Early and delayed assessments of quantitative gait measures to improve the tap test as a predictor of shunt effectiveness in idiopathic normal pressure hydrocephalus

Table 2 Receiver operating curve analysis

Examination day	Measures (change in time or step number from baseline)	AUC	p	Sensitivity (%)	Specificity (%)	Cutoff
Day 1 (N = 63)	TUG (s)	0.770	0.249	76.6	75.0	2.0
	10Ti (s)	0.645	0.102	63.8	68.8	0.8
	10St (step)	0.602	0.027*	36.1	93.8	4.5
Day 4 (N = 52)	TUG (s)	0.677	0.283	43.9	90.9	5.0
	10Ti (s)	0.711	0.383	63.4	81.2	2.3
	10St (step)	0.579	0.053	34.1	90.0	5.5

Examination day	Measures (change in percentage from baseline)	AUC	p	Sensitivity (%)	Specificity (%)	Cutoff (%)
Day 1 (N = 63)	TUG (%)	0.808	–	78.3	80.0	11.3
	10Ti (%)	0.641	0.193	63.0	66.6	9.4
	10St (%)	0.577	0.016*	41.3	86.7	13.6
Day 4 (N = 52)	TUG (%)	0.599	0.109	73.2	50.0	10.0
	10Ti (%)	0.738	0.483	63.4	90.0	15.8
	10St (%)	0.560	0.036*	41.5	80.0	17.0

ROC curve analysis showed the AUCs, sensitivities, specificities and cutoff values. The percent change from baseline of AUC on day 1 was regarded as a control and compared with other variables. Note the TUG on day 1, both for change in time and change in percentage, showed the highest AUC
Statistical differences were noted in most of 10Sts (*), but not in other variables of the TUG and the 10Ti

ISSN: 2045-8118