Introduction: why do we need phenomenology in psychiatry?

Recent years have witnessed a gradual accretion of knowledge about mental disorders, as well as incremental advances in evidence-based treatments. Nevertheless, few new treatments have been developed, and clinical research has fallen short of its promise to deliver better mental healthcare for all (Thornicroft, Reference Thornicroft2007; Leichsenring et al., Reference Leichsenring, Steinert, Rabung and Ioannidis2022). Despite an ever-increasing, evidence-based body of knowledge to aid clinical and policy decision-making, considerable research-practice and treatment gaps remain (Stein et al., Reference Stein, Shoptaw, Vigo, Lund, Cuijpers, Bantjes, Sartorius and Maj2022). The heaviest burden, in terms of years lived with disability, falls upon children and adolescents, with significant implications for young people's ability to participate in education, family and occupational life (Gustavson et al., Reference Gustavson, Knudsen, Nesvåg, Knudsen, Vollset and Reichborn-Kjennerud2018; Dalsgaard et al., Reference Dalsgaard, Thorsteinsson, Trabjerg, Schullehner, Plana-Ripoll, Brikell, Wimberley, Thygesen, Madsen, Timmerman, Schendel, McGrath, Mortensen and Pedersen2020).

While there is reason to hope that neuroscience and genetics will deliver the kind of hard-science certainties that psychiatry – as a medical discipline – aspires to, much work is still required to develop a genuinely personalised and ethically responsive practice. To this end, psychiatry – as the discipline that strives to make sense of abnormal human subjectivity – needs a pluralistic methodological and ethical framework that can connect explaining with understanding and caring (Stanghellini and Broome, Reference Stanghellini and Broome2014). Such a knowledge is integral to the very practice of medicine and clinical care, regardless of perceived scientific maturity. This tension, between the human and the biological sciences, lies at the very core of phenomenological psychiatry and, arguably, of psychiatry generally as a discipline. But this should not be regarded as misfortune. Rather, it points to the strength, complexity and excitement of our field. The challenge is, then, to create an integrative framework that can accommodate – within psychiatry – both sides of the same coin.

Emerging from the philosophical tradition of phenomenology (with its central figures of Husserl, Heidegger, Merleau-Ponty, Sartre and Stein among others), phenomenological psychiatry has a history dating back to Karl Jasper's founding text General Psychopathology (Jaspers, Reference Jaspers1963; originally published 1913). Jaspers was well aware of the aforementioned methodological challenges, and tensions, inherent to the study of psychopathological syndromes. Finding himself in a historical moment – after the ‘first biological psychiatry’ – conceptually not dissimilar to the present terrain laid out by the ‘decade of the brain’, Jaspers reflects critically on the obscurity and lack of common theoretical language in psychiatric discourse (Broome, Reference Broome, Stanghellini and Fuchs2013). While acknowledging the need for ‘certain general concepts and laws’ (Jaspers, Reference Jaspers1963, p. 1) and thus for reliable classifications, Jaspers is also clear that ‘psychopathology is limited, in that there can be no final analysis of human beings as such’. The more we reduce them to what is typical and normative, the more we realise there is something hidden in every individual that defies recognition: ‘We have to be content with partial knowledge of an infinity which we cannot exhaust’ (ibid., p. 1). Psychopathology, he continues ‘is concerned with every psychic reality which we can render intelligible by a concept of constant significance. The phenomenon studied may also be a matter of aesthetic, ethical or historical interest, but we can still examine it psychopathologically’ (ibid., p. 2).

Jaspers’ lesson remains highly relevant in the contemporary culture of categorical classifications and standardised quantitative data collection, dominated by a frantic search for the neurobiological cause that will explain psychopathological experiences once and for all. For instance, there are – within clinical research – a number of widespread but questionable assumptions such as: (1) that specific categories of signs and symptoms (e.g., delusions or hallucinations) will translate directly, relate meaningfully and reduce smoothly, to the lived experience of mental disorders and their neural correlates (i.e., a particular symptom or experience can be redescribed in the language of cognitive neuroscience, without loss of any richness of the phenomena); (2) that ‘statistically significant evidence’ for or against the effectiveness of a certain treatment in controlled conditions can be treated equally to ‘personally significant evidence’ in clinical practice; (3) that patient-centredness and shared decision-making can be unproblematically achieved through a mathematical weighing of patient preferences, research evidence and clinical expertise. But just as the reduction of consciousness to a mere product of neurophysiological events has proven difficult (Levine, Reference Levine1983; Chalmers, Reference Chalmers1996), the bio-reductionist research agenda of the past 40 years in psychiatry has failed to close the explanatory gaps between a given psychopathological phenotype, its modular neuro-cognitive substrates or processes, its proposed pharmacological or psychological intervention, and their translation into meaningful and effective treatment for those in need.

