Page 133 - Reduction of coercive measures

Limitations
Several limitations should be taken into account when interpreting the results of the studies. First, limitations concerning the study described in Chapter two concerned the possibility that the observer did not observe all coercive measures, for example when certain coercive measures were applied out of sight and outside the hearing distance of the observer. Another limitation concerns the bias that could have arisen by non-random selection of shifts by the informant and the selection of stakeholders by the researchers. Finally, no specific methods were used to analyze qualitative data on stakeholders’ reflections. A limitation of the entire dissertation concerns the reliability of the information on coercive measures. The broad definition of coercive measures that refers to any measure that is restrictive for a resident in a specific situation is not sufficient to indicate coercive measures. The dissertation established a subset of coercive measures that can be measured with at least reasonable reliability. However, the broad interpretation may lead to differences in interpretation between the assessments of healthcare practices by different staff members, which in turn reduces the reliability of the data. A limitation of the study described in the third chapter was the absence of methodological analyses of the reflections of the experts. As a result, the study is limited to a summary of expert opinion that legislators and policy developers can use to optimize the registration of involuntary care. With regard to the fourth chapter, a number of shortcomings can be mentioned. The first limitation concerns the reliability of the measures of coercive measures. For the purpose of the study, information on the use of 76 coercive measures was analyzed. Reliability data were only available for a 56 measures. These data were derived from the first study described in the dissertation, and indicated that the overview of 56 measures was partly measurable with at least reasonable reliability. Together with the extension to 76 measures, part of the measures is to be known as unreliable and for another part, reliability was unknown. Second, the division of the total in coercive measures into different subsets is partly based on recent studies and partly, due to the broad definition, based on interpretation
General discussion
131
6