Table 4.
Multivariable linear regression analysis results show the effects of independent variables on perceived coercion. N = 192 patients
| Variable | Beta (95 % CI) | P-valuea | R2c |
|---|---|---|---|
| Female gender | −0.13 (−11.16–1.07) | 0.117 | |
| Age | −0.09 (−0.50–0.12) | 0.441 | |
| Living alone | −0.01 (−5.80–5.57) | 0.713 | |
| Global Score Index: Scl-90-R | 0.19 (0.82–9.35) | 0.015 | 3 % |
| Severity scores | |||
| Injected drug abuse | 0.01 (−5.99–6.09) | 0.720 | |
| Drug overdoses (lifetime) | 0.10 (−2.85–9.99) | 0.970 | |
| Suicide attempts (lifetime) | −0.10 (−6.84–5.40) | 0.340 | |
| Treatment variable | |||
| Treated for somatic diseasesb | 0.12 (−1.39–11.12) | 0.113 | |
| Involuntary hospital admission | −0.03 (−7.90–5.43) | 0.536 | |
a p-value obtained from bivariate linear regression. Only one independent variable showed a p-value <0.20 in bivariate analyses
bDuring the 6 months prior to admission
c R2 adjusted = squared correlation coefficient to obtain a measure of explained variance