Participants
This was a single-center, randomized controlled study carried out in Zhongshan School of Medicine, Sun Yat-sen University between March 2016 and June 2016. Enrollment criteria included: (1) Third and fourth-year medical students; (2) Major subject of clinical medicine; (3) Absence of previous history of mental illness. Finally, a total of 146 medical students were enrolled. All the enrolled participants have read and signed the written informed consent before commencing the study.
Sample size
The power and sample size were calculated with two-sided Chi-square test at a 5% significance level. We estimated that the study would have 80% power to detect 5 points differences with standard deviation of 10 in the primary end point of Empathy between two groups with a sample size of 64 students assigned to each group. In the current study, the sample size was 146, rendering 85% power to detect a 2.5 points difference with a standard deviation of 5 in the quality of life scale and 64% power to detect a 15.7% difference between both groups in the depression.
Study design and randomization
Eligible participants were randomly assigned to the intervention (n= 74) and control (n=72) group according to the results sealed in an opaque envelope delivered by the co-investigators. The randomization coding list was generated by a simple randomization allocation method PLAN procedure (SAS, version 9.4, USA). Each student was assigned a number, which was linked to the corresponding contact information kept by the third party supervisor. Neither the students nor their teachers were blind to the allocated group due to the interactive schedules between both groups. Each group received one session per month for three months.
Data collection
In this trial, we used five validated questionnaires to evaluate the SE, QoL, depression, burnout and empathy. All participants were asked to complete the self-assessment scales and fill in their individual randomized number on the questionnaires before and one month after the whole intervention courses. The indicated scales were introduced as follows:
1. General Self-Efficacy Scale
Self-efficacy (SE) was measured by the General Self-Efficacy Scale (GSE Scale). The GSE scale was designed to measure a broad and stable sense of personal competence to effectively deal with a variety of stressful situations [29]. The scale was composed of 10 itemized questions that were answered on a 6-point Likert-type scale ranging from strongly disagree (score 1) to strongly agree (score 6). A higher total score meant a higher level of SE.
2. Medical Outcomes Study Short Form 8
QoL was measured by the Medical Outcomes Study Short Form 8 (SF-8). SF-8 was a valid quality of life survey that covered the same eight domains of the SF-36 but in a shorter (eight question) form [30]. The domains covered by the SF-8 Likert-type sacle included: physical function, limitations due to physical health problems, bodily pain, general health, energy/fatigue, social function, limitations due to emotional problems, psychological distress and mental well-being. The score of pain domain ranges from 1 to 6, while scores of the other domains range from 1 to 5. Higher scores represented better physical state and mental state.
3. Patient Health Questionnaire-9
Depression was measured by the Patient Health Questionnaire-9 (PHQ-9). PHQ-9 had 9 questions involving core symptoms of depression (anhedonia and depressed mood), suicidal tendency (suicidal thoughts), physical symptoms (trouble sleeping or concentrating, feeling tired, changes in appetite, feeling slowed down or restless) and the feeling of guilt or worthlessness over the past two weeks. Scores > 1 in the core symptoms of depression or suicidal tendency indicated the existence of depressive symptoms, in addition, medical students with a scoring > 9 points were also diagnosed with depression.
4. Maslach Burnout Inventory
Maslach Burnout Inventory (MBI) was a valid and widely used standard survey to measure burnout [31-33]. Burnout was comprised of three domains including emotional exhaustion (EE), depersonalization (DP), and personal accomplishment (PA). The MBI was a 7-point Likert scale including22-items that evaluated all of the three domains (the score range of EE was 0-54, the score range of DP was 0-30 and the score range of PA was 0-48). Medical students who scored ≥ 27 in the EE domain or ≥ 10 in the DP domain were identified as having at least one manifestation of professional burnouts [34].
5. Jefferson Scale of Empathy-Health Care Provider Student version
We measured empathy by the Jefferson Scale of Empathy-Health Care Provider Student version (JSE-HPS) that included 20 items answered on a 7-point Likert scale. Student were required to score each item according to the level of agreement (from 1 = strongly disagree to 7 = strongly agree). Ten items were negatively worded with reverse score. The total JSE-HPS score ranges from 20 to 140, with higher score values indicating higher degree of empathy [35]. Itemized questions in each survey were initially translated into Chinese for data collection, then translated back into English by bilingual researchers at the First Affiliated Hospital of Sun Yat-sen University to ensure the accuracy of translation. All of the questionnaires including the General Self-Efficacy (GSE) scale [36, 37], the Medical Outcomes Study Short Form 8 (SF-8) scale [38], the Patient Health Questionnaire-9 (PHQ-9) scale [39-42], and the Jefferson Scale of Empathy-Health Care Provider Student version scale [28] have been validated for a Chinese population.
