The Effect of a Structured, Home-based Interview With a Patient on First-year Medical Students' Patient-centredness.

Background Doctors are regarded as "professionals" both by the public and by their peers [Medical Protection Society, 2017]. They are the most trusted profession among the public, and this has been the case for many years. Medical professionalism is not a new concept and has been present in the history of medicine in the form of a Hippocratic Oath taken by physicians [Seif-Farshad M, Bazmi S, Amiri F, et al, 2016]. However, medical professionalism is difficult to define and remains poorly understood [Riley S, Kumar N, 2012]. While professionalism has become a widely emphasised subject in medical education and medical practice, there is still a lack of common understanding about the meaning of the concept. Discussions on the subject have thus been unsystematic, as the word "professionalism" has multiple meanings, complexities, and subtle differences [Seif-Farshad M, Bazmi S, Amiri F, et al, 2016].

Professional healthcare groups have their own set of norms (codes) which guide members of that profession in terms of how they should behave professionally. [Monrouxe V, Rees C, 2017] These are designed and implemented by the profession's regulatory body, so differ according to different healthcare professions and countries. Over one hundred different dimensions of professionalism have been identified. Swick [2000] proposes that medical professionalism consists of specific behaviours (Table 1).

Concerns about professionalism in medicine have made the explicit teaching and learning of ethics, professionalism and personal development necessary. The General Medical Council in the UK, and other professional bodies in both Europe and the Americas, have emphasised the need to enhance the teaching and learning of professionalism in medical schools, particularly the development of good attitudes, appropriate and competent skills, and the inculcation of a value system that reflects the principles of professionalism in medicine [Sivalingam, 2004].

A study investigating tutors' and students' perspectives of the delivery of professionalism in the early years of Glasgow's learner-centred, problem-based learning (PBL) medical curriculum found that [Goldie J, Dowie A, Cotton P, et al, 2007]: early patient contact experiences were found to be particularly important, and that learning activities promoting critical reflection were most effective. However, a systematic review found no unifying theoretical or practical model to integrate the teaching of professionalism into the medical curriculum [Birden H, Glass N, Wilson I, et al 2013]. There is no consensus about how best to teach professionalism [Stockley AJ, Forbes K, 2014], and few studies have explored the effectiveness of different teaching and learning methods for professionalism [4].

Research on students' understandings of professionalism identified 19 dimensions [Monrouxe V, Rees C, 2017]. Of these, patient-centredness was the second most discussed dimension (after 'professionalism as rules'). Students' ideas of patient-centred professionalism came from a variety of sources: formal lectures on ethics, informal learning through role models, and from formal assessments. In the patient-centred clinical method, both the physician's and the patient's agendas are addressed by the physician and any conflict between them dealt with by negotiation [Levenstein JH, McCracken EC, McWhinney IR, et al, 1986]. This means that the physician aims to gain an understanding of the patient as well as the disease, and it is in contrast to the disease-centred method in which only the doctor's agenda is addressed. Summarising patient-centeredness elegantly, McWhinney describes the patient-centred approach as one where the "physician tries to enter the patient's world, to see the illness through the patient's eyes" [McWhinney 1989]. Evidence suggests that patient-centred care is associated with a number of favourable biomedical, psychological and social outcomes [Bauman AE, Fardy HJ, Harris PG, 2003], and it has been recognised as an important aspect of quality in health care [Institute of Medicine Committee on Quality of Health Care, 2001].

There have been more than 900 papers published that measure patient-centred care or one of its components [de Silva, 2014]. However, only two measurement instruments for attitudes towards patient-centredness in undergraduate medical students have been identified: the Doctor Orientation Scale and the Patient-Practitioner Orientation Scale (PPOS); of these two, the PPOS has been used much more extensively [Archer E, Bezuidenhout J, Kidd MR, et al, 2014]. The PPOS was developed in 1999 to measure the attitudes of medical students towards patient-centredness [Krupat E, Hiam CM, Fleming MZ, et al, 1999]. It differentiates between patient-centred versus doctor-centred or disease-centred orientation, measuring attitudes along 2 dimensions: 'sharing' and 'caring' [Tsimtsiou Z, Kerasidou O, Efstathiou N, et al, 2007]. It has been used to assess attitude changes towards patient-centredness in medical student cohorts as they progress through the clinical curriculum.