As others have already argued on phenomenological grounds (Schwartz and Wiggins, Reference Schwartz and Wiggins1985; Mullen, Reference Mullen2007; Parnas et al., Reference Parnas, Sass and Zahavi2012; Nordgaard et al., Reference Nordgaard, Sass and Parnas2013; Sass, Reference Sass2022), we believe that psychiatry's enduring tendency to conform to a philosophy of operationalism, at the expense of more genuinely pluralistic and multi-layered methodological enquiries into the person's subjectivity, may have become a self-sustaining form of stagnation and impediment to the generation of new knowledge. In line with other recent calls for more phenomenology in psychiatry (Larsen et al., Reference Larsen, Maschião, Piedade, Messas and Hastings2022), we believe that phenomenological concepts and methods can act as a fruitful corrective for contemporary psychiatry – with the proviso that a stance of openness, provisionality and humility is adopted (Ritunnano et al., Reference Ritunnano, Stanghellini, Fernandez, Feyaerts and Broome2022a). Phenomenological psychiatry is specifically aimed at grasping the existential structures (and alterations thereof) that give formal coherence and meaning to our experience of world. As such, it is not just illness-oriented, but also person-oriented; it makes room for symptoms both as a source of distress and meaning-making process . Thus, phenomenology offers a way to develop an enriched, person-centred, evidence-based psychiatry that takes subjectivity seriously when selecting the object of enquiry, targets of treatment and preferred outcomes (Stanghellini and Broome, Reference Stanghellini and Broome2014).

Our proposal bears significant ethical implications for both research and practice in mental healthcare, where the alleged value-neutrality of operational epistemologies has often led to the dismissal of the perspectives of people who live with mental disorders. In the past, this has led to localised and structural forms of epistemic injustice (Box 1) derived from differential power relations (e.g., patient/physician; participant/researcher; policymakers/communities) across healthcare research and services, where many have voiced feelings of being persistently ignored, dismissed or marginalised by health professionals (Carel and Kidd, Reference Carel, Kidd, Kidd, Medina and Pohlhaus2017; Harris et al., Reference Harris, Andrews, Broome, Kustner and Jacobsen2022; Ritunnano, Reference Ritunnano2022). While there is now (at least in high-income countries) a growing recognition of the importance of patient and public involvement within the field of mental health research and service improvement, it is still the case that meaningful participation of service-users and carers as active collaborators in the research process is not yet systematically sought (Montori et al., Reference Montori, Brito and Murad2013; Schünemann et al., Reference Schünemann, Wiercioch, Etxeandia, Falavigna, Santesso, Mustafa, Ventresca, Brignardello-Petersen, Laisaar and Kowalski2014; Zhang et al., Reference Zhang, Coello, Guyatt, Yepes-Nuñez, Akl, Hazlewood, Pardo-Hernandez, Etxeandia-Ikobaltzeta, Qaseem, Williams, Tugwell, Flottorp, Chang, Zhang, Mustafa, Rojas, Xie and Schünemann2019).

Box 1. Key terms

Therefore, as we move closer to a fuller understanding of subjective life with the potential to improve psychiatric interventions, a new integrative framework is needed that acknowledges and values the role of subjectivity, personhood and existential meanings, alongside traditional research data. By drawing on a range of different value perspectives, this framework can aid decision-making processes in mental health research. Here, we focus on three key actionable areas where we see possibilities for engagement between phenomenology and mainstream psychiatric research: (1) defining the object of interest or ‘caseness’; (2) integrating phenomenological methods: promises and challenges; (3) identifying meaningful outcomes and new targets for psychological treatment. In Box 1 below, we provide accessible definitions of relevant technical terms.

