گزارش مراقب از اثربخشی مداخلات شناختی رفتاری برای خشم

Anger resulting in Aggression can be a significant problem for some people with Intellectual Disabilities. Carers were asked to complete a provocation inventory and an attribution scale before and after a group cognitive behavioral intervention aimed for anger and at similar points in time for a waiting list control. When compared using an analysis of variance results suggest that staff perceive a significant reduction in aggressive responses for participants who took part in the intervention. A regression analysis of factors that may influence the amount of change observed suggests that greater change was achieved if participants were accompanied by carers and had been attributed by carers as having an emotional cause for their behavior.

Aggressive behavior can have a significant impact on quality of life of an individual and can also have an impact on the psychological well-being of their carers (e.g. Hassall et al., 2005 and Jenkins et al., 1997). It can also impact on the quality of care provided by staff (Rose, Jones, & Fletcher, 1998). Aggression is relatively common in people with Intellectual Disabilities with a range of estimates of prevalence varying from 10 to over 50% (Benson & Brooks, 2008). Aggression in adults can have multiple functions (Benson & Brooks, 2008) and has been traditionally been treated within a psychopharmacological or within a behavioral framework. Reviews suggest that there is evidence to support the use of behavioral treatment methods (e.g. Didden et al., 1999, Emerson, 2001 and McClean et al., 2007) but less conclusive evidence for psychopharmacological methods (e.g. Brylewski and Duggan, 1999 and Edwards et al., 2007). While a behavioral approach may remain the most appropriate treatment option for some people with Intellectual Disabilities over recent years there has been a greater emphasis on using cognitive behavioral approaches (Lindsay & Hastings, 2004). Cognitively based anger control packages of treatment for people with learning disabilities have been the subject of a number of studies evaluating their efficacy from the participants perspective (e.g. Benson et al., 1986, Burns et al., 2003, Lindsay et al., 2004, Rose et al., 2005, Rose et al., 2008 and Taylor et al., 2002). Other studies have found reductions in expressed anger of participants by evaluating the groups from a carer perspective (King et al., 1999, Willner and Tomlinson, 2007 and Willner et al., 2002). More recently a study has shown that the skills and strategies learned by individuals in a day service setting can generalize to residential settings using staff ratings across different contexts (Willner & Tomlinson, 2007). While many of these studies suffer from significant weaknesses in design such as small numbers and the lack of an effective control group, all of them demonstrate reductions in either measures of aggression or expressed anger as a result of intervention. The conclusion of reviews (Hassiotis and Hall, 2008, Taylor, 2002 and Whitaker, 2001) are that while methodological difficulties exist the evidence suggests that these treatments show some promise. The mechanism for change remains unclear with some authors suggesting that only the behavioral elements of these interventions are effective in changing the behavior of people with Intellectual Disabilities (Sturmey, 2004). However, whatever the psychological mechanism is for change, therapeutic interventions based on cognitive behavioral models provide a methodology that values the individual and includes them at the centre of the change process. Some studies have used direct care staff within the therapy and there has been a suggestion that involving carers in anger interventions could enhances treatment-outcomes. Rose et al. (2005) also found that individuals with a greater receptive vocabulary tended to assess themselves as reducing their expressed anger more than others. This finding is supported by Willner et al. (2002) who found a positive correlation between general intelligence and improved outcome. Weiner's (1980) model of helping behavior suggests that if staff perceive challenging behavior to be under the personal control of an individual, then this is more likely to provoke negative emotions, such as anger, and results in them offering less help. In contrast, if the cause of challenging behavior is not judged to be under the individual's control it is more likely to provoke a greater willingness to help. Evidence exists for the application of Weiner's (1980) model to ID services (e.g. Dagnan et al., 1998 and Wanless and Jahoda, 2002) although some researchers have argued that it needs to be adapted (Bailey et al., 2006 and Jones and Hastings, 2003). Phillips and Rose (2010) have found that staff attributions of challenging behavior are related to placement breakdown. That is, when staff perceive clients to be in control of their challenging behavior their residential placement is more likely to break down even if reported levels of challenging behavior are similar. Other methods of assessing attributions also exist for example, Hastings (1997) developed the Challenging Behavior Attribution Scale that asks care staff why individuals with Intellectual Disability might engage in challenging behavior. The CHABA has six subscales and Grey, McClean, and Barnes-Holmes (2002) found that the learned positive and emotional subscales were most likely to be endorsed by staff as attributions for challenging behavior. That is, the staff would attribute the behavior to either positive reinforcement processes or emotional reasons. This suggests an association between staff attributions and the expression of aggression in people with Intellectual Disability that would be interesting to explore in relation to cognitive behavioral intervention programs. This paper aims to examine changes in carer ratings and attributions of individuals they care for over the course of a therapeutic group based on a cognitive behavioral model that aimed to help them manage their anger more appropriately and to reduce inappropriate aggression. It was hypothesized that greater change would result if clients were accompanied by staff and if they were more able. It was also suggested that carer attributions of client behavior may influence outcome however no specific hypotheses were made in respect of what attributions may be important.

