Contribution
The paper turns a family of complaints about DSM-era psychotherapy research into one programme with a name, process-based therapy (PBT), and one foundational question. It is written jointly by leading figures of the two camps it reconciles: Hofmann as the Beckian, traditional CBT side and Hayes as the acceptance and commitment therapy, “new generation” side. The programme makes the process of change, not the syndrome or the named protocol, the unit of evidence-based care. The paper adds no data. Its contribution is the framing and a list of testable predictions about where the field will go.
Key insight
Gordon Paul’s 1969 question already asked “how does it come about?”. The field answered the rest of the question with “protocols for syndromes” and dropped that part. The authors’ bet is that syndromes are the wrong unit, because they never resolved into diseases. The right unit is the modifiable process that a procedure moves and that in turn moves the outcome. Once that is the unit, a therapy’s name, its school and even the diagnostic category become secondary. Two problems that change by the same processes under the same procedures need not be distinguished.
Assumptions
- Processes are real, modifiable and measurable at the level the paper needs. A therapeutic process is defined as “a set of theory-based, dynamic, progressive, and multilevel changes that occur in predictable empirically established sequences oriented toward the desirable outcomes”.
- Mediation and moderation can carry the evidential load, with the stated caveat that “a statistically significant mediator is not synonymous with a generative mechanism because third variables can always exist”. Mediators and moderators are “first steps” toward causal and functional accounts (Kazdin 2007; Nock 2007).
- Pragmatic outcomes are common ground across philosophies of science. Contextualists take outcomes as truth criteria and “elemental realists” take them as the outgrowth of ontological knowledge; both value them (Biglan & Hayes 2016).
- Change runs through context. “If a dependent variable is going to change in psychology, ultimately it needs to be done by changing history or situational circumstance.”
Key results
The paper has no theorems or experiments. Its results are the definitions, the argument against the syndrome programme, and the predicted trends.
- Foundational question: “What core biopsychosocial processes should be targeted with this client given this goal in this situation, and how can they most efficiently and effectively be changed?”
- Definition of PBT: “the contextually specific use of evidence-based processes linked to evidence-based procedures to help solve the problems and promote the prosperity of particular people”. Procedures are techniques. Processes are the change mechanisms that procedures act through.
- The case against the latent disease model. It assumes symptoms reflect latent disease entities. The DSM-5 itself concedes that validators are insufficient and falls back on clinical utility, and the authors say there is “scant evidence” the model met even that. NIMH funded protocols for syndromes after DSM-III (1980) and began withdrawing from them after 2010 toward RDoC.
- Cited process-level findings (as the paper reports them):
- In panic, verbal information and perceived control changed anxiety under CO₂ challenge, supporting Clark’s catastrophic-misinterpretation model over a purely organic one.
- Exposure-based CBT for anxiety “works in part through cognitive defusion” (Arch et al. 2012).
- Patients high in catastrophic misappraisal did better with capnometry-assisted respiration training than with CBT (Meuret, Hofmann & Rosenfield 2010).
- Anxiety-only patients may do better with traditional CBT and those with comorbid mood disorders with ACT (Wolitzky-Taylor et al. 2012).
- Psychological flexibility can account for part of the effect of the therapeutic alliance as well as of ACT (Gifford et al. 2011).
- Admission rule. “If a procedure reliably produces gains and manipulates a process that mediates these gains, then it is ready to be admitted into the armamentarium of PBT.” If the specified process “cannot be shown to be consistently applicable, the underlying model is incorrect and needs to be set aside”. Conversely, if a procedure fails to move a process, the model “remains untested”.
- Predicted trends: decline of named therapies and of general schools; rise of testable models; rise of mediation and moderation studies; new forms of diagnosis and functional analysis, including complex networks; a move from nomothetic to idiographic methods; processes that specify modifiable elements; the importance of context; component studies returning to the laboratory; new training; integration with the life sciences; new delivery forms (apps, books, calls); a science of the therapeutic relationship; and attention to culture.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | The syndrome-based research programme (1980–2010) did not yield diseases, and its treatment utility was weak | moderate: a historical judgement, partly in the DSM-5’s own words | APA 2013 p. 20; Insel et al. 2010 |
| C2 | Evidence-based therapy should be redefined as processes linked to procedures | normative proposal | argument |
| C3 | A model whose specified process of change does not consistently hold should be set aside even if its intervention works | normative, methodological | argument |
| C4 | Traditional CBT methods work partly through processes named by third-wave therapies (e.g. defusion in exposure) | weak here: one cited study | Arch et al. 2012 |
| C5 | Moderated mediation can ground a more useful nosology: problems with the same processes under the same procedures need not be distinguished | speculative | argument; Kazdin 2007 |
| C6 | Named therapies, broad schools and eventually the term CBT will decline in importance | prediction | trends listed |
| C7 | Processes of change in evidence-based therapy appear not to be strongly culture-bound | weak: “so far appears to be reassuring” | one cited study (Monestes et al. 2016) |
Concepts
- therapeutic procedure: a technique or method the therapist uses to reach the client’s agreed goals.
