Protocol for a Qualitative Study on Doctors’ Opinions on and Experiences of Exceptionally Good Doctors

Plain Language Summary

There is evidence that some doctors are exceptionally good, even if at least in terms of patients’ physical health outcomes, however, there is little research on how such exceptionally good doctors differ from their peers. There is research showing that good doctors deal well with people, are competent, and ethical but almost none on doctors who are exceptionally good. This paper outlines how the authors will try to find out more about what doctors think it takes to make an exceptional doctor and learn more about doctors’ experiences with exceptionally good doctors, through a series of interview. The understanding gained from these interviews may then be used in future to improve medical doctors’ abilities and performance.

Strengths and Limitations of This Study

• One strength of this work is the relevant experience of the authors in recently systematically reviewing over 10,000 studies on doctors’ or surgeons’ effect on patients’ physical health.
• Another strength is that the interviewer is not a clinician and therefore the doctors will not feel the need to take account of the opinion of a colleague during the interview.
• A further strength is that there are no pre-conceived notions on what makes an exceptionally good doctor beyond the opinion that there are such doctors, allowing for a wide range of possible outcomes of this research study.
• A limitation is that this is a qualitative study with a limited number of interviewees on a wide subject, and therefore may not achieve saturation, ie it is possible that many more interviews than the considered 10–15 may be needed.
• A second limitation is that, as all of the subjects will be English-speaking doctors and most of them are expected to be residing in Australia, with perhaps some from North America and Europe, it is uncertain whether the findings could be generalised to other regions, particularly developing nations.
• Finally, due to the heterogeneity regarding medical/surgical specialties, types of interventions, and types of outcomes relevant to different interventions, there may be differing criteria on what makes an exceptionally good doctor for different medical/surgical specialties.

Background

In a classic¹ paper from 1955, Michael Balint² raised the possibility that “the most frequently used drug in general practice was the doctor himself” and over the next seven pages he details how a doctor should “prescribe himself”. At the time of writing, this paper has 86 citations as per Scopus³ and over 4800 citations in google scholar, a discrepancy which points to an impact that is largely outside the traditional scientific literature. On a similar theme, a special issue of The British Medical Journal (BMJ) in 2002 attempted to answer the question, “What’s a good doctor and how do you make one?”.⁴ In one article, letters from individual doctors and others who tried to answer this question were published. Among which, one quote was: “There is not a single piece of evidence or the means to measure whether a doctor is good or bad”.⁵

Previous research has shown that there can be exceptionally poor performing doctors,^6,7 to the point of legal consequences⁶ and historical precedents of malpractice.^6,7 Systematic reviews have evaluated the influence of surgical experience measured in case volume or years of practice, recognizing it as a key factor of medical performance and acknowledging a learning curve.^8,9 Recent evidence also suggests that the provider expectations can substantially affect patient outcomes.¹⁰ While research on doctors’ effect in many non-surgical specialties is rare, there is evidence in primary care,¹¹ obstetrics¹² and acute care¹³ that physician factors had a sizeable effect on health outcomes. Furthermore, there are publications where author-selected outstanding practitioners are investigated.^14–16 “Outstanding” is defined as the mastery of identified skills that would improve patient care and reaffirm medicine’s calling,^14,16 and recognized practice of patient-centred care.¹⁵

The authors of this proposal have recently completed two systematic reviews on existing research of the doctors’ effect on patients’ physical health. Both studies have revealed that there is an unknown doctors’ effect even after controlling for the intervention and all known information such as patient risk factors, hospital and doctor demographics, and experience. It has further been established that there are positive outliers among doctors, ie doctors whose patients’ physical health outcomes are substantially above average. Hence, detecting a significant doctors’ effect also implies that there are doctors who perform better than others. Knowing what identifies exceptional doctors could make it possible for other doctors to learn the same characteristics or practices,¹⁴ and thereby improving patients’ physical health outcomes. If such skills can be passed on to other doctors, then this could be an effective way to improve healthcare.

