Page 40 - Reduction of coercive measures

Chapter 2
Conclusion and discussion
Findings revealed a subset of coercive measures that were recorded with reasonable reliability, and that could provide the basis for routine registration of the use of coercive measures. This registration can be used to improve care and protect the rights of persons with intellectual disabilities, at the level of individual care plans, institutional policies, as well as national policies. However, registration of coercive measures yielded reliable data for only 25 out of 57 types of coercive measures. Despite standardized definitions for each coercive measure (Matson & Boisjoli, 2009; Williams, 2010), registration that covers the broad definition of coercive measures (“any measure that is restrictive”) is due to yield unreliable and variable prevalence outcomes.
The data revealed patterns of disagreement between registrations of support staff members, independent observers, and colleague support staff members. Both the observer and colleague staff member more often registered a coercive measure when the support staff member did not than the reverse. The stakeholder group, which reflected in the second part of the study on the findings concerning reliability, suggested that decisions to register particular care practices as coercive measures may be dependent on the encoding of practices performed and observed during the shift as restrictive, which would require awareness of the full set of 57 coercive measures. This awareness may have been heightened among the observers, because they were specifically trained and only had to focus on observing, rather than providing care and support. In addition to factual knowledge about practices that could be restrictive, differences in norm setting (e.g., the importance of self-determination) and being accustomed to restrictive measures may influence the encoding and interpretation of care practices, leading to differences in retrieval at the end of a shift when coercive measures were recorded (Frederiks et al., 2017). These potential explanations do not apply to the heightened prevalence according to the registrations by colleague support staff.
The need to have a broad definition of coercive measures and to have a registration that is as broad as possible was underscored by the relative independence of the use of the 57 different coercive
38