Defining the object of interest or ‘caseness’

Perhaps the one area where phenomenology has the greatest potential to be swiftly employed, to improve the quality of interventional and outcomes research in mental health, is that of ‘caseness’. In this context, we use the term caseness to refer to the degree to which accepted standardised diagnostic criteria, or psychometric tools for a given condition, can validly and reliably distinguish cases as cases rather than controls, or distinguish between different clinical groups within a study (for instance on the basis of severity or risk stratification), and define the boundaries between such groups.

Depending on the study design, caseness is a key research strategy required to ensure diagnostic and prognostic homogeneity, and draw reliable conclusions. For example, in randomised controlled trials, failure to assemble participants into groups which are (as much as possible) prognostically similar may lead to biased findings that cannot reliably or meaningfully guide practice. In non-interventional cohort studies, poor caseness may lead to erroneously identifying participants as having developed a certain pathological condition, again leading to biased findings about its aetiology. Without being able to identify who is or is not affected, whom is to treat and what is most likely to work, clinicians may also struggle to make informed clinical decisions.

By providing a more detailed psychopathological characterisation of the individual case, we believe that phenomenology may help clinical researchers with the task of assembling prognostically homogeneous patient groups, for the purpose of investigating the effectiveness of a new intervention. It may also help guide aetiological and prediction research within non-interventional study designs. A tangible example of this potential is provided by the application of phenomenological insights for the purpose of early identification and prediction of psychotic disorders, holding potential for translation into early treatment and prevention of deleterious outcomes.

Over the last 25 years, advances have been made in identifying young people at heightened risk of schizophrenia and other psychoses (see, for instance, Fusar-Poli et al., Reference Fusar-Poli, Borgwardt, Bechdolf, Addington, Riecher-Rössler, Schultze-Lutter, Keshavan, Wood, Ruhrmann and Seidman2013; McGorry et al., Reference McGorry, Hartmann, Spooner and Nelson2018). However, we are still unable to identify which individual patients are most likely to progress to full-threshold psychosis. While this is inherently a complex task involving several methodological challenges, part of the problem may be ascribed to the oversimplified nature of current psychopathological descriptions incorporated into many of the rating scales used to measure psychopathology. The use of yes/no self-report instruments in research studies seems to be particularly detrimental for the identification or delineation of ‘caseness’. For instance, Nordgaard et al. (Reference Nordgaard, Buch Pedersen, Hastrup, Haahr and Simonsen2019) investigated the validity of self-rated questionnaires for ‘psychosis-like’ symptoms in the general population. They found that the use of self-rating scales resulted in 82.5% of the cases being false positives when re-tested against a semi-structured interview conducted by staff trained in psychopathology. Phenomenology has been suggested as a useful corrective to these research trends, by way of adding depth, richness and nuance to standard clinical data (Nordgaard et al., Reference Nordgaard, Sass and Parnas2013; Nelson et al., Reference Nelson, Hartmann and Parnas2018).

Nelson et al. (Reference Nelson, McGorry and Fernandez2021) have suggested a way to take this forward. For instance, in psychosis research, phenomenology could be integrated with the clinical staging approach to add depth and nuance to stage-based clinical phenotypes. Importantly, this approach promotes a multi-layered understanding of the unique (i.e., idiographic) as well as shared (i.e., nomothetic) features of the experience of mental disorders. The integration of phenomenological insights could also open new research paths for clinical studies of delusions: here, it can help capture widely neglected areas of mental and experiential life beyond simple clinical severity, without lumping together forms of delusions that may only be loosely linked (Ritunnano et al., Reference Ritunnano, Broome and Stanghellini2021). When combined with standard clinical data, such as symptom severity or clinical stage, the integration of a phenomenologically informed framework allows us to increase the granular resolution of the psychopathological phenotype, thus contributing to improved, more accurate identification of caseness.

