In What Works for Whom, Roth and Fonagy (1996) provided a critical review of the evidence base for psychotherapy. The key findings are that there is good evidence for the efficacy of psychotherapies, but there was more evidence for cognitive–behavioral than for psychodynamic psychotherapy. References to research with people who have mental retardation, however, were notably absent from the review. The Handbook of Psychotherapy and Behavior Change (Bergin & Garfield, 1994), published 2 years earlier, did not contain any reference to people with mental retardation at all. There is a statement in Roth and Fonagy in relation to people with mental retardation: “Although there are reports of effective psychodynamic treatment (e.g., Sinason, 1992), systematic outcome research has focused on behavioral training techniques” (Target & Fonagy, 1996, p. 312). There was no comment on cognitive–behavioral psychotherapy or any other approaches, which probably reflects the state of the research literature at the time. Psychological treatments provided and researched for people with mental retardation then were either behavior modification or behavioral skills training. These approaches lie outside the scope of reviewers of psychotherapy research, whose main interests are cognitive and psychodynamically based treatments for mental health concerns. Reviewers of psychotherapy with people who have mental retardation published in the mid-1990s only found case studies on the use of psychodynamic psychotherapy, and only two reports of cognitive–behavioral psychotherapy for mental health issues (Hurley, Pfadt, Tomasulo, & Gardner, 1996; Nezu & Nezu, 1994). The reference to Sinason (1992) in Roth and Fonagy (1996) is also surprising because she did not report any data on effectiveness.
In a more recent review of the effectiveness of psychotherapy with people who have mental retardation, Prout and Nowak-Drabik (2003), covering the period 1968 to 1998, found 92 reports. Their definition of psychotherapy was fairly broad and included a range of behavioral interventions, including relaxation, desensitization, and various skills training interventions. They concluded that the area is dominated by case studies, with few controlled comparisons or clinical trials. Cognitive–behavioral and psychodynamic interventions only accounted for 28% of the reports reviewed. The largest proportion of the theoretical orientation of the 92 studies could not be determined (37%), and a further third were behavioral interventions. Only 9 studies contained sufficient data for inclusion in a meta-analysis, yielding a mean effect size of 1.01. All but one of these were behavioral interventions. However, in view of the wide acceptance of the effectiveness of behavioral interventions, it is surprising that only 8 were found. The 92 studies were reviewed by an expert panel, which concluded that psychotherapy effects are modest with persons who have mental retardation. The reviewers could only indicate that the effects could be demonstrated across theoretical approaches.
This commentary focuses on the effectiveness of the more controversial use of psychodynamic and cognitive–behavioral psychotherapy with people who have mental retardation. The progress, or lack of it, since the publication of What Works for Whom? is considered.
Cognitive–Behavioral Interventions With People Who Have Mental Retardation
Dagnan and Chadwick (1997) identified two broad-based approaches to cognitive–behavioral therapy that have been used with people who have mental retardation: self-management approaches and cognitive therapy.
In this approach it is assumed that emotional and behavioral difficulties are due to a lack of cognitive skills or presence of deficits. Interventions such as self-monitoring, self-instruction, self-control, and problem-solving/decision-making are used. These methods are employed in conjunction with relaxation techniques, education, skill acquisition, and social skills training.
There are now numerous case reports and a series of small numbers of uncontrolled case studies concerning the effectiveness of these methods, especially with criminal offenders who have mental retardation (Lindsay, Marshall, Neilson, Quinn, & Smith, 1998; Lindsay, Neilson, Morrison, & Smith, 1998). Not all reports, however, have been positive (e.g., Rose, Jenkins, O'Connor, Jones, & Felce, 2002). A few attempts at controlled studies have also been made and are reviewed here. These researchers have evaluated problem-solving and anger management packages.
Nezu, Nezu, and Arien (1991) conducted a study comparing problem-solving with assertiveness-training group treatment and no treatment. Their participants were 28 adults with mental retardation and co-morbid mental health problems. They found that both treatments produced better outcomes than did no treatment, and benefits were maintained at 3-month follow-ups. However, no differences between the treatment conditions were found. Loumidis and Hill (1997) attempted to evaluate the impact of problem-solving group treatment on problematic behavior. They compared the outcome for 29 recipients of group therapy with 17 people in a control group. They found little effect for problem-solving on its own or in conjunction with skills training. It is also difficult to draw conclusions from this study because the intervention package included education about emotions, relaxation, and self-instruction.