The current level of knowledge on attributes of good, not necessarily exceptionally good, doctors is also limited, as Abu-Hilal et al’s study illustrates.¹⁷ The authors constructed a “standard questionnaire” to identify attributes considered important in being a good doctor. The attributes identified included clinical ability, to be approachable, appreciate limitations, and knowledge. However, this type of study does not allow a deeper understanding of the phenomenon. A similar but more rigorous qualitative study by Churchill & Schenck identified eight pivotal skills of exceptionally good doctors regarding patient-relationship¹⁴ “do the little things; take time; be open and listen; find something to like; remove barriers; let the patient explain; share authority; and be committed”.

Furthermore, on good doctors, Steiner-Hofbauer et al¹⁸ have systematically reviewed the existing literature regarding the topic of “good doctors” in different stakeholder groups, and they have found 6 studies on doctors of different specialities,^19–23 plus studies gathering the opinions of medical students, patients, the general population, nurses, and children. Steiner-Hofbauer et al¹⁸ have stated that what makes a good doctor is not clearly defined yet, and that the answer to such a question will definitely vary depending on the type of population you ask.²⁴

Others have provided opinion pieces,^25–35 an analysis of publications on stakeholders’ opinions on character strengths of a good doctor,³⁶ patient and physician perspectives,³⁷ and a thesis on the changing perceptions from 1910–2010 on the good doctor in education.³⁸

Currently it is still not known what makes an exceptionally good doctor. This research project aims to address this knowledge gap by conducting a study to identify and describe exceptional doctors. As there is little literature on this subject, an exploratory qualitative study is needed. This approach is commonly used to generate theory when previous knowledge is limited. The selected approach consists of semi-structured interviews and Braun and Clarke’s thematic analysis.^39–41

Materials and Methods

Study Design

The qualitative approach for this study involves semi-structured interviews of medical doctors of any specialty who are available for interviewing. The only other inclusion criterion is that they speak English. Recruitment will be achieved through acquaintances, contacting authors of relevant research papers, contacting medical practices or medical doctors with an internet contact, the GP recruitment program of the Institute of Evidence-based Healthcare of Bond University, and any doctors referred by any of these groups.

Semi-Structured Interviews

Participants will be interviewed in person or via zoom or via telephone or any other similar way at a time and place of their convenience. It will be emphasised that participants are free to terminate the interview at any time or not to answer any question.

The potential participants will indicate their consent by clicking a link that takes them to an online survey consisting solely of the participant information sheet and the consent form, and they can confirm their consent by clicking on a “I agree to participate in this research project” or “I do not agree”. Each potential participant will be given an individual link so the survey can greet them by name and record their consent.

The interviews are expected to take between 15 minutes to an hour, depending on the amount a participant wishes to contribute. There may be a subsequent, shorter interview to clarify matters.

The interview guide will include three open-ended questions, as follows:

a. What, in your opinion and experience, makes an exceptional doctor?

b. Have you ever met an exceptional doctor?

c. If yes, how did this doctor differ from other doctors?

There is scope for participants to raise issues not anticipated. The interviews will be, as recommended in qualitative research, flexible and responsive to the participant. Interviews will continue either until 10-15 interviews have been done or until saturation has been achieved, ie when additional participants only add marginal new relevant data; whichever is earlier.

The interviews will be recorded in audio or video format. Where available, video via zoom, skype or Microsoft Teams (Teams) will be used, otherwise audio-only zoom, skype or Teams, or, if the interviews are done in person or by telephone, an audio recorder will be employed.

The interviewees will receive a transcript of the interview and will be able to edit their answers. This will occur after the interviews have been transcribed. The transcripts will be sent to the interviewees by email, or mail if email is not available, and they will be offered to edit their answers before further processing by the researchers. If they choose to edit their answers they will be followed-up until they have done so, and they are free to change their decision to not wishing to edit, or to indicate their withdrawal from the study.

The Interviewer (Reflexivity)

The lead researcher will be doing all the interviews. He is a 62-year-old PhD student whose profession is being a financial adviser. He has experience in running focus groups to design a lengthy health survey for a group of women. In his professional work, he has extensive experience talking to a varied group of people with very different backgrounds about their financial matters. Many of these conversations become very personal without being intrusive, and without imposing his own opinions as that would impair the client’s ability to outline their situation and their own, at times unexpected, ways to respond to their situation.

This background will be useful for several reasons. One is that describing an experience with an exceptionally good doctor can be relaying a very personal event. Another is that any displayed bias can reduce the effectiveness of the interview.