In this way, researchers may be able to better demarcate the diagnostic, prognostic and therapeutic subgroups in a way that is relevant, for instance, for the translation of findings from clinical trials to aid decision-making in clinical practice. In fact, when evaluating a patient's complaints and choosing treatments, the clinicians may not only consider the diagnosis or the severity of symptoms, but also their experiential quality, the meanings they bear for the person experiencing them, the social and cultural context in which they are embedded, the interactional dynamics that shape them and their consequences for the person's sense of identity.

In the context of data analysis, phenomenological variables may also provide potentially useful information for moderation, mediation or path analyses by foregrounding previously unacknowledged experience-based variables with a significant effect on illness onset or recovery processes. They may also inform the iterative development and validation of new tools and measures grounded in the lived experience of the person. More accurate measures, and thresholds for caseness, informed by phenomenology may eventually improve our ability to diagnose, treat and potentially prevent serious mental disorders.

Importantly, phenomenology should not necessarily be constrained by existing taxonomies, but can aid scientific openness and the discovery of new knowledge by virtue of its rejection of strong theoretical assumptions, including that of our current classifications. This may be relevant for instances where psychopathological phenomena show an underappreciated transdiagnostic potential as an investigational and therapeutic target. For the case of mood instability, see Broome et al. (Reference Broome, Saunders, Harrison and Marwaha2015).

Integrating phenomenological methods: promises and challenges

Phenomenology offers sound empirical methods for exploring and describing the patient's subjectivity (Box 1). The use of these methods is not, however, without challenges. Phenomenological practice has often been accused of requiring too much in-depth training, or of being too time consuming for it to be effectively embedded in mainstream psychiatric research – and therefore being unable to deliver on its promise. There is no denying that phenomenological interviews are lengthy processes, taking up a great deal of resources both in terms of training researchers, and conducting the necessary fieldwork. However, there is also no denying the fact that massive financial investments have been made in the past to support costly genetic testing and functional brain imaging studies, with relatively minimal or modest gains in terms of patient benefit.

We believe that the time has come to reflect on the assumptions and guiding principles that shape editorial and funding policies in mental health research. The ‘hard’ kind of scientific evidence, supposedly delivered by neuro-centric and bio-oriented research, may well seem reassuringly objective – with its allure of certainty and its promise of unshakeable empirical foundations – but does it deliver valuable, actionable information when it comes to understanding troubled human existence? There is an unjustified optimism in the faith that a narrow biomedical conception of mental ill health will deliver improved outcomes, echoing the criticisms of the ‘neuromythology’ of late 19th century German psychiatry made by Jaspers and his contemporaries. Is this approach as ‘neutral’ or ‘objective’ as it purports to be? Does it provide us with useful, effective tools to make sense of mental suffering? Does it challenge the forms of epistemic injustice that affect many people with mental disabilities? Does it provide psychiatry with the tools required to deliver improved care? As Bilsbury and Richman note, ‘a quest for statistical psychometric virtue is futile if the instrument is so ill-focused that it is irrelevant to the individual’ (Bilsbury and Richman, Reference Bilsbury and Richman2002, p. 10).

Ultimately, to expect that quantitative, bio-psychiatric research alone is going to lead to better mental healthcare for all is probably unrealistic: we are currently lacking in strong evidence for such optimism. Joining forces may be a better way forward. But how? Echoing Jaspers once again, we believe that psychiatry should be concerned with the ‘human being as a whole’ (ibid., p. 1) as its main object of investigation, including the environmental and social contexts in which altered experiences may occur (Pienkos, Reference Pienkos2020).