Cognitive–behavioral treatment for anger is based on the work of Novaco (1975), who utilized a stress inoculation paradigm. The main elements of this intervention are relaxation, behavioral skill training, education, self-instruction, and problem-solving. Whitaker (2001) reviewed 16 published reports on interventions based on this approach. All but 2 of them concerned a single case (n = 6) or a small series of uncontrolled single cases (n = 8). Benson, Rice, and Miranti (1986) compared relaxation with self-instruction, problem-solving, and a combined condition in group treatment with 54 recipients. Although they found statistically significant improvements on their outcome measures for all conditions, they did not include a no-treatment control group. The only study in the review with a no-treatment control condition was conducted by Rose, West, and Clifford (2000). They treated 25 participants who had histories of assault, damage to property, or aggression in groups and compared them with 19 people in a control group. A novel aspect of this intervention was that a direct caregiver accompanied and remained with the recipient throughout group sessions. The purpose of this was to encourage collaborative working and facilitate transfer of skills to every day settings. Rose et al. reported that participants in the therapy condition had better outcomes than those in the waiting list control group on measures of anger, self-concept, and depression. However, it would appear that the waiting-list control group was not entirely independent, and the presence of direct caregivers was a novel component of treatment and could contribute to outcome.
Since the publication of Whitaker's (2001) review, two further controlled evaluations of anger management have been published. Taylor, Novaco, Gillmer, and Thorne (2002) provided a preliminary report on an evaluation of a modified cognitive–behavioral treatment package for anger over 12 sessions. The recipients were men with mild to borderline mental retardation who had histories of criminal offending and were living in secure accommodation. Participants were randomly allocated to either treatment (n = 9) or routine-care control (n = 10) groups. Self-report of anger intensity on a shortened Novaco Provocation Inventory (Novaco, 1975) was reported to have significantly reduced in the treatment group but not in the control group. Staff ratings of clients' anger disposition and coping behavior after treatment provided some modest support for treatment effectiveness. Willner, Jones, Tams, and Green (2002) carried out a small, controlled trial of a cognitive–behavioral anger management group. They alternately allocated 14 clients to a treatment and no-treatment waiting-list control group. They reported that the groups were similar in age, IQ, and pretreatment anger scores. Outcome was evaluated through participant and caregiver ratings on an Anger Inventory and Novaco's Provocation Index. The treatment group scores significantly reduced, whereas those in the no treatment group did not. The authors also conducted various other analyses but because of the small number of participants, their reliability and generalizability is questionable.
Cognitive therapy is concerned with cognitive distortion and has developed from the psychotherapeutic tradition. In this approach, unhelpful or irrational emotions or behaviors are considered to be the products of distorted cognitions in the form of beliefs, attributions, inferences, and evaluations. The aim of treatment is to help people examine and test the utility of the meanings they make of their experience. To date, the evidence base for this approach is very limited and is comprised of case reports and a report of a series of cases (Lindsay, 1999).
Psychodynamic Psychotherapy With People Who Have Mental Retardation
Since Hurley et al.'s (1996) and Nezu and Nezu's (1994) reviews were published, some pre–post open trials have been reported. Frankish (1989) found reductions in behavior problems for series of 6 recipients. Beail (1998) reported the outcome of weekly outpatient psychoanalytic psychotherapy with 20 men with mental retardation who presented with behavior problems or who had committed a range of criminal offenses. The problem and offending behaviors were eliminated in most cases and maintained at 6-month follow-ups. Beail (2001) reported on recidivism rates among 18 male criminal offenders with mental retardation. Of the 13 who completed psychodynamic psychotherapy, the majority remained offense free at 4-year follow-ups. Of the 5 men who refused treatment, all had re-offended within 2 years. Beail and Warden (1996) and Beail (2000) reported a study of the outcome of psychodynamic psychotherapy with 20 adults who had mental retardation and co-morbid mental health problems. They found significant reductions in symptoms of psychological distress, improvements in interpersonal functioning, and increases in self-esteem following psychodynamic psychotherapy.
