Behaviour Research and Therapy

For decades the development of evidence-based therapy has been based on experimental tests of protocols designed to impact psychiatric syndromes. As this paradigm weakens, a more process-based therapy approach is rising in its place, focused on how to best target and change core biopsychosocial processes in specific situations for given goals with given clients. This is an inherently more idiographic question than has normally been at issue in evidence-based therapy over the last few decades. In this article we explore methods of assessment and analysis that can integrate idiographic and nomothetic approaches in a process-based era. 1. The role of the individual in the coming era of process-based therapy Questioning assumptions in science is disruptive. Within a defined area of study, a priori analytic assumptions provide the scaffolding for which questions are asked, which methods are used, and which data are deemed relevant. Professionals often view questions, methods, and analytic units simply as the required tools of good science – not reflections of assumptions – and as a result there can be a sense of disorientation when times of upheaval arrive and assumptions are pointed out and critically examined. So it is today within the domain of mental health, and the intervention science linked to it. For decades is has been assumed that a satisfactory field of evidence-based treatment could emerge based on adequate experimental tests of protocols focused on psychiatric syndromes. This protocols for syndromes era had a coherent set of key strategic assumptions built into its scientific and public health strategy – but every one of them is now being openly questioned. At the same time, a powerful alternative strategic agenda is emerging that echoes some of the process-based and idiographic assumptions of the earliest days of behavioral research, as well as the therapy based upon it. We are reminded of that history by the very name of this, the oldest of all of the applied behavioral journals. However, revitalization of the study of change processes that apply idiographically is not a mere repeat of the past, since it encompasses questions, methods, and data that are distinct and new (Hayes & Hofmann, 2017, 2018; Hofmann & Hayes, 2018). 1.1. Foundational assumptions of the protocols for syndromes era of evidence-based therapy With the advent of DSM-III in the late 1970s and the arrival of a syndrome-focused approach to applying for research funding, a set of key strategic assumptions gradually shaped the very definition of intervention science. They included: 1. Notable and treatable human biopsychosocial problems reflect latent diseases (i.e. mental illnesses). Mental health is synonymous with or at least is based on the absence of mental illness. 2. While we do not yet know how many mental illnesses there are, they form a discrete and discoverable set. 3. To discover them, a good strategy is to gather together signs (features we can see) and symptoms (features of complaints) into syndromes and sub-syndromes and to use these identified clusters to guide a search for their underlying etiology, course, response to treatment, and mechanisms of pathology and response. 4. Sets of techniques that target biopsychosocial problems should be forged into well-specified protocols that are tested in randomized controlled trials for their ability to change the signs and symptoms of known syndromes. 5. As evidence gathered, meta-analyses, organizational reviews, or state-sponsored examination systems would identify which named treatments are evidence-based, and over time, which ones are best for which syndrome. 6. Public health outcomes will improve when practitioners are made to https://doi.org/10.1016/j.brat.2018.10.005 Received 27 April 2018; Received in revised form 19 September 2018; Accepted 13 October 2018 ∗ Corresponding author. Department of Psychology, University of Nevada, Reno, NV, 89557-0062, United States E-mail address: hayes@unr.edu (S.C. Hayes). Behaviour Research and Therapy xxx (xxxx) xxx–xxx 0005-7967/ © 2018 Published by Elsevier Ltd. Please cite this article as: Hayes, S.C., Behaviour Research and Therapy, https://doi.org/10.1016/j.brat.2018.10.005 use evidence-based treatments, competently and to adherence. Once such treatments are identified, they should be disseminated by all necessary means. 7. Component analyses and studies of processes of change ultimately need to be done, but only for established evidence-based treatments, with the purpose of rounding out our scientific understanding and providing clues about how to simplify our best protocols. 