Cognitive-behavioural therapy for chronic pain: determining what works for the person in front of you

Meta-analyses of randomised-controlled trials (RCTs) show cognitive-behavioural therapy (CBT) produces small to moderate improvements in functioning and mood for people with chronic pain.¹ This evidence supports the routine inclusion of CBT-based approaches within specialist interdisciplinary pain management programmes. More recently, CBT-based chronic pain pathways have also been implemented within primary care mental health services in England.²

Despite the advantages of RCTs for experimental control, group-level data cannot tell a clinician whether CBT will benefit an individual client sitting in front of them. We presently have inconclusive data upon which to predict who will respond well to CBT for pain.³ Moreover, samples included in RCTs are highly selected. For example, many RCTs of CBT for pain exclude people with complex mental health challenges, such as severe affective disorders, substance abuse, psychosis and interpersonal difficulties. Therefore, clinicians encountering people with these comorbidities have limited data to judge the usefulness of CBT for pain management for these people. To improve the effectiveness of psychotherapeutic approaches for pain, we need to better understand treatment mechanisms.⁴ There has been some progress to this end.^5–7 However, the standard pre-, post-treatment and follow-up assessments of RCTs limit our capacity to investigate mechanisms with precision.

There is a growing resurgence of interest in single-case designs as a rigorous methodological approach to address some of the identified challenges surrounding RCTs.^8–10 In this issue, Jones and Hurrell¹¹ report a single-case AB design consisting of repeated baseline assessments (‘A’ phase) followed by CBT during which repeated assessments were also collected (‘B’ phase) with a participant with chronic pain and depression. Jones and Hurrell collected standardised measures of psychological symptoms, pain catastrophizing and self-efficacy before and after CBT. Crucially, they also collected daily idiographic measures of cognitive and affective treatment targets formulated in collaboration with the participant as key to his presenting problem. The idiographic measures showed the greatest change, while generic psychological symptoms showed the least.

The study by Jones and Hurrell¹¹ is a useful example of the benefits of single-case designs in contrast to RCTs. Single-case designs provide the opportunity to tailor assessment measures to the priorities of the individual participant which may not be adequately captured by standardised measures used in RCTs.^8,11 This collaborative assessment approach is arguably more empowering to participants and is consistent with recommendations for involving people with the condition under study as active research partners.^12,13 The frequency of idiographic assessment enables more precise investigation of the time-course of changes in treatment process and outcome, which can be used to refine treatment.^8,11,14,15 The flexible, person-centred approach of single-case designs may be more appropriate to test the effects of psychotherapy for people with multiple comorbidities.¹⁶ Therefore, single-case methods may be useful for participants traditionally excluded from RCTs of CBT for chronic pain due to the complexity of their mental health or medical needs.

Of course, RCTs are still incredibly important to provide an evidence base upon which decisions about commissioning treatments at the population level are made. However, in the era of personalised medicine, single-case designs may be tools better suited to answer the questions of utmost importance to scientist-practitioners:¹⁷ what works for the person in front of me, in our context, and why?

Whitney Scott
Health Psychology Section, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, UK