TRANSD recommendations: improving transdiagnostic research in psychiatry

There is no doubt that transdiagnostic research in psychiatry has gained momentum over recent years. However, what is meant by transdiagnostic research, and the impact it has on current psychiatric practice, is much less clear. The adjective “transdiagnostic” itself does not exist in English dictionaries, and even online medical dictionaries recommend searching the words “trans” and “diagnostic” separately. The word “transdiagnostic” is not only a neologism, but it also exclusively applies to psychiatry. While diagnoses are ubiquitous in medical research and practice, there are no consolidated exemplars of transdiagnostic research in other branches of medicine.

To characterize the actual meaning and the clinical impact of transdiagnostic research in psychiatry, a systematic review was recently conducted following state‐of‐the‐art evidence synthesis guidelines1. Although, as a matter of fact, the word “transdiagnostic” has been historically introduced by cognitive behavioral theories and treatments for eating disorders2, in that review1 there was no restriction on any a priori definition of transdiagnostic research. On the contrary, the review focused on articles reporting on any transdiagnostic topics: interventions (45%), cognition and psychological processes (28%), neuroscientific topics (13%), classification (4%) and prediction studies (10%).

To systematically appraise the evidence without superimposing a priori conceptual schemata of transdiagnostic research, the review performed an epistemological test and empirically included and interrogated articles that self‐proclaimed transdiagnostic by explicitly using the word “transdiagnostic” in their title1. High‐order conceptual reviews of research initiatives that have implicitly adopted a transdiagnostic approach, such as the Research Domain Criteria (RDoC) project, the Hierarchical Taxonomy of Psychopathology (HiTOP) approach, the p‐factor construct (none of which have yet replaced the current classification systems in clinical routine), and the clinical staging model, have been recently presented and fully debated in this3, 4, 5 or other6 journals, and as such were not the main focus of the systematic review1.

The core finding of this review was that transdiagnostic designations in psychiatry are applied in a loose and unstandardized way, encompassing several different and often incoherent conceptualizations1. For example, one would expect studies that self‐proclaim transdiagnostic to somewhat address issues relating to the diagnosis of mental disorders. Paradoxically, some of the studies reviewed were intrinsically incompatible with a transdiagnostic framework because they investigated symptoms and not disorders or, to the extreme, reported no diagnostic information at all1.

Another illustrative example is the fact that authors themselves disagree on the ultimate aim of transdiagnostic research. Some of them claim that transdiagnostic research is a fundamental pathway to clinical utility for improving psychiatric classification and diagnosis7, while others argue that the transdiagnostic approach does not primarily target the improvement of psychiatric classification and diagnosis, but rather tests a general theory of psychopathology8. A further example is the fact that, until the publication of this systematic review1, the empirical limitations and reporting quality of transdiagnostic research remained unaddressed: appraising and acknowledging the specific limitations of a certain domain of knowledge is equally, if not more, important as celebrating its successes.

It may well be that some versions of a transdiagnostic approach are going to be necessary to improve psychiatric classification and care7. What is certain is that, until studies continue to loosely and incoherently self‐proclaim transdiagnostic without acknowledging any diagnostic information, it is unlikely that transdiagnostic research will bear any real‐world meaning for clinicians, patients, and medical practice. Similarly, poor reporting on the number and type of (trans)diagnostic spectra prevents the appraisal, refinement, and eventual integration of categorical and dimensional approaches in psychiatric classification.

The systematic review acknowledged that transdiagnostic categorical approaches that respect dimensionality are possible in organic medicine as well as in psychiatry1, but this requires transparent reporting of the results. For example, the largest transdiagnostic study published to date demonstrated that it is possible to report the diagnostic information for almost all ICD‐10 mental disorders9. Furthermore, while it is possible that transdiagnostic interventions may display superior efficiency, cost‐effectiveness, accessibility, and patient‐reported satisfaction compared to specific‐diagnostic interventions8, demonstrating this would require robust comparative analyses specifically conducted to test the non‐inferiority or superiority of the transdiagnostic approach. These analyses are infrequent in the current literature1.

The systematic review leveraged these caveats to put forward six empirical transdiagnostic research recommendations: TRANSD1. The TRANSD recommendations are pragmatic and focus on improving the quality of appraising and reporting transdiagnostic constructs. Importantly, they do not provide any a priori restrictive definition of the transdiagnostic schemata; as such, they can be applied to different topics and stimulate critical research in the field.

The first recommendation is to have a transparent definition of the gold standard (ICD, DSM, other), including specific diagnostic types, official codes, primary vs. secondary diagnoses, and diagnostic assessment interviews. Second, the primary outcome of the study, the study design, and the definition of the transdiagnostic construct should be reported in the abstract and main text. Third, the conceptual framework of the transdiagnostic approach – across‐diagnoses (comparing different ICD/DSM categorical diagnoses against each other), beyond‐diagnoses (employing ICD/DSM diagnostic information to go beyond it, testing new diagnostic constructs such as biotypes), other (with an explanation of the conceptual framework) – should be appraised. Fourth, the diagnostic categories, diagnostic spectra, and non‐clinical samples in which the transdiagnostic construct is being tested and then validated should be indicated. Fifth, the degree of improvement of the transdiagnostic approach should be shown against the specific diagnostic approach through specific comparative analyses. Sixth, the generalizability of the transdiagnostic construct should be demonstrated through external validation studies.

It is hoped that these recommendations will improve the transparency and consistency of the next generation of transdiagnostic research, overcoming the current limitations of knowledge and benefitting psychiatric care.