In Japan, those wishing to practice clinical medicine enter a compulsory clinical graduate program after obtaining a national license for physicians. All trainees alternate between several clinical departments (junior residents; kenshui in Japanese) for two years. It is only after the two years of training that doctors proceed to higher postgraduate clinical training in specialized areas, which usually lasts more than three years (senior residents; senkoui in Japanese) .
In this study, we used a cross-sectional multicenter design. Approval to conduct this study was granted by the Institutional Review Board of the University of Tokyo (IRB approval number: 2021108NI).
Framework and participants
Thirty graduate clinical training hospitals in Japan were selected using information from the Electronic Residency Information System, which is a database of teaching hospitals developed and maintained by the Ministry of Health, Labor and Welfare of Japan. Twenty-five of the thirty hospitals agreed to collaborate in our research. Participating hospitals were geographically distributed throughout Japan. These hospitals varied in size and included both teaching and community hospitals (Table 1).
Anonymous questionnaires were distributed to all eligible participants [n = 1038] by their training program administrators in September 2021. Eligible study participants were all first- through sixth-year postgraduate (PGY) trainees in the training programs of the 25 hospitals that had committed to participate in the study. The participants filled out the questionnaires, put them in their respective envelopes and handed them over to those in charge of the training program. The administrators mailed them to the researchers. Participants were informed that participation was voluntary. About a week after the survey was distributed, they received a reminder. A second reminder was sent a week later.
Outcome Variable: Patient Ownership Scale
Ownership of patient care was measured using J-PCOS . J-PCOS is a 13-item instrument. Participants were asked to assume hospital care settings when completing the survey. Each item was scored on a 7-point Likert scale ranging from 1 = strongly disagree to 7 = strongly agree. Total J-PCOS scores were calculated by summing the item scores for each participant. Therefore, the total J-PCOS ranges from 13 to 91, with higher values indicating better PCO.
Based on a literature review [7, 11,12,13,14, 17]various environmental factors that may be associated with PCO were included as explanatory variables: hospital size (≤ 500 beds; 501–800 beds; 801–1000 beds; or ≥ 1001 beds), type of hospital (community hospital vs university hospital), medical care system (single shujii system vs multiple shujii (as described below in detail)), number of team members (≤ 2; 3-4; 5-6; or ≥ 7), number of inpatients in charge (≤ 3; 4-6; 7 -9; or ≥ 10), average number of hours worked per week, number of off-peak calls per month (0 vs ≥ 1), and perceived level of the workplace as a learning environment.
With regard to the question of the medical care system, participants were asked to choose between two of the following options: a single shujii (the physician primarily responsible for the patient) system and several shujii system (team system). In Japan, most hospitals used a single shujii system. In one shujii system, a single doctor takes charge of a patient’s care until discharge. In this system, even if other doctors are on call during their off hours, hospital doctors cannot just take a break because they have to constantly check on patients’ condition and treat them accordingly. . When a resident becomes a shujii for a given patient, the legal responsibility for patient care is considered to rest with the supervisor, but the majority of management (eg, scheduling exams, planning treatment, and explaining the condition to the patient) is left to the resident. However, in recent years, overwork of doctors has become a problem in Japan. [19, 20]and some hospitals have started adopting multiple shujii system as a counter-response. In many shujii system, multiple physicians are responsible for patient care as a team. As either of these two systems is likely to affect the development of trainees’ OCP, we decided to include it as an explanatory variable in this study.
We asked participants about the average working hours on weekends and weekdays and the number of night shifts per month they work. Based on previous studies [21,22,23]we calculated the average number of working hours per week using the following formula:
Average working hours per week = 5 * (Average working hours on weekdays) + 2 * (Average working hours on weekends) + 7 * (Number of night tasks per month/30) * (24–hours of average weekday work).
Previous research (both qualitative and quantitative) has shown that the learning environment can impact PCO training [12, 14]. In particular, the article validating the PCO scale showed that residents trained in a positive learning environment had a significantly higher PCO through bivariate analysis using the Mini-Rez scale of Linzer et al. . Accordingly, we decided to include workplace level as a learning environment as an explanatory variable. We chose a single-item global rating scale for the following three reasons. First, in the field of medical education and work environment research, the usefulness of global rating scales has been proposed because of their excellence in capturing nuanced elements. [24,25,26]. The learning environment is a multi-faceted concept, and its nuances can best be captured by a comprehensive rating scale. In fact, a one-item measure of the learning environment has been used in previous studies in the field of medical education. [27,28,29], which would justify our use of an overall score in this study. Second, in terms of response rate, shorter questionnaires generally give better results. . Since the survey covered many explanatory variables, there was concern that the response rate would decrease if a large number of questions were required for each explanatory variable. Third, to our knowledge, there is no valid Japanese version of the Mini-Rez scale. Therefore, we used a single measure of workplace level as the learning environment as follows: “Using any number from 0 to 10, where 0 is the worst possible learning environment and 10 is the best possible learning environment, what number should you use to rate your current department as a learning environment? »
We also included some personal factors as possible explanatory variables related to PCO based on previous studies. [7, 11,12,13,14, 17]; they are: educational level (PGY 1–2 (junior residents); kenshui) vs. PGY 3–6 (senior residents; senkoui)); the gender of the participants (female, male or other identities); and participant service (internal medicine, surgery or other services).
In Japan, internal medicine and surgery and other departments treat a large number of hospitalized patients. We considered that the differences by department could affect the PCO and decided to add the departments of the participants to the explanatory variables.
In this study, a linear mixed-effects model was used (random intercept model), which includes random effects for hospitals and explanatory variables (i.e. gender, education level, department, hospital size, type of hospital, medical care system, number of member teams, number of inpatients in charge, average weekly working hours, post-work on-call obligations and level of workplace as learning environment) as fixed effects. The full case analysis approach was chosen due to the small amount of missing data. One to two tails p a value