ABSTRACT
BACKGROUND: Diagnostic uncertainty is common, but its communication to patients is under-explored. This study aimed to (1) characterise variation in doctors' communication of diagnostic uncertainty and (2) explore why variation occurred. METHODS: Four written vignettes of clinical scenarios involving diagnostic uncertainty were developed. Doctors were recruited from five hospitals until theoretical saturation was reached (n = 36). Participants read vignettes in a randomised order, and were asked to discuss the diagnosis/plan with an online interviewer, as they would with a 'typical patient'. Semi-structured interviews explored reasons for communication choices. Interview transcripts were coded; quantitative and qualitative (thematic) analyses were undertaken. RESULTS: There was marked variation in doctors' communication: in their discussion about differential diagnoses, their reference to the level of uncertainty in diagnoses/investigations and their acknowledgement of diagnostic uncertainty when safety-netting. Implicit expressions of uncertainty were more common than explicit. Participants expressed both different communication goals (including reducing patient anxiety, building trust, empowering patients and protecting against diagnostic errors) and different perspectives on how to achieve these goals. Training in diagnostic uncertainty communication is rare, but many felt it would be useful. CONCLUSIONS: Significant variation in diagnostic uncertainty communication exists, even in a controlled setting. Differing communication goals-often grounded in conflicting ethical principles, for example, respect for autonomy versus nonmaleficence-and differing ideas on how to prioritise and achieve them may underlie this. The variation in communication behaviours observed has important implications for patient safety and health inequalities. Patient-focused research is required to guide practice. PATIENT OR PUBLIC CONTRIBUTION: In the design stage of the study, two patient and public involvement groups (consisting of members of the public of a range of ages and backgrounds) were consulted to gain an understanding of patient perspectives on the concept of communicating diagnostic uncertainty. Their feedback informed the formulations of the research questions and the choice of vignettes used.
Subject(s)
Communication , Physician-Patient Relations , Physicians , Humans , Uncertainty , Male , Female , Physicians/psychology , Adult , Middle Aged , Interviews as Topic , Diagnosis, Differential , Qualitative ResearchABSTRACT
BACKGROUND: Studying clinician-patient communication can be challenging, particularly when research seeks to explore cause-and-effect relationships. Video vignettes - hypothetical yet realistic scenarios - offer advantages to traditional observational approaches by enabling standardisation and manipulation of a clinician-patient encounter for assessment by participants. While published guidelines outline stages to create valid video vignette studies, constructing high quality vignettes which are accessible to a wide range of participants and feasible to produce within time and budget restraints remains challenging. Here, we outline our methods in creating valid video vignettes to study the communication of diagnostic uncertainty. We aim to provide practically useful recommendations for future researchers, and to prompt further reflection on accessibility issues in video vignette methodology. METHODS: We produced four video vignettes for use in an online study examining the communication of diagnostic uncertainty. We followed established guidelines for vignette production, with specific consideration of how these might be applied pragmatically to save time and resources. Scripts were pilot-tested with 15 laypeople, and videos with 14 laypeople; pilot-testing involved both quantitative and qualitative analysis. RESULTS AND DISCUSSION: We demonstrate the usefulness of existing guidelines, while also determining that vignette production need not necessarily be expensive or time-consuming to be valid. Our vignettes were filmed using an iPhone camera, and featured a physician rather than a professional actor; nonetheless, pilot-testing found them to be internally and externally valid for experimental use. We thus propose that if care is taken in initial script development and if pragmatic choices are made regarding filming techniques and pilot-testing, researchers can produce valid vignettes within reasonable time and budget restraints. We also suggest that existing research fails to critically examine the potential benefits and harms of online video vignette methodology, and propose that further research should consider how it can be adapted to be inclusive of those from underserved backgrounds. CONCLUSIONS: Researchers creating video vignette studies can adapt the video vignette development process to suit time and budget constraints, and to make best use of available technology. Online methods may be harnessed to increase participant accessibility, but future research should explore more inclusive vignette design.