The PPOS is available in 13 languages. It has been translated into German and the degree of medical students' patient-centeredness assessed in 2 student samples in Freiburg, Germany and in Basel, Switzerland [Kiessling C, Fabry G, Rudolf Fischer M, et al, 2014]. Construct validity was tested using factor analysis. Based on factor analysis and tests of internal consistency, a shortened version with 6 items for each of the 2 subscales "sharing" and "caring" was generated (PPOS-D12). PPOS-D12 (see Appendix 1) was found to be a reliable instrument to assess patient-centeredness among medical students in German-speaking countries.

In a longitudinal survey of medical students' attitudes toward patient-centred care in Greece, students' attitudes were significantly less patient-centred at the end of their studies compared to the beginning of their clinical curricula (mean score in year 4: 3.96; mean score in year 6: 3.81; P < 0.001) [Tsimtsiou Z, Kerasidou O, Efstathiou N, et al. 2007]. In a South African study, medical students from all undergraduate six years took the PPOS survey. There was a decrease in mean scores (from 2.65 in first-year students to 2.25 in final-year students), with the most pronounced decrease in the first two years of study [Beattie A, Durham J, Harvey J, et al, 2012]. In a year-long study of resident physicians at a university hospital in Tokyo, PPOS scores reduced significantly over the year (mean score at start of year: 4.5, SD 0.48; mean score at end of year: 4.39, SD 0.51; change: -0.11, SD 0.42) [Ishikawa H, Son D, Eto M, et al. 2018].

The investigators have only identified one study that assessed the effect of an intervention on students' levels of patient-centredness. In this uncontrolled UK study, first-year dental undergraduates were given an attitudinal questionnaire to complete before and after their behavioural science course. No significant difference was found between their mean pre- and post-course PPOS scores (pre-course mean scores: 3.44 SD 0.33; post-course mean score: 3.37, SD 0.19; P > 0.05). The investigators have found no controlled trials that test the effect of interventions on medical students' levels of patient-centredness, and no trials that look at the effect of early patient contact in the form of an unaccompanied visit to a patient in their home.

Aim of the study The aim of this study is to assess how a structured, in-depth, home-based interview with a patient with a chronic illness affects first-year medical students' patient-centredness.

A nested study will assess whether the GP's level of patient/doctor-centredness affects changes in their student's level over the year.

Methods

Study setting The study will take place in the Berner Institut für Hausarztmedizin (BIHAM) at the University of Bern, Switzerland, and a subset of the >700 General Practitioner (GP) teaching practices that are affiliated to it.

Study participants The population will be medical students who are in the first year of their studies (their first Bachelor year) at the University of Bern, during their longitudinal placements in primary care. The GP teachers are family doctors that are accredited by BIHAM to take medical students.

Study design This will be a randomised controlled trial. During their first Bachelor year, each medical student will spend a series of six half-days at the practice of a GP teacher that she/he has been allocated to. Students will be randomly assigned to either the intervention or the sham comparator which will take place during their last half-day in their practices. All students will complete an on-line version of the PPOS-D12 questionnaire at the start of the academic year and once more after the last half-day at their GP teachers' practices.

Intervention The intervention will be a structured in-depth interview with a patient with a chronic illness that has been chosen by the student's allocated GP teacher. These chronic diseases are the four conditions at the top of a list of diseases with high disability-adjusted life years (DALY) scores in Switzerland: ischemic heart disease, low back pain, major depressive disorder and COPD [Institute for Health Metrics and Evaluation, 2010]. GP teachers who only see children are asked to choose a patient who has a chronic cardiac condition, a chronic lung disease, or another chronic illness that has a significant effect on the child's quality of life, the interview to be carried out at the child's home with one or both parents, and the child if she/he is old enough to take part in the interview.

