2.1 Study Design

We conducted a qualitative study of primary and speciality care clinicians, their patients, and caregivers of patients (when present) to identify what information from a clinic visit was important and why. This study received IRB approval from the Committee for the Protection of Human Subjects (CPHS) at Dartmouth College (STUDY# 29708). This study is reported in accordance with the ‘Consolidated Criteria for Reporting Qualitative Research’ (COREQ) guideline [14].

2.2 Setting & Participants

We recruited clinicians, patients, and caregivers from four specialty clinics (orthopaedics, cardiology, radiation oncology, and gynaecologic oncology) within a large academic medical centre and an affiliated outpatient primary care clinic in the Upper Valley of New Hampshire between November 2016 and June 2018. Participants had to be at least 18 years of age and able to communicate in English. Participants who could not provide informed consent or had hearing impairment were excluded from the study. Convenience samples of clinicians were recruited via email. Recruited clinicians were asked to identify patients who would be receiving a new health-related diagnosis or making a medical decision during their visit (i.e., new treatment, screening decision), rather than those coming in strictly for a procedural visit (i.e., blood draw or immunization) where little information is exchanged. If caregivers were present and the patient agreed to their participation, they were invited to complete informed consent and participate in the project. We aimed to recruit a minimum of six to eight patients from both primary care and specialty care, reflecting a number of interviews considered adequate to reach data saturation [15].

2.3 Study Procedures

Patients received a portable digital audio recorder and were instructed to begin recording at the start of their clinic visit and stop the recording upon the visit conclusion. Immediately after the clinic visit, patients were brought into a separate clinic room by a session facilitator. Patients were instructed to listen back to the clinic visit and identify (1) important information that could include something they may want to recall later or want a family member to hear, and (2) why the segment was important to them. Patients used a secure application developed by researchers at Dartmouth College to play back the recording and identify the parts of the visit that were most important. To highlight important information, patients would press and hold the spacebar, releasing the spacebar when the important information concluded (Figure 1). Patients' responses to why the segment was important were written down verbatim by the facilitator. Caregivers completed the same study procedures independently. Both received $20 for their participation. Due to time constraints, audio recordings from clinic visits were transcribed verbatim by Acusis, a HIPAA-compliant medical transcription company, and shared with clinicians. Clinicians were instructed (1) to mark on the transcript information from the visit they, as clinicians felt, was most important and what they would want the patient to remember, and (2) to provide their rationale for each section of the text identified. Clinicians were not paid for their participation in the study.

2.4 Data Collection

We collected patient and caregiver demographics (age, gender, race, ethnicity, education), recordings of clinic visits, written and verbal responses on what information from the recordings or transcripts was important, and reasons information was deemed important.

2.5 Data Analysis

We conducted a summative content analysis, counting and comparing the utterances considered most valuable by stakeholders, to identify what themes of information from the clinic visit were important [16]. Our approach to coding was informed by findings from our case study of clinics where audio/video recordings of clinic visits are routinely shared with patients and caregivers; in the case study, interviewed stakeholders stated what information was most valuable from recordings, but they did not review and identify information [17]. We also conducted a directed content analysis to identify themes for why the information was identified as important. Two coders independently applied an initial codebook to a subset of the clinician rationale data. The two coders then met to create a revised codebook, which was applied to the existing codes. Emergent codes were added to existing codes. This process was repeated for the patient and caregiver rationale data. The research team decided to create separate codebooks for each stakeholder group to capture nuanced differences in the codes that emerged between groups. Codes were reviewed to create qualitative memos which allowed the researchers to interpret the data and identify emerging themes. Memos were reviewed to search for themes across stakeholders. Potential themes were reviewed and refined with a third reviewer. To increase trustworthiness and credibility, qualitative study results were reviewed independently by a methods expert who was not part of the core research team (KS). Statements made exclusively by the clinician were classified as the clinician speaking, statements made exclusively by the patient were classified as the patient speaking, and statements where both the patient and clinician spoke/contributed to the same topic (i.e., dialogue) were classified as patient and clinician speaking. All quotations presented in the results section have been deidentified.