- therapeutic process: the change mechanism through which a procedure produces the outcome, defined as theory-based, dynamic, progressive and multilevel. It may include common factors such as the alliance, if they meet the same theoretical and empirical standards.
- process-based therapy (PBT): defined above. PB-CBT is its CBT form.
- latent disease model / medical illness model: the assumption that symptoms reflect underlying, latent disease entities, each to be targeted by a specific protocol.
- functional analysis: idiographic identification of “important, controllable, causal functional relationships” for a client’s target behaviours (Haynes & O’Brien 1990). It is presented as behaviour therapy’s original foundation, which psychiatry’s structural nosology displaced.
- mediator / moderator: as Baron and Kenny define them, with the caveat that clinical use blurs baseline moderators into ongoing reassessment.
Connections
- Network theory (LIT-547). The paper’s “complex network approach” is Hofmann, Curtiss and McNally (2016). It describes the approach as “an extension of functional analysis” in which “psychological problems are not expressions of underlying disease entities but rather are interrelated elements of a complex network”. That is Borsboom’s theory in all but citation. Borsboom’s network interventions are a natural reading of what a process is in network terms.
- The single-mechanism alternative (LIT-538). Salkovskis, Sighvatsson and Sigurdsson ask the same Paul question, in its 1967 form without “how does it come about”. They propose one mechanism for all effective therapies: a shift from a stuck negative meaning (“Theory A”) to a tested alternative (“Theory B”), which they call psychological flexibility. PBT expects many processes in a hierarchy, and it suggests that reappraisal might sit under a broader cognitive flexibility. The two converge on “flexibility” as the candidate at the top of that hierarchy, from opposite starting points.
- Engel (LIT-548). The foundational question’s “biopsychosocial” is Engel’s word. The paper does not cite him, and its target is the DSM’s latent-disease model rather than the biomedical model of medicine.
- Meditation’s adverse effects (LIT-066). The paper lists mindfulness among the processes CBT has absorbed and warns that “few processes are always positive regardless of context”. Lindahl et al.’s taxonomy of meditation-related difficulties is an instance of the moderation that warning calls for, though the paper does not discuss it.
Bearing on the record
- No THEORY is indicated by this reading alone. It is a programme, and its strongest content is normative (C2, C3). It is one of three sources for the convergence proposed in the batch report: that disorder is maintained by modifiable processes rather than expressing a latent disease.
- No ML instruction. Nothing here belongs in the anthology.
- Its C3, with LIT-538’s p. 599, is filed as THEORY-053 (Active).
Limitations
- A position paper. Its evidence for the field’s direction is a selection of cited studies, and its trends are predictions.
- It names few processes concretely. Flexibility, defusion, reappraisal and the alliance appear as examples; the hierarchy of processes is explicitly left to future research.
- The mediator problem is acknowledged, not solved. The paper grants that statistical mediation is not mechanism, then makes mediation the admission test. Its answer is the laboratory component study and idiographic measurement, both described rather than shown.
- Interest. The authors disclose book royalties and advisory work. The programme also favours models with a “tight link” to processes, and the paper names ACT, Hayes’s own therapy, as a possible example.
Open questions
- What are the core processes, how many are there, and how do they nest?
- Can moderated mediation in fact recover a nosology, and would it agree with network clusters (LIT-547) or with the single mechanism of LIT-538?
- Is psychological flexibility, as ACT measures it, the same construct as the meaning-making flexibility of LIT-538?
Corrections
- none to a seeded skim (there was no seed)
- Author order. The brief cites “Hayes & Hofmann (2019)”; the paper is Hofmann & Hayes. It appeared online on 29 May 2018.
- Reference slip. The manuscript’s reference list gives the third author of Hofmann, Curtiss and McNally (2016) as “McNally MJ”; Crossref gives Richard J. McNally.