Evaluating nuanced practices for initiating decision-making in neurology clinics : a mixed-methods study.

Background We report follow-on research from our previous qualitative analysis of how neurologists offer patients choice in practice. This focus reflects the NHS’s emphasis on ‘patient choice’ and the lack of evidence-based guidance on how to enact it. Our primary study identified practices for offering choice, which we called ‘patient view elicitors’ (PVEs) and ‘option-listing’. However, that study was not designed to compare these with recommendations or to analyse the consequences of selecting one practice over another. Objectives To (1) map out (a) the three decision-making practices – recommending, PVEs and option-listing – together with (b) their interactional consequences; (2) identify, qualitatively and quantitatively, interactional patterns across our data set; (3) statistically examine the relationship between interactional practices and self-report data; and (4) use the findings from 1–3 to compare the three practices as methods for initiating decision-making. Design A mixed-methods secondary analysis of recorded neurology consultations and associated questionnaire responses. We coded every recommendation, PVE and option-list together with a range of variables internal (e.g. patients’ responses) and external to the consultation (e.g. self-reported patient satisfaction). The resulting matrix captured the qualitative and quantitative data for every decision. Setting and participants The primary study was conducted in two neurology outpatient centres. A total of 14 neurologists, 223 patients and 114 accompanying others participated. Results Distribution of practices – recommending was the most common approach to decision-making. Patient demographics did not appear to play a key role in patterning decisional practices. Several clinical factors did show associations with practice, including (1) that neurologists were more likely to use option-lists or PVEs when making treatment rather than investigation decisions, (2) they were more certain about a diagnosis and (3) symptoms were medically explained. Consequences of practices – option-lists and PVEs (compared with recommendations) – were strongly associated with choice by neurologists and patients. However, there was no significant difference in overall patient satisfaction relating to practices employed. Recommendations were strongly associated with a course of action being agreed. Decisions containing PVEs were more likely to end in rejection. Option-lists often ended in the decision being deferred. There was no relationship between length of consultation and the practice employed. Limitations A main limitation is that we judged only outcomes based on the recorded consultations and the self-report data collected immediately thereafter. We do not know what happened beyond the consultation. Conclusions Patient choice is harder to enact than policy directives acknowledge. Although there is good evidence that neurologists are seeking to enact patient choice, they are still more likely to make recommendations. This appears to be partly due to concerns that ‘choice’ might conflict with doctors’ duty of care. Future guidance needs to draw on evidence regarding choice in practice to support doctors and patients to achieve the wider goal of shared decision-making. Future research To advance understanding of how interactional practices might have effects beyond the clinic, a priority is to investigate associations between decision-making practices and external outcomes (such as adherence). Funding The National Institute for Health Research Health Services and Delivery Research programme