3.1 Participant Characteristics

Seventeen unique clinic visits, including eight primary care and nine specialty care visits, were recorded and included in the analysis (Table 1). Six clinicians reviewed a median of three visit recordings, as they were the treating clinicians for more than one patient in the sample. A total of 16 patients and 4 caregivers reviewed their visit recordings. One primary care clinician was unable to review three of their visit recordings due to illness and one patient did not review their visit recording. Patients had a median age of 68 years (25–84 years), more than half were female (56%) and more than half (63%) possessed a bachelor's degree or higher (Table 2). Caregivers had a median age of 57 years (46–78 years), were predominantly female (75%), and had a graduate degree or higher (75%). Four out of the six clinicians were male, and all had been in practice for at least 10 years, with three having practised for over 25 years. Clinician specialties included family medicine (n = 2), interventional cardiology, gynaecologic oncology, radiation oncology, and orthopaedic surgery.

The average length of visits was 33:43 min [8:10–61:05]. The median number of total words exchanged during a visit was 5271.5 [1466–8971], of which 37% were spoken by patients [7%–60%], 59% by clinicians [40%–92%], and 4% by others [1%–12%].

3.2 Information Identified As Important

A total of 440 statements were identified as important by stakeholders across the visits (Table 3). Clinicians identified a median of 12.5 statements [range 6–50 statements] while patients identified a median of 7.5 statements [range 3–20 statements], a ratio of 1.6 statements identified by clinicians for every 1 statement identified by patients. Caregivers identified a median of 18 statements [range 4–31 statements] per visit reviewed. Overall, information related to patient education was most frequently identified as important (38.6%; 170 statements). This included background information on a medical condition and explaining how a medical procedure works (see Table 3). Education was followed by symptoms and problems (18%; 78 statements), the clinician's recommendation (18%; 77 statements), and the discussion of medications (15%; 64 statements). When doctors identified information related to medications as important, it tended to focus on side effects, whereas patients also identified information about medication benefits as important.

Clinicians overwhelmingly selected their own statements as important (85.5% of statements). Caregivers also identified primarily information conveyed by the clinician as important (92.3% of statements). Patients were equally likely to identify the information they shared or their dialogue with the clinician as important (48.6% of statements), compared to the clinicians' statements alone (51.4% of statements). None of the statements that were identified as important was spoken by a caregiver (Table 4).

3.3 Thematic Analysis of Why Information Is Important

Four main themes emerged for why the information was identified as important: knowledge transfer on key information, the patient experience, the care plan and the provision of support/reassurance by the clinician.

4.1 Discussion

4.2 Practice Implications

In this exploratory study, both clinicians and caregivers placed more value on clinician statements, compared to patients who also identified their own contributions as of high value. The implications for how we communicate with patients are important—for example, an after-visit summary could benefit from a more detailed description of what the patient contributed to the clinic visit, their thoughts, concerns and questions. These implications align with views from experts in the field of shared decision-making who report that it is insufficient to focus on ‘dispassionate and objective evidence of medical science’ alone when communicating with patients [25]. It is also necessary to understand the patient's values and what matters most to each patient. By focusing on medically oriented goals, clinicians may inadvertently miss or not provide enough attention to information considered valuable to patients; this can also result in contextualized errors, ‘the failure to adapt research evidence to patient context, when it results in an inappropriate plan of care’ [26]. In a study of 4496 clinic visits at the VA, clinicians addressed such patient contextual factors 67% of the time, which improved to 72% after clinicians were more systematically made aware of these contextual errors, which improved patient outcomes and saved an estimated $25.2 million from avoided hospitalizations [26].

4.3 Conclusion

Our conclusions regarding all stakeholders can only be preliminary, given the limited number of participants, particularly clinicians and caregivers. That being said, all stakeholders in our study found the exchange of information, especially patient education and care planning, to be valuable. There was overlap in what stakeholders identified as important; however, notable differences exist, such as the patients' focus on their own contributions to the visit, which was infrequently cited by clinicians. Additionally, clinicians, as reported elsewhere, dominate what is said. While this is important, more must be done to recognize the importance that patients place in their own words. The patient voice is the raw data that drives medicine; we must value it more to achieve the true coproduction of care.