The paradox of non-adoption

Natoli et al, outline both the established limitations of categorical approaches and the evidence supporting the utility of dimensional models. Yet this evidence highlights a stark reality: despite a strong evidence-base (Hopwood Reference Hopwood, Kotov and Krueger2018a) and demonstrated clinical utility (Bach Reference Bach and Tracy2022), implementation remains severely limited. DSM-5's alternative model for personality disorders (AMPD) is now over a decade old, yet remains largely unknown to clinicians in practice, absent from most post-graduate training curricula and disconnected from government and insurance remuneration schemes. This is particularly perplexing as dimensional approaches are readily learnable, with both undergraduate and professional graduate students demonstrating high inter-rater reliability and score equivalence with expert ratings (Few Reference Few, Miller and Rothbaum2013; Zimmermann Reference Zimmermann, Benecke and Bender2014; Garcia Reference Garcia, Skadberg and Schmidt2018; Morey Reference Morey2018; Garner Reference Garner, Blocher and Tierney2022). Further, clinicians rate dimensional approaches more useful for treatment formulation and communication, with an overall preference for dimensional approaches. Perhaps some resistance lies in these same studies finding that some form of hybrid approach is still preferred, suggesting something alluring about a qualitative label (Bernstein Reference Bernstein, Iscan and Maser2007; Morey Reference Morey, Skodol and Oldham2014, Reference Morey and Hopwood2020). The gap between evidence and practice extends from theoretical and empirical challenges to institutional inertia – layers of bureaucracy, entrenched billing systems and decades of categorical treatment approaches create substantial barriers to adoption (Brown Reference Brown, Sellbom and Bach2023).

Natoli et al's proposed hierarchical framework may address several of these barriers by matching common factors to severity assessment and treatment intensity. Although this approach aligns with existing practices and provides the qualitative labels needed by clinicians and institutions – particularly given that personality disorder severity appears to be the primary prognostic indicator (Crawford Reference Crawford, Koldobsky and Mulder2011) – it raises complex questions about treatment targets. Evidence suggests that severity (personality functioning) changes over the course of therapy whereas personality style may be much slower to change, suggesting that individuals ‘stay essentially who they are’ (Natoli 2005) while developing more adaptive environmental interactions. This suggests a need to develop more adaptive manifestations of high traits while potentially ‘taking the edge off' these extremes. Previous work suggests that personality functioning is the more malleable and important target (Wright Reference Wright, Hopwood and Skodol2016), yet individual cases may require a focus on reducing trait elevations when adaptive manifestations at certain extremes prove unrealistic.

A step forwards on a much-needed track

The proposal for modularised treatments for specific trait and impairment combinations is well-reasoned, building on existing evidence-based approaches (Hopwood Reference Hopwood2018b; Ruggero Reference Ruggero, Kotov and Hopwood2019). This initial emphasis on guidelines was necessary – developing treatments before establishing stable models risks wasting resources and potentially discouraging clinicians back to categorical approaches. However, with more mature dimensional models now available, the field must shift focus towards practical implementation. The treatment of Morris in Natoli et al's case vignette is a nice illustration and guidance for future research to follow, and I hope the challenge of real-world evidence is accepted soon.

Changes to diagnostic systems must demonstrate improved patient outcomes to justify the disruption to healthcare systems (Lahey Reference Lahey, Moore and Kaczkurkin2021; Zimmerman Reference Zimmerman2022). Although initial frameworks could serve as scaffolding for building this evidence base, we still lack the randomised controlled trials across multiple sites that would demonstrate superior outcomes compared with existing approaches (Zimmerman Reference Zimmerman2022). Larger scale trials also meet another barrier, as standardisation becomes more challenging when moving away from categorical diagnosis-specific protocols towards personalised approaches based on trait profiles (Krueger Reference Krueger and Markon2014; Waugh Reference Waugh, Hopwood and Krueger2017). This is not to say that treatment approaches should not be individualised, and there are respectable arguments for the perils of providing this type of ‘cookie-cutter’ treatment; however, standardisation to some degree underpins reliable clinical trial data.

Demonstrable outcomes could manifest as shorter treatment durations, better recovery rates, lower readmission rates or – perhaps most compelling to institutions – reduced healthcare costs. To achieve this, we will likely need to see broader systemic changes. Training programmes should integrate dimensional approaches into their core curricula, moving beyond brief introductions to provide sustained practical experience (Zimmermann Reference Zimmermann, Benecke and Bender2014; Monaghan Reference Monaghan and Bizumic2023). This will mean that the next generation of clinicians will be familiar and confident enough with these models to integrate them into their practice.

To further reduce the implementation hurdle, treatment protocols could integrate dimensional-based frameworks with existing evidence-based treatments to draw on existing institutional knowledge, confidence and resources (e.g. cognitive–behavioural therapy, dialectical behaviour therapy, interpersonal psychotherapy). The psychodynamic tradition's emphasis on modelling adaptive interpersonal styles might be particularly relevant for certain trait profiles. Active work on the clinical demonstration is progressing within the Hierarchical Taxonomy of Psychopathology framework (Ruggero Reference Ruggero, Kotov and Hopwood2019). Data gained from these approaches might provide enough evidence to convince larger healthcare systems to change – if these approaches are indeed superior in practice.