© Springer International Publishing Switzerland 2015Ulrich Schnyder and Marylène Cloitre (eds.)Evidence Based Treatments for Trauma-Related Psychological Disorders10.1007/978-3-319-07109-1_27
27. What Works for Whom?
Division of Dissemination and Training, National Center for PTSD, Menlo Park, CA, USA
Department of Psychiatry and Child and Adolescent Psychiatry, New York University Langone Medical Center, New York, NY, USA
School of Psychology, University of New South Wales, Sydney, NSW, Australia
Department of Psychiatry and Psychotherapy, University Hospital Zurich, Zurich, Switzerland
We are very fortunate to have a volume with chapters that provide rich and detailed descriptions of various treatment programs across the spectrum of trauma-related disorders. This approach allows readers to select the chapter, which describes the type of patient or type of approach of interest, and obtain the desired information on a specific issue. It is also useful, however, for readers to consider the issues raised in chapters that may not seem to directly relate to their usual practice domain because the convergent lessons emerging from these chapters can inform and enhance our treatments for different types of patients. To this end, in this chapter we step back and summarize not only commonalities observed across treatments (see Chap. 1) but also consider differences across treatments and ask the larger question of how best to match treatments to patients. The phrase “what works for whom” asks the question what interventions are of demonstrated benefit to which specific patients groups. This question is of relevance to clinicians and patient consumers alike, as well as clinic administrators, insurers, and policy makers.
The organization of the book reflects basic assumptions about matching treatments to different types of patients. Treatments are presented in a spectrum organized by acuity and complexity. Indeed, there are interventions for those who have severe stress reactions immediately post-event (ICD-10, acute stress reaction; DSM-5, acute stress disorder) and develop sustained reactions (PTSD) or more complex forms or variations of trauma reactions (complex PTSD, prolonged grief disorder). In doing so, it includes diagnoses described in both DSM-5 and ICD-10/ICD-11 but importantly extends beyond trauma-related diagnoses by including considerations of trauma-related disorders comorbid with borderline personality disorder and chronic pain.
The consideration of treatment matching to patient needs extends beyond symptom acuity and complexity. Age is recognized as an important consideration in intervention selection. Treatment interventions are described which have been designed with the cognitive, emotional, and social developmental needs of children in mind and are sensitive to the trajectory of rapid change and growth associated with the first decades of life. The slow decline in health, memory, and cognition of older adults is also addressed and includes recognition of age-specific life events such as the inevitable deaths of friends and family and the loneliness and loss of meaning and identity which may follow.
The volume also takes into consideration the social environment, identifying specific interventions that recognize the role of partners and family, as well as that of neurobiology via identification of state-of-the-science knowledge regarding the use of pharmacological agents and mapping of brain activity to behavior. Treatment matching to patients also requires consideration of access to care and logistical barriers. For this reason, we have included a chapter that discusses the way in which technology can be used to provide services to patients who cannot obtain or do not want face-to-face care. This is sometimes due to geographical barriers or work or family commitments or, alternatively, simply a preference to remain at a distance from provider sites. This may be the case with women veterans who have experienced sexual assaults by fellow military personnel during their service and state a preference for staying away from veteran healthcare facilities.
In this chapter, we first briefly discuss patient-treatment matching issues in regard to critical components of trauma treatment for which there seems to be broad agreement, specifically the therapeutic alliance, and the review and analysis of the traumatic experience. Yet, even here, unknowns remain regarding how to refine these elements to provide maximum benefit to each patient. We will discuss what has been traditionally considered the essential component to patient-treatment matching, namely, patient characteristics. It will be shown that, to date, research in the trauma field has provided little insight about how to optimize patient care. We provide some considerations about directions for future research. Lastly, we will discuss considerations regarding patient-treatment matching for which there seems to be less consensus. This includes the use of multicomponent therapies; the integration of different types of interventions into trauma-focused work such as coping skills training, emotion regulation, and stabilization strategies; and the debate about sequencing versus simultaneous implementation of interventions.
27.2 Common Factors
27.2.1 Therapeutic Alliance
Despite the fact that it is rarely explicitly discussed in any of the treatment chapters, a positive therapeutic alliance is uniformly revealed in the case examples and the interactions presented between the patient and therapist. It is expressed in the therapist’s appreciation for and understanding of the patient’s experience, the framing of the interventions in the context of the patient’s experience, and a sense of warmth and kindness that emanates from the therapist. The therapeutic alliance is the most consistently identified predictor of psychotherapy outcome across types of treatments and patients (Horvath and Symonds 1991; Martin et al. 2000). It is typically defined as comprised of various dimensions including the patient’s sense of being understood, sense of being liked by the therapist, agreement on treatment goals, and agreement on tasks or means toward reaching those goals.
