Treatment of refugees at the crossroads: The need for an evidence base

The paper by van Wyk and colleagues (2012) represents a milestone in the Australian research literature in that it is the only follow-up study of its type that examines the impact of counselling on traumatic stress symptoms amongst recently arrived refugees. The findings are encouraging in that they indicate general improvement in post-traumatic stress disorder (PTSD) and other symptoms amongst Burmese refugees over the course of counselling.

It is well recognized that naturalistic studies of this type do not allow close analysis of the mechanisms of change; in this case, whether improvement can be attributed to counselling, to the honeymoon effect of arriving in a safe country or to other influences. Nor is it possible to make fine-grained comparisons between studies undertaken across countries. For example, we can only speculate why the Australian study recorded positive outcomes, whereas a similar naturalistic inquiry undertaken in the well-established Rehabilitation Centre for Torture Survivors in Denmark reported only minimal improvement in symptoms amongst refugees following extensive therapeutic interventions (Carlsson et al., 2005, 2006). It is possible that sample-specific differences account for the disparity: the Australian study involved refugees whose post-traumatic stress symptom scores were in the moderate range; in contrast, the Danish treatment group included a large number of torture survivors whose baseline symptom scores were notably high. The modest treatment outcomes in Denmark therefore may reflect one of the paradoxes of treatment: well-established clinics such as the Danish centre tend to congregate complex and chronic cases who are most likely to be treatment resistant.

These ongoing uncertainties about the effects of treatment only serve to emphasize the need for rigorous intervention trials to be undertaken in the field of refugee mental health. The reasons for the existing dearth of studies undertaken in Australia are readily evident to those working in the field. The past 25 years have been the pioneering period in which mental health services for refugees were established for the first time in Australia. The priorities have been to establish and consolidate these fledgling services across all states of Australia; to ensure their sustainability and growth; to develop durable relationships with the diverse communities they serve; to foster state-wide and national networks; and to formulate and implement effective training and supervision programs. Leaders in the field have had to devise novel conceptual frameworks and treatment approaches in a field where the existing theoretical and empirical base has been slender. Taking these constraints into account, the achievements of Australian services have been remarkable: they have emerged as leaders in the field internationally in developing comprehensive, multimodal models of intervention that integrate key principles from diverse disciplines including mental health, human rights, transcultural psychiatry and psychology, psychiatric traumatology and community development.

Even in the best circumstances, the field faces particular challenges in undertaking intervention research. The contentious political climate surrounding the admission and treatment of asylum seekers in Australia has done nothing to promote dispassionate research. There are realistic concerns that negative treatment results could fuel propaganda portraying refugees as disabled persons who impose an ongoing drain on health and welfare services. Within the field, controversy has continued about the epistemological validity of applying western diagnostic constructs such as PTSD and positivistic methods of scientific inquiry within a transcultural setting (Miller et al., 2006; Summerfield, 1999). From a practical perspective, the pressures of service provision often limit the capacity of personnel to pursue rigorous and time-consuming research protocols. Ethical considerations also loom large, with service providers rightfully being wary of study designs that allocate distressed patients to wait-list or placebo conditions. Finally, recently arrived refugees are preoccupied with the practical exigencies of survival and resettlement and these pressures may temper their motivation to engage in interventions focusing on trauma-related psychological reactions – at least in the short-term (Silove et al., 2002). Furthermore, the immediate stresses refugees experience on arrival could attenuate or distort the impact of therapy offered at that stage.

Notwithstanding these challenges, there are critical questions that need to be resolved about the preferred psychotherapeutic interventions for survivors of refugee trauma. For example, we need to know whether the treatments recommended for PTSD in civilian settings can be applied equally to refugees, particularly techniques that focus on systematic exposure to past traumatic experiences (Forbes et al., 2007). The debate about such uncovering therapy has remained contentious from the early period of the establishment of services. For example, more than 20 years ago, therapists from Denmark described their approach to therapy as follows: ‘If a victim (of torture) has strong repressive mechanisms he is asked to tell his story in chronological order. The victim must describe precisely what he experienced during isolation and the time spent waiting for torture’ (Somnier and Genefke, 1986).

The present author suggested an alternate perspective, as follows: ‘… therapists need to consider carefully the patient’s capacity to cope with the stress of self-disclosure and catharsis while struggling with the immediate demands of resettlement … The adverse effects of premature uncovering of traumatic experience may be particularly dangerous in (some) torture survivors …’ (Silove et al., 1991).

Remarkably, the ensuing debate has remained unresolved over the decades, at times erupting into acrimonious exchanges amongst experts in the field. For example, in a commentary published in the British Medical Journal, Metin Basoglu decried what he claimed to be the failure of refugee services to apply evidence-based interventions (meaning trauma-focused cognitive behavioural therapy (CBT)) in the treatment of torture survivors (Basoglu, 2006). The paper provoked a large volume of rejoinders from prominent mental health workers in the field. Most repeated the widely accepted view that therapy must be based on a multimodal framework that addresses the full range of psychosocial and acculturation challenges faced by refugees. Moreover, some commentators identified the potential risks of exposing torture survivors to memories of the abuses they have suffered.

The most important issue that this debate has uncovered, however, is the dearth of high-quality research in the field. Recent reviews and meta-analyses invariably have concluded that the absence of high-quality data preclude any definitive conclusions being drawn about preferred therapies for refugee and post-conflict populations experiencing ongoing stress reactions (Nickerson et al., 2011; Tol et al., 2011). At best, the data that exist provide provisional evidence supporting the use of trauma-focused CBT for the treatment of PTSD amongst refugees and other survivors of mass conflict; there is even less evidence to support the efficacy of multimodal therapies. The stark reality is that we continue to work in a field where interventions are shaped more by the proclivities of the therapist than by rigorous scientific evidence.

As refugee services in Australia reach their 25-year anniversaries, it is timely to take stock and consider the challenges we face over the next quarter of a century. In this author’s view, a critical issue is whether we can continue to apply models of treatment, however credible, without subjecting them to scientific testing. In spite of the daunting human rights, ethical, cultural and service-related challenges in undertaking the necessary studies, it behoves us to demonstrate that the treatments we use actually work. The core interlocking questions are evident: what we need to know, ultimately, is how to select the most appropriate interventions for refugees from diverse cultural backgrounds who are located at different points in their resettlement trajectories and who present with a range of psychological reactions related to past trauma and ongoing stresses. No single study or methodology will answer all the outstanding questions. It will require a systematic program of research, ideally undertaken across centres, to begin to piece together the necessary data.

In pursuing this endeavour, creative research designs will need to be devised to address all the aforementioned challenges. It is feasible and ethical, for example, to add one or more components of therapy to multimodal interventions using a randomised controlled trial design, especially if there is no extant evidence to indicate whether these individual elements contribute to positive therapeutic outcomes. Furthermore, many services are compelled to have waiting lists, making it ethical to apply wait-list-controlled treatment studies. The issue of individual differences in therapeutic response can be addressed by including qualitative components that may throw valuable light on the specific nature of personal, familial and ecological factors that influence the outcomes of therapy.

It is a natural tendency to want to protect refugees from ‘experimentation’ given their background experiences of persecution and victimization. We should not overlook the fact, however, that refugees have as much interest in the advancement of science and the development of therapeutic methods as we do. Their courage in braving dangerous voyages when the outcomes are uncertain bears testimony to their capacity to confront challenges and to maintain hope. They may be equally courageous in their willingness to engage in novel treatments that offer some prospect of assisting them and future refugees, as long as they are assured that the interventions being offered are well-founded and directed by an ethically informed science.