8. Regardless of why they occur, notable changes in established protocols such as adding or subtracting elements, creating online versions, changing dosage, should be re-validated experimentally (see step 4). As a public relations matter, the idea of mental illness (assumptions 1 through 3 above) has spread across the globe. Worldwide, people across the kitchen table discuss what syndromes their acquaintances, loved ones, or public figures “have”. The idea of evidence-based treatments is far less clear in the public eye (Carman et al., 2010). Commercial interests have ensured that classes of medications are well known (particularly in the United States where commercials for psychoactive medications are legal and common). To a much lesser degree, educated members of the public may know that some psychosocial treatment protocols are better supported empirically than others, but they often don't understand what “empirically supported” means (Carman et al., 2010). As a public health matter, it is hard to argue that this era has been a success. The health care burden of mental health problems is increasing (e.g., Larkin, Claassen, Emond, Pelletier, & Camargo, 2005). Treatment is becoming more medical, and chronic. In the ten-year period from 1998 to 2007, people using psychotherapy alone to address mental health issues fell nearly 50% and by 2007 only 1 in 10 people with mental health problems addressed them that way. Using psychotherapy and medications also fell about 30%. Meanwhile 60% of people use nothing but medications – up a third in ten years (Olfson & Marcus, 2010). Psychiatric polypharmacy is rising in vulnerable populations (Maust et al., 2017). Thirteen percent of the United States population over 12 used anti-depressants during 2014, a 65% increase from 15 years earlier; and a quarter of those now on anti-depressants have used them for a decade or more (Pratt, Brody, & Gu, 2017). As a scientific matter, there were enormous increases in data and information secured during the era of protocols for syndromes, but this period did not produce the outcomes it promised. For one thing, research on syndromes did not in fact yield identifiable diseases. Despite decades of pouring significant research dollars into a strategy that was supposed to reveal them, not even one good example exists of this occurring. Given the assumptions behind disease reasoning, it was disheartening that the full mapping of the human genome failed to lead to identifiable genes or gene systems that could explain psychopathology in a direct and robust fashion. For example, in a meta-analysis of the genetic basis of major depression based on genomic analysis of over 100,000 participants, the authors concluded they were “unable to identify robust and replicable findings” (Sullivan et al., 2013, p. 497). High levels of so-called “comorbidity” suggested that clusters of signs and symptoms were not linked to discrete functions (Kupfer, First, & Regier, 2002). When the DSM–5 was released it was to disinterest or controversy, not applause. Meanwhile, on the intervention side, empirically validated protocols proliferated but overlapped, and rarely applied to specific syndromes. Instead, the protocols often lacked treatment specificity (Kupfer et al., 2002), and failed to identify the key components and processes of change. In the US, the NIMH launched a multi-year effort to back away from the assumption that syndromes should be the guiding focus of mental health research (Insel et al., 2010). The latent disease model arguably has had other negative effects. It tended to blind treatment developers to the key role of normal psychological processes in behavioral outcomes, and to the centrality of pragmatic outcomes desired by consumers such as social effectiveness or quality of life, instead prioritizing the referred list of signs and symptoms. It tended to reduce human suffering to brain abnormalities and biological dysfunctions and de-emphasize the centrality of the individual (Greenhalgh, Howick, & Maskrey, 2014) and with it the cultural and biopsychosocial context of individual human lives. Meanwhile practitioners, government entities, and the public in many parts of the world remained unconvinced about the value of evidence-based care. Protocols were at times difficult to deploy, and the lack of known components and processes of change made them difficult to fit to individuals and their complexity. Most patients given psychosocial treatment did not receive evidence-based care (Wolitzky-Taylor, Zimmermann, Arch, De Guzman, & Lagomasino, 2015). Success of the “protocols for syndromes” approach depended on rapid progress toward functional disease entities, or at lease highly specific treatment effects organized by syndromes. When neither appeared, the scientific path toward a mature form of evidence-based therapy turned into a kind of brute force empiricism in which almost everything should be compared to almost everything else in all syndromes or sub-syndromes. The mathematics of that kind of research approach makes it impossible to mount, even if the number of new intervention methods and syndromal or sub-syndromal entities could magically be held to its current number, which it cannot. 2. Process-Based Therapy The earliest days of behavior therapy were characterized by an attempt to apply ev