Page 41 - Reduction of coercive measures

measures. Only nine measures were found to be associated with other coercive measures. To some extent, this incoherent pattern can be explained by the low interrater reliability, which attenuated correlations. But the number is still small relative to the 25 coercive measures that were registered with adequate agreement between support staff members and independent observers. Even for these nine coercive measures, it is possible that these correlations are the result of similarly worded items. As it clearly is impossible to make categories of coercive measures on the basis of empirical clustering or underlying factors, one could use a priori defined categories on the basis of specific characteristics of coercive measures such as physical or mechanical measures, as proposed by Matson and Boisjoli (2009). Concerning registration of coercive measures this could lead to a clear order of measures and perhaps a way to recognize coercive measures more easily.
The second part of the study raises the concern that a mandatory and structural registration system which is part of the residents’ electronic personal file may yield an unreliable and incomplete picture, even after training of support staff and professionals as this was found important by the group of stakeholders. Insufficient registration risks persistent use of coercive measure against policies to reduce their use. Consensus on whether particular care measures were coercive measures or not by the team of professionals and support staff was conditional on the registration in the system. 84% coercive measures were identified and talked through during training and agreement was reached on 46% of coercive measures. In their reflections stakeholders emphasized the importance of awareness of coercive measures. However, systematic identification and training on awareness of coercive measures did not lead to consensus on coercive measures and a complete registration of all measures. Moreover, only four measures (7.0%) had a reasonable reliability in part one and were registered as coercive measure in part two of the study, underscoring that reliability may come to the expense of coverage.
Stakeholders also suggested that the meaning of the context in which a measure is applied is part of the determination of coercive measures by support staff. This could be in line with difficulties
Measurement of coercive measures
39
2