Blood tests in primary care: A qualitative study of communication and decision‐making between doctors and patients

Abstract Objective Blood tests are commonly used in primary care as a tool to aid diagnosis, and to offer reassurance and validation for patients. If doctors and patients do not have a shared understanding of the reasons for testing and the meaning of results, these aims may not be fulfilled. Shared decision‐making is widely advocated; yet, most research focusses on treatment decisions rather than diagnostic decisions. The aim of this study was to explore communication and decision‐making around diagnostic blood tests in primary care. Methods Qualitative interviews were undertaken with patients and clinicians in UK primary care. Patients were interviewed at the time of blood testing, with a follow‐up interview after they received test results. Interviews with clinicians who requested the tests provided paired data to compare clinicians' and patients' expectations, experiences and understandings of tests. Interviews were analysed thematically using inductive and deductive coding. Results A total of 80 interviews with 28 patients and 19 doctors were completed. We identified a mismatch in expectations and understanding of tests, which led to downstream consequences including frustration, anxiety and uncertainty for patients. There was no evidence of shared decision‐making in consultations preceding the decision to test. Doctors adopted a paternalistic approach, believing that they were protecting patients from anxiety. Conclusion Patients were not able to develop informed preferences and did not perceive that choice is possible in decisions about testing, because they did not have sufficient information and a shared understanding of tests. A lack of shared understanding at the point of decision‐making led to downstream consequences when test results did not fulfil patients' expectations. Although shared decision‐making is recommended as best practice, it does not reflect the reality of doctors' and patients' accounts of testing; a broader model of shared understanding seems to be more relevant to the complexity of primary care diagnosis. Patient or Public Contribution A patient and public involvement group comprising five participants with lived experience of blood testing in primary care met regularly during the study. They contributed to the development of the research objectives, planning recruitment methods, reviewing patient information leaflets and topic guides and also contributed to discussion of emerging themes at an early stage in the analysis process.

consequences when test results did not fulfil patients' expectations. Although shared decision-making is recommended as best practice, it does not reflect the reality of doctors' and patients' accounts of testing; a broader model of shared understanding seems to be more relevant to the complexity of primary care diagnosis.
Patient or Public Contribution: A patient and public involvement group comprising five participants with lived experience of blood testing in primary care met regularly during the study. They contributed to the development of the research objectives, planning recruitment methods, reviewing patient information leaflets and topic guides and also contributed to discussion of emerging themes at an early stage in the analysis process. Qualitative studies with doctors and patients have shown that blood tests also fulfil psychosocial roles in the consultation, for example, providing reassurance, helping to build an empathic connection and providing validation for patients' symptoms. [2][3][4] Inflammatory markers are a type of blood test commonly used for the diagnosis and monitoring of infections, autoimmune conditions and cancers, 5 and as a nonspecific screening test for patients with unexplained symptoms. 6 Inflammatory markers rarely offer a specific diagnosis, but can offer 'clues'. Falsepositive inflammatory markers can lead to increased rates of follow-on consultations, blood tests and referrals. 7 It is therefore important that doctors and patients have a shared understanding of the reasons for inflammatory marker testing and the potential pitfalls of testing.
Shared decision-making is a process in which patients and clinicians work together to make decisions based on evidence and patients' preferences, and is widely recommended as 'best practice'. 8 Awareness of the importance of shared decision-making is widespread; yet, most research focusses on treatment decisions rather than diagnostic testing. Where research into diagnostic testing does exist, it mostly focusses on specific tests, for example, prostatespecific antigen for prostate cancer. 9 The aim of this study was to examine communication and decision-making around inflammatory marker blood tests in primary care. Although we examined testing decisions through a lens of shared decision-making, we chose not to limit our analysis to shared decision-making, as we wished to consider the process of communication from the initial consultation to test results. Participants were informed that the interviewer was a GP; it was emphasized that the interviews were nonjudgemental, and were focussed on exploring communication around testing, not on scrutinizing the clinical decision-making.

| METHODS
After patient recruitment, the GP who had requested the blood tests was contacted to arrange a telephone interview. Each GP could complete a maximum of two interviews (about different patients), to maximize the range of GPs.

| Interviews
Interviews were carried out by J. W., a practising GP with experience and training in qualitative research methodology. Interviews were semistructured, using topic guides based on the research questions, but flexible enough to allow exploration of issues raised by the participant.
The topic guide (see the Supporting Information Material) was adapted iteratively during the study, using information emerging in early interviews to inform subsequent interviews. The initial patient interviews Interviews were continued until a diverse sample had been recruited and data saturation was achieved, meaning that the topic guide was stable, with no new codes arising. 10

| Analysis
Audio recordings were transcribed verbatim by an experienced transcriber. Analysis began when the first transcripts were available, so that data collection and analysis were conducted concurrently.
Transcripts were analysed using thematic analysis, involving a mixture of inductive and deductive coding and constant comparison. 11 Two members of the research team (J. W. and J. B.) independently reviewed four transcripts to develop an initial coding framework that reflected the research objectives. This framework was adapted following discussions with the study team and tested on a further three transcripts by J. W. and J. B. J. W. then took responsibility for ongoing coding and categorization of the data, using NVivo qualitative data management software. Categories of data and thematic relationships were identified and written up as descriptive and interpretive accounts.   experience) The perception that testing is 'still one of those areas' seems to reflect an awareness amongst GPs of the changes over time in medical decision-making, and the gradual move away from paternalism. Some doctors reflected on this during the interviews and felt that this was an area where they could try to improve their practice.
I'm probably not as good as I might be at sharing that decision, so it's probably more mine actually but I think There was a perception among both doctors and patients that sharing information about testing was the right thing to do 'in an ideal world', but multiple barriers to information sharing were identified. Both doctors and patients felt that blood tests were complex and too technical for patients to fully understand. There was a lack of shared language and a lack of resources for sharing information about tests, which meant that even when doctors tried to share information, patients did not always understand or retain information about their blood tests. In the overall context of the consultation, tests were perceived by GPs as relatively trivial interventions and were therefore low priority for information sharing; time and workload pressures added further barriers to a shared understanding of tests.