Development of the Intervention courses
To increase the efficacy of the intervention courses, we attempted to know about the psychological disturbances of the 3rd and 4th year medical students before the intervention development. We randomly recruited five pre-clinical medical students (3rd year) and five interns (4th year) to serve as a pilot study group for the preliminary data collection to determine the themes of the interventional courses. We met with the pilot group members and discussed their impressions of the current clinical work and the related mental troubles for three months. Students often mentioned the lack of motivation and achievement in the daily repetitive medical work, the pressure to cope with the doctor-patient relationship and the need to acquire action-oriented strategies or skills to help them face the workload and potential medical negligence. Thus, topics concluded from the pilot study were organized into modules entitled “Establishing the sense of Achievement”, “Means for Efficient Patient-Doctor Communication”, and “Strategies to Manage Medical Errors”. To develop the problem-oriented interventions, we introduced three experienced clinicians to narrate the theme-related stories based on their own real clinical experience. In the “Establishing the sense of Achievement” module, we hoped to deliver four positive aspects of medical practice to our students including the professionalism of physicians to save lives, the meaning of gratitude or good faith from the patient, the importance of any newly acquired clinical knowledge from the daily work and the reputation award from colleague. These stories should help the students establish a good feedback of cognition and behavior in their daily clinical study and work, thus reducing the burnout and enhancing the self-efficacy. In the “Means for Efficient Patient-Doctor Communication” module, we emphasized the importance of the art of patient-doctor conversation and the empathy ability. The lecturer shared some successful examples (No regret or complaint on the clinical decision from patients or their family members) involved with the ethical background such as the expensive treatment cost and poor economics of the patient’s family, the unexpected blood relation of the parent and children, etc. The communication skills and empathy reflected by these vivid stories should help our students optimize the clinical decision or advise with the respect for the patients, thus promoting the students’ empathy and also reducing the depressive symptoms due to the released pressure. In the “Strategies to Manage Medical Errors” module, the lecturer shared the medical errors that she had made and the strategies to prevent and solve similar problems in her future work. These stories were of importance to tell our students that everyone could make mistakes even she was a very experienced doctor. In addition, how to report the mistake to the superior doctor and communicate with the patient and their family members after medical errors were essential skills for the students. These stories should help the students establish confidence, thus reducing depression and enhance self-efficacy. As a result, the quality of life should be improved due to the reduced stress. Building on previous literature [43-45], we combine the form of facilitated discussion groups that incorporated some elements of mindfulness, shared experience, and small-group learning to give the intervention. The same general structure was followed in each course: (1) check-in and welcome, (2) environment preparing (eg, leisure, friendly atmosphere, simple reflective exercise), (3) story narration, (4) skills and solutions learning, (5) facilitated group discussion, (6) summary and checkout.
Courses administration
The intervention course focused on three main subjects: (1) Establishing the sense of Achievement, (2) Means for Efficient Patient-Doctor Communication, (3) Strategies to Manage Medical Errors. Three experienced clinicians with at least 10 years of clinical service in our affiliated hospital were selected to serve as the lecturers. The lecturers were trained by a member of our team to ensure that they were familiar with the protocols and were able to narrate these stories fluently. The shared stories involved with various theme-related experience or skills about, for example, the positive aspects of medical practice such as letters of gratitude, awards or newly acquired experience, the strategies for efficient patient communication and ways to break bad news, and the dealing with medical errors.
The interventional courses were delivered in leisure, friendly atmosphere after the regular medical courses. The students were divided into three small groups (25students/group) based on the order of their code number and each group was guided by a corresponding lecture. During the course, each lecture narrated a story and summarized the skills for 20-30 minutes, then the students were given 30 min for group discussion and sharing their own feelings. Thus, each intervention lasted about 60 minutes. Students in the control group has random conversation under similar leisure, friendly atmosphere, however, no specific topics, coping strategies and skills were guided and discussed.
Follow up
In order to evaluate the long-term impact of the intervention courses, we contacted all the interventional group students two years after the intervention (June, 2018) by e-mails. They were requested to reflect on the interventional courses by answering the following three questions: (1) Please write down the most impressive story you remember from the interventional courses? (2) Which aspect of the interventional course influence you the most? (3) How did you apply strategies from the interventional courses in your daily clinical practice? The qualitative data were collected and analyzed.
Statistical Analysis
Questionnaires with missing information in three or more items were excluded from subsequent analysis. If the scores of few scale items were missing, the median of the scores of the same items in the remaining students was used to replace the missing data. The baseline characteristics of the participants were described with the mean (SD) if the variable follows a normal distribution, and if the data did not follow a normal distribution, the median (IQR) was used. Independent sample t test or Wilcoxon rank sum test was used to compare differences for continuous variables between two groups while Chi-square test or Fisher exact test was used for analyzing the categorical variables. The statistical differences between the baseline and follow up scores for each outcome separately in intervention and control group were also compared using the paired t test, in which the difference between the baseline and follow up data was calculated and compared. All of the statistical analysis were performed using SAS (version 9.4, USA). A P value < 0.05 was considered to be statistically significant.