Table 4.
Logistic regression analyses to investigate whether changes in secondary outcome measures are associated with RTWa
| Univariable associations | Multivariable associations | |||||||
|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | p | OR | 95% CI | p | |||
| Changes on secondary outcome measures | Lower | Higher | Lower | Higher | ||||
| Improved CORE-OM total | 1.06 | 0.99 | 1.14 | 0.099c | ||||
| Improved work ability | 1.16 | 1.03 | 1.32 | 0.019c | ||||
| Improved generalized self-efficacy | 0.99 | 0.47 | 2.09 | 0.981 | ||||
| Improved expectancy of future work ability (yes)b | 5.90 | 1.81 | 19.24 | 0.003c | 5.90 | 1.81 | 19.24 | 0.003 |
The dependent variable was RTW; an odds ratio (OR) for RTW of >1 indicates a higher probability of successful RTW
aN varies between 121 and 164 due to missing cases
bOn this variable, being a case indicates either a change from negative to positive expectancy of future workability, or a maintained positive expectancy of future workability
cThese variables were included in a backwards stepwise multivariable logistic regression with p < 0.05 as cut-off for including independent variables in the final model. This resulted in a final model in which only improved expectancy of future work ability was statistically significant associated with RTW