Various phenomenologically informed methodologies, and forms of phenomenological interviewing, have been developed and used worldwide across qualitative and quantitative research designs. For instance, in qualitative research, Interpretative Phenomenological Analysis (IPA) is a widely used approach, informed by phenomenology, hermeneutics and idiography, committed to the investigation of how people experience and make sense of major life experiences (Smith et al., Reference Smith, Flowers and Larkin2022). In quantitative designs, the use of the Examination of Anomalous Self-Experience (EASE) scale (Parnas et al., Reference Parnas, Møller, Kircher, Thalbitzer, Jansson, Handest and Zahavi2005) is a good example of how a phenomenologically informed approach can enrich translational research in psychiatry. The EASE is a semi-structured phenomenologically informed psychometric instrument, providing both qualitative and quantitative data on subjective anomalies that may indicate a disorder of self-awareness or self-disturbance (Box 1) – see also Nelson et al. (Reference Nelson, Parnas and Sass2014) for a clarification of the concept. The EASE has been used in empirical studies to explore both psychotic and non-psychotic self-disorders, and their association with clinical variables and diagnostic outcomes. Notably, a recent systematic review of 53 empirical studies using the EASE scale by Henriksen et al. (Reference Henriksen, Raballo and Nordgaard2021) supports the notion that self-disorders hyper-aggregate in schizophrenia spectrum disorders, but are less prevalent in other mental disorders or healthy controls. The results also show that self-disorders are far more prominent in first-episode psychosis and ultra-high-risk (UHR) groups compared to non-psychotic and health controls, and that they are a strong independent predictor of future schizophrenia onset in UHR patients (Nelson et al., Reference Nelson, Thompson and Yung2012), non-psychotic adults (Parnas et al., Reference Parnas, Raballo, Handest, Jansson, Vollmer-Larsen and Saebye2011) and youth clinical populations (Koren et al., Reference Koren, Tzivoni, Schalit, Adres, Reznik, Apter and Parnas2020).

While larger observational studies are still ongoing (e.g., Reference Krcmar, Wannan, Lavoie, Allott, Davey, Yuen, Whitford, Formica, Youn, Shetty, Beedham, Rayner, Polari, Gawęda, Koren, Sass, Parnas, Rasmussen, McGorry, Hartmann and NelsonKrcmar et al., in preparation), this knowledge holds promise as a powerful diagnostic and predictive tool in clinical settings. It is also particularly valuable to research investigating the pathogenic mechanisms of onset of schizophrenia and related disorders. In this context, for instance, phenomenological data on self-disorders are being used alongside neurocognitive and neurophysiological measures (e.g., source monitoring deficits and aberrant salience) with the aim of developing more accurate predictive models for the identification of UHR patients who are most likely to progress to full-threshold psychosis (Nelson et al., Reference Nelson, Lavoie, Gaweda, Li, Sass, Koren, McGorry, Jack, Parnas, Polari, Allott, Hartmann and Whitford2019). If validated, such models could be translated into tools for use in clinical practice to inform diagnostic, prognostic and treatment decision-making. It is notable that despite the phenomenological knowledge accumulated in this area, the effect of pharmacological or psychotherapeutic interventions on self-disorders has not yet been investigated. Notwithstanding the high levels of distress, often reported by patients with psychosis, in relation to alterations in the sense of self and identity (Griffiths et al., Reference Griffiths, Mansell, Edge and Tai2019; Bögle and Boden, Reference Bögle and Boden2022), the specific treatment of self-disorders remains, to our knowledge, unexplored.

Identifying meaningful outcomes and new targets for psychological treatment

Identifying and selecting the appropriate outcome variables to assess healthcare interventions and services is one of the biggest challenges faced by researchers and providers today. Mental states are complex, fluctuating, strongly individualised experiences that often resist the kind of quantitative measurement pursued by standardised rating scales, and it is fortunate that many studies have now moved away from cross-sectional symptom reduction as a primary or sole outcome. Similarly, we know that recovery is a deeply personal and unique process, which goes beyond a simple reduction in symptom severity as captured by a numerical score. Key dimensions of recovery in mental health include, for example, ‘connectedness, hope and optimism about the future, identity, meaning in life, and empowerment’ (CHIME) (Leamy et al., Reference Leamy, Bird, Boutillier, Williams and Slade2011). For these and related reasons, there has been growing interest in the development of patient-centred approaches to assessing treatment outcomes (Thornicroft and Slade, Reference Thornicroft and Slade2014), and many calls to action have been made to build and deliver patient-centred care in collaboration with patients (Santana et al., Reference Santana, Ahmed, Lorenzetti, Jolley, Manalili, Zelinsky, Quan and Lu2019; Schroeder et al., Reference Schroeder, Bertelsen, Scott, Deane, Dormer, Nair, Elliott, Krug, Sargeant, Chapman and Brooke2022). To this end, patient-reported outcome measures (PROMs), measuring patients' perspectives on health outcomes, are increasingly used in health care. However, the extent to which these measures are developed through a meaningful and systematic engagement with patients and lived experience researchers has been questioned (Trujols et al., Reference Trujols, Portella, Iraurgi, Campins, Siñol and Cobos2013; Wiering, de Boer and Delnoij, Reference Wiering, de Boer and Delnoij2017).