GP teachers and students will be told that the students' intervention interviews need to be unaccompanied and at patients' own homes, but in justified, exceptional cases, and after consultation with a member of the research team, the interview may take place in the GP teacher's practice premises. The interview will be followed by a structured interview with the practice nurse and then a structured debriefing interview with the GP teacher.

Sham comparator In the sham comparator, the student's allocated GP teacher will give the student time to read a document that gives information about consultation skills, and asks questions that the student will need to discuss with the GP teacher. The document is designed to have real educational value, and to complement BIHAM's department-based consultation skills teaching. The use of this approach as the sham intervention is based on a study that found that a behavioural science course (including consultation skills teaching) had no effect on students' PPOS scores [Beattie A, Durham J, Harvey J, 2012].

Data collection At the start of the academic year (i.e. before the interventions), students will complete a SurveyMonkey questionnaire that asks for demographic information (gender, history of serious chronic illness in the participant, a friend or close relative, history of training or patient contact other than in medical school) and the PPOS D12 form.

At the end of their GP attachment (i.e. after the interventions), students will complete a SurveyMonkey questionnaire that asks whether they had the intervention or sham comparator or neither (and if neither, why), and the PPOS D12 form. Consent to use these data for research will also be requested.

At the end of their students' attachments, their GP teachers will complete a SurveyMonkey questionnaire that asks them to complete the PPOS D12 form. Consent to use these data for research will also be requested.

Each of these questionnaires will ask for the participant's name, so that pre-intervention, post-intervention, and GP's data can be linked and compared.

Outcome measures The primary outcome measure will be the change in students' PPOS-D12 scores from base-line (at the start of the academic year) to the end of their year-long primary care attachment.

Secondary outcomes will be the effect of students' gender, previous experience and prior exposure to chronic illness in the participant or her/his close relatives and friends on their PPOS-D12 scores A nested study will measure the strength of association between the GP teachers' own levels of patient/doctor-centredness and changes in their students' levels over the year.

Development of the intervention Following a literature review and discussion, RF, A-LC and MH developed a patient interview pro-forma which was designed to identify patients' views on their illnesses, how it affects their lives in physical, psychological and social terms, and how their relationships with their GPs impacts on their lives. The sections of the interview pro-forma map across to Mead and Bower's conceptual framework of patient-centredness given above [Mead N, Bower P 2000].

Six medical students who had been visiting their GP teachers for eight half-days per year since their first year of study agreed to pilot the intervention in November 2017. The GP teachers were asked to select a patient with a chronic illness for their students, and to organise 2.5-hour interviews, half of the students at the patients' home and in the other half in the teaching practices. Students were asked to use the questionnaire pro-forma. In addition, the students each had a 30-minute interview with a practice nurse to assess their perceptions of the effect of the illnesses on their patients, and finally a one-hour discussion with their GP teacher to reflect on their encounters with the patients, discuss their findings, and talk about any difficulties that may have arisen.

Following this, RF conducted a focus group interview with the participating students. The discussion was recorded and transcribed in full. A-LC and MH analysed parts of each the transcript independently, then independently developed a coding frame. Their coding and thematic analysis was similar. The findings (the experiences, suggestions and criticisms of the piloting students) were used to make improvements in the organisation and in the information sheets for patients, students, MPAs and GPs. The themes, findings and improvements needed are shown in Appendix 2.

Three medical students then piloted the re-designed interview pro-forma with one of the researchers (A-LC) role-playing the patient), and they gave feedback with recommendations on how the pro-forma could be further improved. Following this, RF, A-LC and MH reviewed the pro-forma and designed the final version. The final version of the interview proforma, in German, is shown in Appendix 3. The interview proforma and instructions on implementation of the intervention will be sent to students and their GP teachers before the students' final visits to their teaching practices.