The main finding from this study is that attendance at the group intervention reduces carer ratings on the provocation inventory. This supports previous studies that have used participant ratings. While the effect size obtained in this study is smaller than analyses of self-report data from participants (e.g. Rose et al., 2005), it is still of reasonable size for a therapeutic intervention, especially as there were relatively few dropouts (14%). This indicates that staff were generally rating participants as less likely to express inappropriate anger and aggression after group participation however, they do not change while waiting for treatment. While the effect size observed was lower than that obtained in studies of change in self-rated data this might be expected as you would expect greater change in emotions and cognitions rather than symptoms (Birchwood & Trower, 2006). While these measures essentially had the same content it can probably be assumed that the participants would be rating them on the basis of emotions, cognitions and symptoms whereas staff would have access to a more limited range of data with judgments being primarily based on overt symptoms. As a result this apparent difference in effect size may provide some support for cognitive changes however, this must be the subject of more rigorous evaluation in the future. When an analysis is conducted to see which variables may contribute to the observed changes, a significant amount of the variability can be explained by three variables (33.5%). Given the small number of variables used in the regression equation this is a relatively large proportion of the variance. In this instance receptive vocabulary was not significantly related to outcome suggesting that relative ability as measured by the BPVS was not a predictor of outcome. This result is in contrast to that found when group participants have completed assessments themselves (Rose et al., 2005) and it is interesting to speculate why this difference may occur. For example, more able individuals may be liable to self-report more socially acceptable responses on completion of the intervention. If participants were accompanied by staff to the group this was associated with a significant improvement in outcome. This finding is consistent with data obtained directly from participants and a number of potential explanations are possible. For example, carers may develop formulations and intervention strategies as a result of their attendance at the group (Rose, West, & Clifford, 2000). If carers rate clients high on the emotional subscale of the CHABA, this suggests a more significant improvement will result. That is, attendance at the group may help carers to understand the participants’ perspective more fully and thus support them to empathize with participants and change their responses more adaptively (Wanless & Jahoda, 2002). Staff participation in groups may also lead directly to changes in the client's interpersonal environment which may lead to improvements in their feelings and expression of anger (Jahoda, Trower, Pert, & Flyn, 2001). At the level of treatment, this raises the question of whether treatment strategies that purposefully target carers’ understanding of and sensitivity towards their clients would carry benefits over and above involving carers in treatment groups. There may also be a social desirability effect with carers who attend the group being more likely to report change than those who do not. However, even if this latter point is important the fact that they are prepared to report change may contribute to systemic changes within the individual's life that may result in actual change over time. There are a number of methodological difficulties with this study for example, the Anger Inventory was adapted from a self-rated provocation inventory that examined responses to a set of hypothetical situations, therefore measuring anger arousal rather than aggressive behavior, as such it only provides a limited measure of change. It was also adapted for completion by carers rather than participants, while an initial psychometric evaluation seems to suggest that this adaptation is appropriate (Rose & Gerson, 2009) a more thorough development would be desirable. It would also seem important to consider more objective changes in behavior in future research. The CHABA was also adapted by asking staff to rate an individual's behavior rather than vignettes and the internal consistency of the majority of the scales did not meet acceptable standards, it is possible that other aspects of the questionnaire did not meet conventional standards and results should be treated with caution. However, the results do partially support those of Grey et al. (2002) who found that the emotional subscale was likely to be associated with challenging behavior. It seems reasonable that using a cognitive behavioral approach would be associated with greater improvements for individuals whose difficulties staff rated as being more emotional as this approach focuses more on emotions than some other interventions. This intervention may therefore be more suitable for individuals where anger has an actual or perceived emotional cause. Further work is required on the development of both of these measures. It would also have been useful to collect data using more objective measures such as direct records of challenging behavior, however, this was not possible due to resource limitations. Information was not routinely collected on the presence of syndromes, psychiatric diagnosis, or the medication that participants in the study were receiving as resources were not available to collect and monitor this information. This information would have been useful to describe participants more clearly and monitor changes. This evaluation was conducted as part of ongoing clinical work and it was therefore impossible to truly randomize the allocation to groups even though some aspects of randomization were evident as allocation was performed on the basis of treatment availability. In order to improve the experimental design true randomization would be required however, this would pose ethical difficulties that could not be overcome in the setting for this research. Additionally it was not possible to randomize whether staff attended with participants or not and the results obtained here suggesting improvements when staff accompanied participants may simply reflect greater motivation to assist participants by staff who came with them. Despite these difficulties groups were very well matched and there did not seem to be any systematic bias operating in terms of allocation to group or waiting list conditions. Groups could not run until sufficient referrals had been received to make them viable, while individuals with more acute difficulties may have received other interventions while they were waiting (e.g. medication changes or individual therapy) however, this did not seem to influence overall levels of expressed anger within the waiting list. Finally, it would have been useful to collect follow up data while this was possible with some individual's staff turnover and availability meant that it was not possible to collect data from the same staff in a number of cases. This report contributes to a growing literature that suggests that cognitive behavioral interventions reduce anger when assessed from a range of perspectives, in this case by staff ratings. However, further improvements are required in the design of interventions unfortunately these developments will require a considerable investment in resources.