With its focus on patients' subjectivity and narratives, there is enormous potential for phenomenological knowledge and methods to be used to develop patient-focused healthcare systems and outcomes that are better tailored to, and centred around, patient experience. Indeed, phenomenology is by no means restricted to the description of psychopathological symptoms (Fuchs et al., Reference Fuchs, Messas and Stanghellini2019). Insights from phenomenological studies can inform, for instance, the development of novel targets for treatment and care strategies, particularly in the field of psychotherapy (Nelson and Sass, Reference Nelson and Sass2009; Pérez-Álvarez et al., Reference Pérez-Álvarez, García-Montes, Vallina-Fernández, Perona-Garcelán and Cuevas-Yust2011; Škodlar and Henriksen, Reference Škodlar and Henriksen2019). Phenomenological knowledge and concepts can also help identify, refine or develop new PROMs that are based upon and truly incorporate the patient's experiences and perspectives. This is especially important in the psychotherapy of schizophrenia and other psychoses, where the effectiveness of currently available, evidence-based treatments such as CBT has been repeatedly found to be sub-optimal against several standard outcomes (Jones et al., Reference Jones, Hacker, Cormac, Meaden and Irving2012, Reference Jones, Hacker, Meaden, Cormac, Irving, Xia, Zhao, Shi and Chen2018; Bighelli et al., Reference Bighelli, Salanti, Huhn, Schneider-Thoma, Krause, Reitmeir, Wallis, Schwermann, Pitschel-Walz, Barbui, Furukawa and Leucht2018; Jauhar et al., Reference Jauhar, Laws and McKenna2019).

For instance, the anomalies of self-awareness described above as core clinical and vulnerability features of schizophrenia could be a potent experience-based target of psychological treatment – as they are purported to underly and generate a wide range of the disorder's more obvious symptoms and signs (such as positive and negative symptomatology). This approach should begin from the patients' subjective experiences (and not the researcher's third-person interpretations of the patient's behaviour or utterances) as the starting point for developing patient-centred interventions, and for identifying patient-focused outcomes. With the appropriate training and investment, the EASE and other phenomenologically informed instruments (Box 1) could be considered when selecting outcome variables for the evaluation of treatment in psychosis research. However, it is crucial to keep in mind that ‘phenomenologically informed’ does not always imply ‘patient-valued’ as there is always a risk that phenomenological measures, although based on experiential accounts, prioritise the values and concerns of clinicians and researchers over those of patients.

For this reason, a mixed-methods approach that integrates qualitative and participatory research techniques in study designs may be a better way forward to identify patient-centred outcome domains, develop patient-valued measures and select new treatment targets. A recent example of this is Sheaves et al. (Reference Sheaves, Johns, Loe, Bold, Černis, Molodynski and Freeman2022)'s work using lived experience accounts to build a novel theoretical framework and developing two new measures of voice-related distress (Sheaves et al., Reference Sheaves, Johns, Loe, Bold, Černis, Molodynski and Freeman2022). In this study, qualitative interviews with people experiencing derogatory and threatening voices formed the basis for the generation of two psychometrically robust assessments, providing a new perspective on voice distress. This experience-based, patient-generated framework can then be translated into patient-valued targets for psychological intervention. Similarly, phenomenological insights gained from co-written bottom-up reviews of the lived experience of psychosis (Fusar-Poli et al., Reference Fusar-Poli, Estradé, Stanghellini, Venables, Onwumere, Messas, Gilardi, Nelson, Patel, Bonoldi and Aragona2022) or from systematic reviews and qualitative meta-syntheses (Ritunnano et al., Reference Ritunnano, Kleinman, Oshodi, Michail, Nelson, Humpston and Broome2022b) could help create measures and develop treatments that are more faithful to the first-person perspective. Without this approach, the risk is that we continue to rely on outdated, researcher-generated constructs that may or may not reflect the real nature of the phenomena under investigation, and may or may not matter to patients.