It could be an advantage that the lead researcher is not a medical or health care practitioner as he will have fewer or no pre-conceived notions on exceptional doctors and the doctors will not feel the need to modify their opinions in the presence of a fellow professional. It could be a source of bias that, due to the two systematic reviews, the lead researcher is aware that there are doctors who have exceptionally positive effect on the physical health of the patients that they treat. However, the results of those previous systematic reviews show that there are exceptionally good doctors but do not suggested a reason for outperformance, so the only prior assumption is the lack of knowledge regarding what makes a good doctor.

Analysis

This is a qualitative study using Thematic Analysis as practised by Braun and Clarke.^39–41 As per Norris et al,⁴² Thematic Analysis

is a highly flexible methodology that can result in rich, complex accounts from different research participants, underlying similarities and differences, as well as generating unanticipated insights.

Thematic analysis will be used in an experiential framework, ie seeking to understand what doctors think and experienced in regards to exceptional doctors. The approach will be inductive while using a realist epistemological (what is the nature of knowledge) position which assumes that it is possible to obtain truthful knowledge on what makes exceptional doctors. As Braun and Clarke⁴³ state

Inductive analysis is therefore a process of coding the data without trying to fit it into a preexisting coding frame, or the researcher’s analytic preconceptions.

Phase one is transcribing and familiarisation with the transcripts.

Phase two is systematic coding of the data using codes that are relevant to the research question.

Phase three is development of candidate themes by examining the codes for similarities and overlap, similar to principal component analysis in survey responses.

Phase four is reviewing the candidate themes against the coded data for fit, and whether the themes tell a coherent story about the transcripts.

Phase five is going backward and forward between phases three and four, until the themes have been finalised.

Phase six is writing up, assembling and editing the resulting themes, including short data extracts to support any analytical claims.

Patient and Public Involvement

There is no patient or public involvement at this stage, though a survey for the public may be constructed after this project.

The results will be disseminated through scientific journals as well as magazines or websites aimed at medical doctors and the wider public.

Ethical Considerations

We will use consent forms for all interview participants. Names will be erased from the interviews or any other delivery document and, in the summary of the findings, contact details will be erased as well. All documents will be kept in secured premises of the lead researcher. Contact details will be destroyed after the summary reports have been sent.

One version of the transcripts will have identifying data but will not be processed further. A version of the transcript without identifying data will be used for further analysis. Video data is by definition identifiable but will only be available outside the research team if there is explicit informed consent from all participants in any interview.

The NVivo software is a secure qualitative data analysis tool used by the lead researcher on his password-protected computer.

Discussion

There is very little research on whether there are exceptionally good doctors, and, if they exist, the data is limited regarding who they are and how they differ from their peers. Previous qualitative research in this area was either on author-selected outstanding practitioners,^14–16 or about what patients think a good doctor is.^44–46 What this research adds is details on what doctors themselves consider to be criteria for exceptionally good doctors, and what they have experienced about them. Future research could then extend this approach comparing the definitions and experiences obtained from such doctors to the patients’ point of view.

Who is a good doctor? And what it takes to make an exceptionally good practitioner? Such questions are crucial to anyone involved in the healthcare practice, and more importantly for medical education. However, yet, responses to these questions are still elusive and not easy to answer, since there is no clear definition about the essence of a good doctor, and the answers most certainly vary depending on whom you ask according to their different needs and interests.⁴⁷ Prior studies seeking the answer of who a good doctor is have involved students, doctors, patients, and nurses. The aspects covered have included the general interpersonal qualities, communication skills and patient involvement, medical experience and skills, the doctor–patient relationship ethics, management and leadership qualities, teaching and supervision skills, research, and continuous education.¹⁸ Interestingly, it was found that patients have tended to evaluate the doctor according to their communication skills, while doctors were more focusing on medical skills. Herein, we intend to retrieve more answers about the essence of an exceptionally good doctor and the qualities behind their outperformance via semi-structured interviews of broader samples of medical doctors of any specialty.

In the future, the results of this project could be compared with the opinions and experiences of doctors who have been objectively identified as being exceptionally good in large cohort studies and those doctors and other practitioners who have worked with such doctors. This could provide useful information about how much the opinions of doctors are shared by their exceptionally good peers.