Table 4.
The psychometric properties of weakness for the FRIED-Q compared to the FRIED-P
| Performance-based test dynamometer | Total | Positive predictive value | Negative predictive value | ||
|---|---|---|---|---|---|
| + | − | ||||
| Replacement questions | |||||
| + | |||||
| N | 24 | 23 | 47 | ||
| % | 46.2 | 16.3 | 24.4 | 51.1 | |
| Weakness | |||||
| − | |||||
| N | 28 | 118 | 146 | ||
| % | 53.8 | 83.7 | 75.6 | 80.8 | |
| Total | |||||
| N | 52 | 141 | 193 | ||
| Sensitivity | |||||
| % | 46.2 | ||||
| Specificity | |||||
| % | 83.7 | ||||
FRIED-P stands for the Fried Phenotype with slowness and weakness operationalized as performance-based tests. FRIED-Q stands for the Fried Phenotype with slowness and weakness operationalized as self-report questions. + indicates a participant’s weakness was lower than the cutoff determined by the Fried Phenotype or had a higher score than the cutoff on the replacement questions determined by Op het Veld and colleagues (see Online Resource 2), − indicates a participant's weakness was higher than the cutoff determined by the Fried Phenotype or had a lower score than the cutoff on the replacement questions determined by Op het Veld and colleagues