کاهش رفتار تهاجمی پس از درمان خشم بیماران با ناتوانی ذهنی بستری شده در بیمارستان

Anger is related to violence prior to hospitalization, during hospitalization, and after discharge. Meta-analyses have established treatment efficacy in reducing anger, but few studies have addressed whether reduced anger leads to lowered aggressive behavior. This study concerns individually-delivered anger treatment, specialized for offenders with intellectual disabilities, delivered twice weekly for 18 sessions to 50 forensic hospital patients. Assessments involved patient self-report of anger, staff ratings of anger and aggression, and case records of assaultive incidents. Physical assault data were obtained from records 12 months pre-treatment and 12 months post-treatment. Significant reductions in assaults following treatment were found by GEE analyses, controlling for age, gender, length of stay, IQ, and pre-hospital violence. Following treatment, physical attacks reduced by more than half, dropping from approximately 3.5 attacks per patient 6 months prior to treatment, versus approximately 1 attack per patient in the 6–12 month interval post-treatment. In hierarchical regressions, controlling for IQ, reduction in physical assaults was associated with pre-to post-treatment change in anger level. These findings buttress the efficacy of anger treatment for patients having histories of violence and have significance for patient mental health, hospital staff well-being, therapeutic milieu, hospital management, and service delivery costs.

For detained hospital patients, both forensic and civil commitment, assaultive behavior is an important problem that affects patient care and staff well-being. Assaultive behavior by patients impairs the treatment milieu, results in increased restrictions and longer periods of detention, constitutes a risk for harm among staff, and has considerable costs for the institution in workers' compensation claims, sick leave and employee turnover. Anger has been found to be predictive of assaultive behavior by psychiatric patients, before, during, and after hospitalization. While there is considerable evidence that cognitive-behavioral anger treatment results in diminished levels of anger, there is little evidence for it reducing patient violence. The present study, which concerns forensic patients with intellectual disabilities, investigates whether assaultiveness in hospital is reduced following a specialized anger treatment (Taylor & Novaco, 2005) and whether that reduction is associated with therapeutic gains in assessed anger levels. The issue of hospital patient assaultiveness was brought to the fore by Fottrell (1980) in a study at three British psychiatric hospitals, finding that 10% of the patients had been violent. That rate was soon eclipsed in research by Larkin, Murtagh, and Jones (1988) at one of the British Special Hospitals, where 36.6% of the patients were assaultive in a six-month interval. These and other seminal studies (e.g., Convit et al., 1988 and Palmstierna and Wistedt, 1989) drew research attention to the problem, the international scope of which is clear. For example, Cheung, Schietzer, Tuckwell, and Crowley (1996) found that the rate of physical assaults in an Australian hospital was 97.6 per 100 patients per year. For over 4000 California State Hospital patients, Novaco (1997) reported 14% had physically assaulted someone in hospital in a 30-day period. In Sweden, a survey of 731 nurses and 320 psychiatrists in all the health districts of Stockholm found that 57% had been victimized by violence in the previous 12 months (Soares, Lawako, & Nolan, 2000). Among the 1494 nurses in 27 hospitals in Japan studied by Ito, Eisen, Sederer, Yamada, and Tachimori (2001), 41.3% had experienced assault within the past year, and the risk of assault by patients was significantly related to their intentions to leave their job. With respect to the hospital in England involved in the present study, 47% of male patients had been physically assaultive and 34% had carried out two or more physical assaults post-admission (Novaco & Taylor, 2004). The importance of anger for patient assaultiveness was established early in this field by Craig (1982) and by Kay, Wolkenfeld, and Murrill (1988), whose studies respectively found anger to be the strongest variable associated with physical aggression before hospital admission and during hospitalization. It is now been established in multiple studies with control variables that anger is predictive of physical aggression prior to hospital admission (Craig, 1982; McNeil, Eisner, & Binder, 2003; Novaco, 1994), during institutionalization (Doyle & Dolan, 2006a; Linaker & Busch-Iversen, 1995; Novaco, 1994 and Novaco and Taylor, 2004; Wang & Diamond, 1999) and in the community after discharge (Doyle et al., 2012 and Doyle and Dolan, 2006b; Monahan et al., 2001; Sadeh & McNeil, 2013; Skeem et al., 2006; Swogger et al., 2012 and Ullrich et al., 2014). Among all adult patients in five California State Hospitals, 35% were rated by their primary clinician as someone who “gets angry and annoyed easily” (Novaco, 1997). To the extent that anger is an antecedent variable in assaults by patients, it serves as a focus for intervention. Nine meta-analyses on the effectiveness of psychotherapy for anger have been published (Beck and Fernandez, 1998 and Del Vecchio and O'Leary, 2004; DiGuiseppe & Tafrate, 2003; Edmonson and Conger, 1996, Gansle, 2005, Ho et al., 2011, Saini, 2009, Sukhodolsky et al., 2004 and Tafrate, 1995), which overall have found medium to strong effect sizes, indicating that approximately 75% of