Table 3.

Internal and external responsiveness of RMDQ and PSFS at short-term follow-up

	n	Effect size (a) (84% CI)	Effect size (b) (84% CI)	Correlations of the change scores with GPE at discharge	P*	AUC (GPE cut-off for improvement = 3 or better)	P**	AUC (GPE cut-off for improvement = 4 or better)	P**	AUC (GPE cut-off for improvement = 5 or better)	P**
LBP
RMDQ	831	1.01 (0.96–1.06)	0.92 (0.87–0.97)	0.57	0.03	0.85	0.48	0.81	0.001	0.83	0.01
PSFS	831	1.70 (1.63–1.78)	1.11 (1.07–1.16)	0.61		0.85		0.85		0.87
Low activity limitation
RMDQ	453	1.08 (0.99–1.17)	0.81 (0.74–0.87)	0.59	1.0	0.81	0.34	0.80	0.06	0.81	0.01
PSFS	453	1.55 (1.46–1.65)	1.11 (1.04–1.17)	0.60		0.82		0.84		0.87
High activity limitation
RMDQ	378	2.84 (2.67–3.01)	1.20 (1.12–1.27)	0.70	0.01	0.92	0.01	0.88	0.37	0.90	0.09
PSFS	378	2.16 (2.02–2.30)	1.12 (1.06–1.21)	0.65		0.88		0.88		0.87

Short-term follow-up refers to time-point at discharge from treatment: acute = 4 weeks, sub-acute = 6 weeks, chronic = 8 weeks

Effect size (a): Standardized effect size defined as the difference between the mean baseline scores and follow-up scores divided by the standard deviation of baseline scores [7]

Effect size (b): Standardized response mean defined as the difference between the mean baseline scores and follow-up scores divided by the standard deviation of the change score [7]

The effect sizes are significantly different if the 84% CI between the RMDQ and PSFS do not overlap [11]. Since the CI do not overlap for each cohort listed in this table, the effect sizes are significantly different for the PSFS and RMDQ

* P values of two-tailed test for paired Pearson’s r correlations Cohen and Cohen [22, 29]

** P values are from DeLong’s test of paired AUC values [24]