Formulation: A plea for parsimony

Dear Editor,

Formulation – a set of explanatory hypotheses that answer the question ‘why does this patient suffer from this problem at this time’ – is considered an essential skill of a psychiatrist. The recent paper by Parker et al. ¹ highlights the various approaches to formulation and distils several themes into a comprehensive approach. However, few authors have seriously considered the utility and reliability of psychiatric formulation.

A formulation may be said to possess utility if it provides useful information about prognosis and treatment outcomes or testable hypotheses. ² This has never been demonstrated in a rigorous and prospective study of psychiatrists. Ridley et al. have outlined the manifold limitations of the utility of formulation, including its lack of precision and empirical inadequacies. ³ A formulation may be said to be reliable if common themes are identified by several psychiatrists independently assessing the same patient and blinded to other assessments. The reliability of formulation by psychiatrists has never been empirically demonstrated. A 2015 systematic review found studies of the reliability of formulation were mostly low quality and lacked fidelity to clinical practice. ⁴

Why does all of this matter? Because the stories that psychiatrists tell patients about their illnesses can influence the course of their lives. ⁵ By influencing the perception of the causes of the patient’s current problem or problems, shared psychiatric formulation may aid patient recovery but may also contribute to iatrogenic harm. There has been significant harm perpetuated by our profession, in part through ideological adherence to theoretical models that may have had face validity but have later been shown to be unscientific. ⁶

Where to from here? Formulation should be approached with epistemic humility. While behavioural scientists have developed statistical models that can predict some human behaviour at an aggregate level with fair accuracy, there remains enormous uncertainty in explaining and predicting complex human behaviour at an individual level. In formulating, we should restrict ourselves to what is measurable and testable and utilise psychological theories that have survived replication. If hypotheses are proposed, these should be falsifiable. Finally, psychiatrists should continue to question the scientific foundations of clinical practice. It is advantageous for the profession and our patients to ask why we do what we do in our day-to-day work. The reply that it has always been done that way doesn’t suffice.