those receiving anger treatment improved compared to controls. Cognitive behavioral therapy (CBT) approaches have greatest efficacy (DiGuiseppe & Tafrate, 2007). Anger interventions with offender populations, however, have in some studies been less successful. For example, Howells and Day and their colleagues (e.g., Heseltine et al., 2010, Howells et al., 2005 and Watt and Howells, 1999) have generally found low efficacy for their anger management programs, which were group-based interventions, largely delivered in prisons. They have highlighted anger problem complexity and insufficient intensity of treatment as potential sources of ineffectiveness and have found that anger declined when “treatment readiness” was present (for a review, Novaco, 2013). Many anger intervention studies with non-forensic populations have not concerned patients with serious clinical problems, so violent behavior has not been within their purview. However, even in anger treatment research with forensic populations, including our own controlled studies discussed below, violence or assaultiveness has been missed in the outcome criteria. Because aggressive behavior is an important part of the anger construct (Novaco, 2000), this is a gap that needs to be addressed. Perhaps the only study to have assessed the effect of anger treatment on violent behavior against a treatment control condition is that of Lindsay et al. (2004) with intellectual disabilities clients in the community. In that study with 47 clients, at post-treatment follow-up assessment, 14% of those who had received anger treatment had been physically assaultive, compared to 45% in the treatment control condition. Anger interventions for clients with intellectual disabilities Aggressive behavior is a prominent problem among people with intellectual disabilities, particularly those in institutional settings (e.g., Harris, 1993; Hill & Bruininks, 1984; Sigafoos, Elkins, Kerr, & Attwood, 1994; Smith, Branford, Collacott, Cooper, & McGrother, 1996). The prevalence of physical aggression in these studies is 35% or higher for persons in institutional settings. The all too common tendency, however, has been to attribute their emotional difficulties and challenging behavior to their disability, rather than to their emotional state or needs. In the past decade, various implementations of cognitive-behavioral anger treatment have occurred with clients with intellectual disabilities, reviews of which are provided in Nicoll et al., 2013 and Taylor and Novaco, 2013 and Willner et al., 2013 and Willner et al., 2013. Nevertheless, except for Lindsay et al. (2004) and Willner, Rose, et al. (2013), reduction in aggressive behavior has not been the anger treatment target in controlled studies. Our previous therapeutic intervention studies (Taylor et al., 2005, Taylor et al., 2002 and Taylor et al., 2004) have shown anger treatment gains compared to control conditions, but aggressive or violent behavior measures were not included. This has also been the case for the group-based anger treatment studies of Rose and his colleagues (e.g. Rose, Dodd, & Rose, 2008). Outside of studies with control treatment conditions, there have been several case series utilizing cognitive behavioral, mindfulness and behavioral skills training approaches reporting clinically significant reductions in aggressive behavior (Allen et al., 2001, Lindsay et al., 2003, Rose, 1996, Singh et al., 2008 and Travis and Sturmey, 2013). Finding no differences in anger between offenders and non-offenders, yet higher aggressive behavior for offenders, Nicoll and Beail (2013) questioned the validity of “the rationale that reduction in anger levels in offenders with intellectual disabilities would reduce aggression/offending behavior that would effectively place them in a non-offending population” (p. 469). Setting aside the complexities of the latter part of that statement, it is the case that evidence for aggression/violence reduction associated with anger treatment remains sparse. Lindsay et al. (2004) found that anger treatment, compared to a wait list control, was associated with significantly lower aggressive incidents in the community. Willner, Jahoda, et al. (2013) and Willner, Rose, et al. (2013), in a cluster randomized control trial of a group-based anger intervention for ID clients in day services, found that those in the treatment condition had significantly lower challenging behavior, but there were no differences on their aggressive behavior measure. The link between patient anger and hospital assaultiveness was demonstrated by Novaco and Taylor (2004), as patient-rated anger significantly accounted for patient assaults, controlling for age, IQ, length of stay, prior violent offending, and personality variables. Those findings conjoined with anger treatment research results signals the need for an investigation of whether reductions in patient assaultive behavior follow from a therapeutic intervention focused on anger. Study design The present study is a clinical service evaluation of a specialized anger treatment protocol (Taylor & Novaco, 2005) conducted with intellectual disabilities patients in a forensic hospital. The study focus is on whether there are any reductions in patients' assaultive behavior associated with individual-based CBT anger treatment, comparing assault incidents for 12-month pre-treatment and 12-month post-treatment time frames. Multiple covariates are used in our analyses. Further, we seek to determine whether the hypothesized subsequent reduction in assaultive behavior is associated with reductions in anger during the treatment phase.