Predicting Falls in People with Multiple Sclerosis: Fall History Is as Accurate as More Complex Measures

Table 3

Average sensitivity of baseline variables for predicting falls and injurious falls in the following 6 months.

	Occurrence/score	AUC: any falls in the following 6 months	AUC: any injurious falls in the following 6 months
A fall in the past year n (%)	39 (75%)	0.75; P < 0.001	0.69; P < 0.001
A fall in the past 2 months n (%)	26 (50%)	0.71; 𝑃=0.002	0.64; 𝑃=0.012
ABC total score	78.8 ± 20.9	0.69; 𝑃=0.018	0.66; 𝑃=0.025
FES-I total score	25.8 ± 9.6	0.66; 𝑃=0.029	0.59; 𝑃=0.14 (NS)
MSWS-12 total score	26.1 ± 14.8	0.69; 𝑃=0.013	0.65; 𝑃=0.031
APR latency from posturography; (in milliseconds)	135.7 ± 19.7	0.62; 𝑃=0.082	0.52; 𝑃=0.40 (NS)
EDSS score	2.8 ± 1.5	0.60; 𝑃=0.14 (NS)	0.55; 𝑃=0.30 (NS)
T25FW time in seconds	5.7 ± 3.8	0.71; 𝑃=0.009	0.64; 𝑃=0.043

AUC: area under the curve; ABC: Activities-specific Balance Confidence; FES-I: Falls Efficacy Scale International; MSWS-12: Multiple Sclerosis Walking Scale-12; APR: Automatic Postural Response; EDSS: Expanded Disability Status Scale; T25FW: Timed 25-Foot Walk.