Measuring recovery after open lower limb fractures: combined objective functional tests and global perceived recovery outperform narrower metrics and a standard generic patient reported outcome measure

Table 5 Comparison of the correlations between different Measures

Test 1	Comparison	Test 2	N or N1/N2	p-value	Reference
TUAG	>	APFP	272/252	0.006**	As described by Glass (1996) [23]
ESST	>	APFP	261/252	0.022*
CGS	>	APFP	264/252	0.001**
FGS	>	APFP	264/252	0.024*
GPE	>	APFP	296/252	0.000***
DRI	>	APFP	233/252	0.001***
DRI	.	TUAG	233/272	0.574
DRI	.	ESST	233/261	0.322
DRI	.	CGS	233/264	0.989
DRI	.	FGS	233/264	0.308
DRI	.	GPE	233/296	0.197
TUAG	.	ESST	261	0.306	As described Sheshkin (2004) [24]
TUAG	.	CGS	263	0.302
TUAG	.	FGS	263	0.057
TUAG	.	GPE	265	0.193
ESST	.	CGS	257	0.071
ESST	.	FGS	257	0.469
ESST	<	GPE	255	0.033*
CGS	>	FGS	263	0.004**
CGS	.	GPE	257	0.198
FGS	<	GPE	257	0.011*

Isolated Measure: Ankle Plantar Flexion Passive (APFP)
Combined Measures: Timed up and go (TUAG), Edgren Side Step Test (ESST), comfortable and fast gait speed (CGS and FGS). PROMS: Global Perceived Effect (GPE) score and Disability Rating Index (DRI)
Using the methods described by Glass (1996) and Sheshkin (2004) [23, 24] > indicates Test 1 has a greater correlation with time than Test 2,whereas < indicates Test 2 correlates better with time. N or N1/N2 are the sample sizes (i.e. how many correlations were used). N used paired correlations and N1/N2 used two independent sets of correlations, hence two sample sizes

ISSN: 1471-2474