Development of Taxonomy
Two study team members (an internist and a physician researcher) independently reviewed each transcript in its entirety, identifying distinct topics (themes) and making comments indicating each of these units of text. Repeated or reworded statements of the same idea by the same participant were listed together as one comment.
Each unit of text (a statement that conveyed one idea) from the transcripts was listed by a physician primary care researcher in the order it arose in the transcripts as both a direct quote and as a summary theme on the basis of the comments made by the first 2 study team members. Initially, to avoid imposing any particular framework onto the women’s comments, 2 investigators did independent inductive coding,13-14 in which each unit of text was reviewed in its context from the transcript, categories (labels) were generated, and a list of labels was compiled. When reviewing this exhaustive list, we found that the list of inductive labels (codes) fit fairly well into established conceptual frameworks for primary care. Thus, all units of text from the transcripts were then reclassified independently in duplicate (by a clinical internist and by a physician primary care researcher), using agreed-upon coding rules from the primary care conceptual framework, with the addition of the physician-patient relationship category, which arose as a common theme from the transcripts.
Interrater reliability for the overall coding of distinct units of text into 1 of 6 major primary care content areas was substantial (b = 0.84 overall). Content analysis was performed on the comments for all 4 focus groups, including a count of the number of times a theme was mentioned by different respondents and the primary care content area into which the themes fit.
Results
A total of 24 women participated in the discussions: 8 Latinas, 15 African Americans, and 1 white woman. The mean age of the participants was 46.6 years (median = 44.5; one third were aged 50 years and older.) Eight of the participants had an 11th grade education or less; 5 were high-school graduates; and 11 had some college education. Four were married. The majority worked: 8 full time, 8 part time, and the rest were unpaid, retired, or unemployed. Sixteen of these women cared for dependents part or full time. Eighty-two percent of the participants had a household income of less than $20,000, reflecting our success in recruiting the population we sought. Twenty-two women were uninsured, but most of the African American participants had had Medicaid or private insurance in the past.
The most important conceptual modification arising from the women’s comments was the addition of the physician-patient relationship as an important and unique feature encompassing many of the women’s priorities. The percentages of focus group participant comments falling into each of the major primary care codes were as follows: an accessible source of care (37.4%), the physician-patient relationship (37.4%), a comprehensive range of services (11.5%), coordination across providers (6.8%), continuity with a single provider (3.7%), and accountability (3.2%). Table 2 gives the frequency distribution of participants’ priorities for primary care and some of the more commonly stated priorities.
Within the content area of the physician-patient relationship, themes mentioned most often were communication between physician and patient, having staff who listen, getting personal attention, and most important, a staff that was concerned and respectful. For Latinas, clinicians’ knowledge of the Latin community and of the fear and trust issues experienced by recent immigrants toward the medical system and toward other members of the community were mentioned often.
Specific attributes mentioned frequently within the category of accessibility were a clinic that had evening and weekend hours, was open to all regardless of insurance status, was located in the inner city or was accessible by using public transport, and was attentive to waiting times. Among Latinas, having a doctor fluent in Spanish and from a similar cultural background was an additional priority.
Within the category of comprehensiveness, the most frequently mentioned themes were the availability of multiple services at one site, presence of an intake procedure that recognized one’s needs, coordination of medical and social services on-site, and the availability of counseling and treatment for emotional and mental health concerns. Sample quotes from the focus group transcripts, organized within the 6 content areas, are presented in Table 3
Discussion
Eighty-six percent of participants’ comments fit into 1 of 3 content areas: physician-patient relationship, accessibility, and comprehensiveness. The breadth and depth9 of physician-patient interactions in primary care make its relationship unique. Heavy emphasis on interactions with their primary care physicians (one third of all comments) supports other authors’ statements about vulnerable patients placing a special emphasis on this relationship.15-16 Underinsured people lacking access to alternate providers have a heightened reliance on a physician’s competence, skills, and good will.15 Having a sense that their physician had concern and respect for the patient was the most frequently mentioned priority in the focus groups. When working with low-income minority or immigrant patients, physicians might want to be especially sensitive to their voice, tone, and posture to communicate a sense of respect and concern for patients who may already feel vulnerable. It appears that the category of physician-patient relationship is vital to the conceptual framework of primary care for these low-income women, and it may be a link in the chain without which the other features (continuity, comprehensiveness, coordination, accessibility, accountability) cannot function optimally.