ReferenceJournal of Family Practice 1997; 44: 169–176
Authors and institutionDavid Thom, Division of Family and Community Medicine, Department of Medicine; Bruce Campbell, Department of Health, Research and Policy Stanford University School of Medicine, California, US
BackgroundThe authors thought that trust was important in patient–doctor relationships, but that this might be affected by healthcare plans: patients may be restricted in their choice of physician, or discontinuity of physician may occur
Setting‘San Francisco Bay Area’, US
AimsTo gain an understanding of how patients perceive trust of a physician and how patients relate physicians' behaviours to their perceptions of trust
Research designFour focus groups. Working definition of trust: ‘the patient's confidence that the physician will do what is best for the patient’
SamplingMixed recruitment strategy. Groups one and two: patients from a university-based family practice, recruited from a list of 12 long-time patients generated by the two senior physicians in the practice and 10 recruited from a list of 43 randomly sampled patients who had visited the office within the previous 6 months. Group three: recruited from a random sample of 54 English-speaking Hispanic patients who had recently (within 56 months) visited a family practice residency clinic in San Jose. Group four: recruited by flyers posted in a publicly supported medical clinic in a lower income area. Participants in groups three and four were paid US$20
Data collectionFocus groups were conducted at each clinic site, lasted 1.5–2 hours, and were led by BC, a sociologist who was ‘experienced in focus group research using principles of qualitative research’. An observer was present at each session to take notes regarding the mood, non-verbal communication, and general impressions. Each group was audiorecorded and transcribed. Participants were asked to describe situations they had experienced that led them to trust a physician, and situations that had caused them to lose, or not to establish, trust
ReflexivityNot discussed
Ethical issuesNot discussed
Data analysisTranscripts were independently coded, ‘using techniques of grounded theory’: patient statements were labelled by four independent readers (a physician, a sociologist, and a research assistant, plus either a nurse researcher or a second physician) and attached to the text using Ethnograph. Labelled statements (‘open codes’) were grouped into conceptual categories (‘axial codes’) by consensus over several meetings. ‘The process was repeated for each subsequent focus group, and the categories (axial codes) were modified to incorporate new types of statements. Thus, the final categories included the reported experiences of all participants in all four groups.’ Analysis until ‘saturation’ was not done because of ‘limited resources’