27.2.2 Patient-Therapist Matching
Matching a patient to treatment can include a subset of considerations of matching patient and therapist on ethnic and cultural characteristics. Several studies have found that matches in culture, ethnicity, and gender can improve treatment engagement although effects on outcome are highly variable. Meta-analyses are uniform in finding a strong patient preference for and positive perception of therapists of the same ethnic background, particularly among minorities (Cabral and Smith 2011). However, the benefits of this match on treatment outcome are highly variable – so much so that a recent meta-analysis indicated that the effect size of patient-therapist match on outcome is nearly nil (ES = .09) (Cabral and Smith 2011).
Match on ethnicity and culture may be a proxy for factors potentially more directly aligned with the therapeutic outcome such as shared worldviews, values, and spiritual or religious beliefs. The benefits of the match may be eliminated without shared values and, indeed, potentially result in increased hostility and guardedness. Benefits may vary depending on severity of baseline symptoms and outcome of interest. For example, in a treatment study of PTSD/SUD patients, ethnic/cultural match produced better PTSD outcomes among those with more severe baseline PTSD, while it did very little to influence substance use outcomes (Ruglass et al. 2014). Additional factors may be the patient’s interpersonal skill, determination, and overall resilience in managing the therapeutic relationship and working toward good outcome. For example, it has been found that Black Americans can be less impacted by therapist match than White Americans with the explanation that Black Americans are already quite acculturated to working outside their ethnic/cultural group than Whites (Ruglass et al. 2014).
In summary, ethnic/cultural matching has benefits in facilitating engagement into treatment and in the initial phases of the therapeutic alliance. However, the benefits of match on outcome are strongly influenced by contextual factors such as the presence or absence of shared values, the severity and nature of the problem, and the history behind the relationships of certain ethnic/cultural groups. Therapists need to develop multicultural competencies (e.g., understand different worldviews) while simultaneously avoiding the assumption that any one individual holds the beliefs of his or her ethnicity or culture. Therapists also need to be aware of their local cultural/ethnic context and the fact that racial and ethnic tensions and differences are dynamic and change over time and across regions.
27.2.3 Therapist Characteristics
It is also interesting to speculate whether certain dispositional characteristics in a therapist might be more or less appealing to a patient: the quiet therapist versus the more talkative person, the physically restless versus the more introverted and contained individual, and the directive versus open-ended therapist. For trauma work in particular, patients may be differentially sensitive to therapists’ characteristic reactions and behaviors in response to trauma disclosure (e.g., sadness, outrage, disgust, embarrassment, concern, optimism about recovery) as well as the approach the therapist takes in introducing trauma-focused treatment elements. Some therapists quickly and explicitly introduce discussion of the trauma, while others wait for an inquiry from the patient. It might be valuable to ask patients what their preferred professional and personality characteristics might be as a means to increasing engagement and possibly improving outcome. Lastly, it is worth considering how the therapeutic relationship interacts with treatment activities and process. For example, an active therapist who provides repeated behavioral demonstration of interventions may facilitate better skills training outcomes, while a more contained, reactive (vs. proactive) therapist who focuses on attunement with the patient’s emotional state may facilitate better outcomes during imaginal exposure. The therapist’s capacity to flexibly shift in degree and kind of behavioral and emotional expression as the tasks of the treatment change may be an important therapeutic skill in treatment. To date, there is little research or information about the impact of therapist characteristics, attitudes, and behaviors on the patient and on treatment outcome.
Many therapists put significant emphasis on helping their patients regain control over their exaggerated emotional reactions (e.g., flashbacks) and encourage them to take charge of their lives. In psychoeducation at the beginning of treatment, they explain that one of the core elements of most traumatic events is the experience of an (in many but not all cases, sudden and unexpected) unwanted and extremely unpleasant loss of control. They also explain that the typical symptoms of PTSD, particularly the reexperiencing and hyperarousal cluster symptoms, can be seen as a repetition of that same loss of control. Therefore, they argue that regaining control is one of the main goals of trauma treatment. There is strong experimental evidence that believing one has control enhances capacity to manage distress (Bryant et al. 2014). Therapists also have good clinical reasons to facilitate control in their patients: once patients have regained control over their memories to the degree that they can decide at any given time whether or not they want to go back in their imagination to the time when the traumatic experience happened, much has been achieved! However, the emphasis on control may overwhelm other important goals of the treatment. Helping patients regain control always needs to be counterbalanced with the recognition that much of life is beyond our control. Many things in life, bad things as well as good, just simply befall us. We fall ill, we fall in love, and we can’t and sometimes also don’t want to do anything about it. Rather than pushing our patients to try to control everything in life, we might want to help them be better able to discriminate controllability. In a situation that is important to them and where they have some degree of control, they should learn to be assertive and courageous and try to exert their influence. However, in a situation beyond their control and power, they should learn to be wise enough to accept and adapt.
