Parkinson’s Disease / 2012 / Article / Tab 2

Research Article

Accuracy of Fall Prediction in Parkinson Disease: Six-Month and 12-Month Prospective Analyses

Table 2

(a) Predictive values at 6 months.

Balance measureAUC (95% CI)ScoreSensitivitySpecificity Posttest probability
with test ≤ cutoff value
Posttest probability
with test > cutoff value

BESTest0.89 (0.74–0.95)≤69%0.930.845.81 (3.69–9.14)0.08 (0.04–0.17)0.690.03
Mini-BESTest0.87 (0.72–0.94)≤20/32(63%)0.860.783.97 (2.68–5.70)0.18 (0.11–0.78)0.600.07
BBS0.87 (0.75–0.95)≤47/560.790.865.64 (3.43–9.27)0.24 (0.17–0.36)0.680.09
FGA0.80 (0.62–0.90)≤15/300.640.813.37 (2.19–5.18)0.44 (0.34–0.59)0.560.15

(b) Predictive values at 12 months.

Balance measureAUC (95% CI)ScoreSensitivitySpecificity Posttest probability
with test ≤ cutoff value
Posttest probability
with test > cutoff value

BESTest0.68 (0.45–0.83)≤69%0.460.741.77 (1.19–2.62)0.73 (0.59–0.91)0.460.26
Mini-BESTest0.77 (0.55–0.89)≤20/32(63%)0.620.742.37 (1.66–3.34)0.52 (0.39–0.68)0.530.20
BBS0.68 (0.45–0.82)≤47/560.460.812.42 (1.53–3.82)0.67 (0.54–0.82)0.540.24
FGA0.70 (0.50–0.83)≤15/300.460.812.42 (1.53–3.82)0.67 (0.54–0.82)0.540.24

Article of the Year Award: Outstanding research contributions of 2020, as selected by our Chief Editors. Read the winning articles.