR vs O
DAS28	1.7 (0.8)	1.5 (0.7)	2.3 (1.0)	2.3 (0.7)	< 0.001	< 0.001	< 0.001	0.999
ESR	9 (15)	11 (16)	13 (17)	7 (12)	0.206	0.999	0.510	0.247
SJC	3 (4)	2 (3)	4 (4)	6 (6)	< 0.001	0.011	< 0.001	0.100
TJC	5 (6)	3 (3)	6 (5)	15 (5)	< 0.001	< 0.001	< 0.001	< 0.001
VAS	21 (25)	19 (21)	55 (23)	14 (23)	< 0.001	< 0.001	0.801	< 0.001
BP	5 (10)	3 (9)	16 (12)	2 (9)	< 0.001	< 0.001	0.999	< 0.001
VT	3 (9)	2 (8)	18 (7)	−1 (8)	< 0.001	< 0.001	0.478	< 0.001
MH	3 (9)	1 (8)	14 (10)	0 (9)	< 0.001	< 0.001	0.999	< 0.001

Data are mean (sd) changes at the time of DAS28 Flare categorisation from the preceding assessment. Change during Flare is rescaled to indicate worsening with positive values for all measures (including SF36 subscales). Significant heterogeneity between latent classes was recorded after univariate ANOVA. Pairwise comparisons with p values (shown after Bonferonni corrections) were performed. MvSR = Moderate vs Self-report. MvO = Moderate vs Observed. SRvO=Self-report vs Observed. Bold indicates the values for the most severely affected subgroups when heterogeneity was detected

ESR erythrocyte sedimentation rate, SJC swollen joint count, TJC tender joint count, VAS visual analogue scale for general health, BP SF36-Bodily Pain, VT SF36-Vitality, MH SF36-Mental Health. Changes in SF36 scores = 10 indicate one standard deviation predicted for the UK population