Whitaker (2001) concluded that the evidence for the effectiveness of cognitive approaches to anger management is weak. This could also be said of the evidence for the effectiveness of cognitive–behavioral and psychodynamic approaches with people who have mental retardation overall. Alternatively, the available evidence provides a positive indicator for treatment effectiveness. People with mental retardation do make gains during cognitive–behavioral and psychodynamic psychotherapy that is maintained at follow-up.
The only controlled studies are of cognitive–behavioral psychotherapy and have, in all but one case (Taylor et al., 2002), concerned group treatments. In the exception, Taylor et al. provided individual treatment, but in a secure setting for offenders, thus limiting the generalizability of the results. Evidence for the effectiveness of psychodynamic psychotherapy is comprised of case reports and a few pre–post studies of individual treatment. No controlled studies have been attempted or group treatments empirically evaluated.
Most of the published reports concern studies carried out in normal clinical practice by clinicians with an allegiance to the treatment modality. The range of clinical presentations of recipients was limited; most had problem behavior (including offending) or anger with relatively little attention to the broad range of mental health problems. Further, assessment and outcome variables were small in number and narrow in focus. Prout and Nowack-Drabik (2003) found that in most of the studies they reviewed, investigators used observable behavior as an outcome measure. This was evident, but to a lesser degree, in the studies reviewed here. Many of the measures used were adapted from extant scales or administered with assisted completion, but reliability and validity data are absent. The designs were also compromised because no power analyses were reported. This is of concern when recipients are being allocated to no-treatment or different treatment conditions. Investigators comparing treatments or components of packages found no differences, which is not surprising, due to the small recipient numbers, which reduced the power to detect them. Also, placing people in no-treatment conditions without statistical power to detect differences is poor and unethical practice. Further, no information was provided on how consent to be randomly allocated was obtained. There are problems associated with obtaining consent from people with mental retardation to treatment and research, especially when random allocation is involved (Beail, 2003). Further, inclusion criteria were fairly broad and groups in comparison studies were not homogeneous. In only one study did investigators attempt to match recipients but only on a limited number of variables (Wilner et al., 2002). In the studies in which investigators examined anger management, they did not employ any multi-trait measures of co-morbidity to assess participants or evaluate outcome (Benson et al., 1986: Rose et al., 2000; Taylor et al., 2002; Wilner et al., 2002).
Research on cognitive–behavioral and psychodynamic psychotherapy with people who have mental retardation is negligible when compared to the volumes reported in “What Works for Whom?” Conclusions regarding their effectiveness with people who have mental retardation can only be tentative and modest, but claims of efficacy would be premature. The evidence base for cognitive–behavioral psychotherapy has also progressed more than that for psychodynamic psychotherapy. However, all studies of cognitive–behavioral psychotherapy have concerned self-management approaches. Thus, they are more behavioral than cognitive in content and designed more for skill acquisition than for challenging irrational cognitive processes.
The studies were conducted largely in clinical practice and should be viewed as practice-based evidence rather than evidence-based practice. Both research paradigms complement each other and inform treatment decisions. However, evidence-based research with adequate designs and sufficient participants, reliable and valid measures, and high internal reliability is completely absent. The emerging data suggest that the therapies warrant more thorough evaluation.
Looking back over the last decade, there would appear to be a serious lack of progress with regard to research. As things stand, it is unlikely that persons with mental retardation will be included in reviews such as What Works for Whom? It may be that the attitudinal problems about the provision of psychotherapy for people with mental retardation referred to by Nezu and Nezu (1994) still exist, impacting on funding for services and research. Also, without a convincing body of research on effectiveness and efficacy, such views may be difficult to challenge. We must not forget, however, that the absence of evidence for efficacy is not evidence of ineffectiveness. Alternatively, there may be issues impeding the implementation of research designs of adequate size and methodological rigor, which results in studies not getting off the ground. However, the mental health needs of people with mental retardation are now more clearly identified as being no different than that of the general population or even more prevalent. Thus, there needs to be an ongoing debate and exploration of the issues impeding the development and availability of services and research on the full range of treatments to meet the mental health needs of people with mental retardation.
Author: Nigel Beail, PhD, Consultant, Clinical Psychologist, The Keresforth Centre, Off Broadway, Barnsley S70 6RS, United Kingdom. Nigel.Beail@BarnsleyPCT.nhs.uk