Development of the sham comparator MH wrote an English-language 'Communication skills for BIHAM medical students' didactic document, which included sections on 'Why should I learn about communication skills?', 'What do patients want from a doctor?', What information are patients looking for?', What consultation skills should I use when I interview patients?', 'Do good communication skills really matter?', 'How can I learn good interpersonal and consultation skills?', 'How is a good consultation structured?', and 'Questions to discuss with your GP teacher'. This document was translated into German by A-LC. The resulting 1,850-word, 9-page document and instructions on implementation of the intervention will be sent to students and their GP teachers before the students' final visits to their teaching practices.

Randomisation Participants will be allocated to intervention or sham intervention through random sequences generated in the SPSS statistical package.

Blinding procedures

This study will be partially blinded:

Screening and enrolment. A person blinded to the purpose of the interventions (i.e. that one is the intervention, the other is a sham comparator) will enrol participants and allocate them to their arm of the study. This person will work separately from the rest of the trial team and all team members will be asked to sign a form stating that they will not disclose the purpose of the interventions to this person.

Students. In a tailored informed consent procedure, students will be given a 'high-level description' of the study objectives with only superficial information on the study interventions, as accepted by ethics committees in similar studies. They will be informed that they are randomised to one of two study groups, without revealing that one is an intervention and the other is a sham comparator. The aim of this is to minimise performance and other reporting biases.

GP teachers. A similar approach with be taken with information for the GP teachers, so that they also have only a 'high-level' description of the study objectives.

Statistical analysis Descriptive statistics will be used to describe the relationships between student demographics and PPOS scores. Mean pre- and post-test PPOS scores will be compared using paired t tests. Differences between PPOS scores in the intervention and control groups will be examined with unpaired t tests. For the nested study, GPs' PPOS scores and changes in their students' PPOS scores will be compared using paired t tests.

The sample size calculation for the primary outcome was based on reported mean reduction of 0.11 in the PPOS score over one year in a group without any intervention [20], and a mean increase of 0.05 in the intervention group. The investigators therefore assumed a mean difference 0.16, with a standard deviation of 0.42. The study was designed with a 5% level of significance and a 90% power to reject the null hypothesis of equivalence between the two groups. To achieve this objective, 220 students would be required (110 for each group). Assuming a 20% drop-out, the investigators therefore aimed to enrol a total of 275 students.

Discussion This research will consider the effect of an in-depth, structured interview with a patient with a chronic illness on changes in first-year medical students' levels of patient-centredness. There is existing evidence that medical students' levels of patient-centredness reduce over their student years, and this study will contribute to an understanding of how this reduction can be minimised or reversed.

Strengths To the investigators' knowledge, this is the first randomised controlled trial that has been designed to study the effect of an intervention on medical students' levels of patient-centredness. The intervention materials were carefully developed and piloted by GP teachers and medical students, and therefore grounded in their clinical and educational experience.

Limitations Although the research team will give participants information about the study in a presentation and in correspondence, there is a risk that fewer students than anticipated consent to take part in the study. While the researchers aim to blind participants to the nature of the comparison, i.e. that one intervention is the intervention and the other is a sham comparator, control intervention, participants may guess the researchers' intentions, and this may result in bias in their responses. The power calculations have been based on data from other published studies, but these may not be directly comparable to the investigators' own study. The control intervention has been chosen because of evidence that teaching dental students about communication skills does not affect their levels of patient-centredness, but it is possible that the investigators' communication skills intervention will indeed have an effect the levels of medical students' patient-centredness, and thus not be a truly inactive control.

Expected impact

One of the aims of BIHAM's clerkships in primary care is to shift the medical students' focus towards 'professionalism', and patient-centredness is an important aspect of professionalism. This study will assess whether a single in-depth structured interview with a patient, followed by a de-briefing interview with the patient's GP, can contribute to achieving this aim.