27.2.4 Review and Analysis of the Traumatic Experience
There is general agreement that direct attention to and explicit review of the traumatic events, when effectively done, produces superior reduction in PTSD and related symptoms compared to therapies that do not. In all of the interventions presented in this volume, the traumatic events are described in words and, occasionally, with additional representational media (drawings, letters, objects). Explicit attention to traumatic experiences and memories is carried out in different ways including through imaginal exposure, cognitive reappraisal, or narrative reconstruction. The goals of these activities are the same: to reduce or resolve feelings of fear, anger, shame, guilt, etc., to develop a coherent understanding of the events, and to create meaning from them.
Questions that clinicians often ask concern the timing of the trauma work and its intensity and duration. There is very little research to help guide the clinician in determining what is best and more particularly what is best for any one patient. Some therapies involve detailed and repeated review of the traumatic events, with an emphasis on sensory-perceptual detail (e.g., PE, NET). Others focus on the development of a narrative and emphasize attention to feelings that have been disregarded (e.g., BEPP) or to maladaptive beliefs (e.g., CPT) and do not emphasize repetition. Still others emphasize the development of an autobiography (NET) and focus on linking themes (e.g., self-identity) across narratives of different events (STAIR Narrative Therapy).
Randomized controlled trials comparing evidence-based trauma-focused therapies are few, and those available indicate that outcomes differ little between interventions (e.g., Nijdam et al. 2012; Resick et al. 2002; Rothbaum et al. 2005; Taylor et al. 2003). However, it is probable that patients are not equally interested in and motivated by all interventions. It is also probable that patients do not benefit equally from all interventions. Recent research in both pharmacological interventions and psychotherapies has found that patients do have treatment preferences and that providing patients with their preferred treatment provides superior results relative to assignment to a treatment by chance (i.e., randomization) (Swift and Callahan 2009). From a clinical perspective, these data would suggest that treatment decisions regarding type of trauma-focused treatment might best be made following the patient’s preference, and options that occur during the course of treatment regarding timing, intensity, and duration of trauma-focused treatment include shared decision making. Research regarding the impact of patient preference on treatment selection and outcome is limited but of increasing interest.
27.3 Patient-Specific Characteristics
To date, there have been at least 20 studies assessing patient-specific characteristics as predictors of PTSD treatment outcome. A systematic review of all the studies indicates very little consistency in results (Cloitre 2011). For every study that has identified a particular patient characteristic which predicted outcome, there is at least one other study which obtained null results. Characteristics that have been evaluated include trauma history (e.g., childhood abuse), demographic characteristics such as age and education, severity of PTSD, and co-occurring symptoms including anger, anxiety, depression and dissociation, borderline personality features, and personality disorders as well as factors such as intelligence, self-esteem, and beliefs about self and the world.
Possible reasons for these inconsistencies include low sample size which limits power to detect real differences as well as great differences in study samples (e.g., inpatient versus outpatient) and interventions. More importantly, the statistical and conceptual approach to identifying predictors may be flawed. The traditional approach presented in reports has been to evaluate individual factors as predictors of outcome, in an effort to find the “silver bullet.” However, as is well known, there are great symptom heterogeneity across PTSD patients and, by implication, a multitude of possible predictors, suggesting that the “silver bullet” approach will likely fall short of capturing clinical reality.
An alternative and potentially successful approach has emerged in general medicine which has been investigating disorders that, similar to PTSD, admit of significant complexity and heterogeneity (e.g., diabetes). Statistical analyses have revealed that there are typically multiple moderators of outcome, each of which is weakly predictive and where no single predictor is strong enough to provide information that is clinically meaningful. The alternative approach allows the statistical generation of a combination of factors (i.e., a profile) that provides the strongest predictor of outcome. For example, coronary heart disease is calculated on the severity of risk factors in combination with each other (e.g., age, blood pressure, cholesterol, diabetes, smoking). In this type of modeling, no one risk factor is critical, but rather each contributes a certain amount that leads to an overall “outcome” (occurrence of coronary heart disease). This approach has now been applied to treatment intervention outcomes in psychiatry (e.g., Wallace et al. 2013). For the treatment of PTSD, it may similarly be possible to identify a moderator “profile” that captures the key aspects of a patient that, as in the case of coronary heart disease, include a range of factors such as historical information (e.g., trauma history), symptoms (e.g., anger), current co-occurring disorders (e.g., dissociative disorder), and behaviors that can cumulatively predict outcome.
Although in its infancy, it is worth noting that there are preliminary studies indicating that treatment response is linked to certain neurobiological characteristics of PTSD patients. For example, there is recent work indicating that people differ in treatment response according to their genetic profile, in response to both psychotherapy (Bryant et al. 2010a; Felmingham et al. 2013) and pharmacotherapy (Mushtaq et al. 2012). Further, brain imaging studies indicate that brain structure impacts on how people respond to psychotherapy (Bryant et al. 2008a), as well as brain functioning prior to commencement of treatment (Bryant et al. 2008b; Falconer et al. 2013). Although this work is very new, it underscores the conclusion that at a fundamental level not all patients will respond to treatment equivalently.