Demographic characteristics of the respondents
Prior to the analysis, 17 respondents who chose the same option as the response for 24 items (89%) or more were excluded as invalid data. Data for a total of 406 respondents (response rate 51.4%, male 286, female 115, not reported 5), including 117 (response rate 49.0%) 6th-year medical students, 60 (response rate 82.2%) 2nd-year residents, and 229 (response rate 47.9%) medical doctors at Kagoshima University Hospital, were included in this research. Among 222 medical doctors who responded with their age, 98 were faculty members, 26 were faculty members of research or other time limited projects, 97 were in another clinical position or were graduate school students, and 1 did not provide a response.
Medical specialties were classified into internal medicine, surgery, pediatrics, obstetrics and gynecology, psychiatry, emergency medicine, and others.
The mean ages of 6th-year medical students, residents, and medical doctors were 25.2, 27.8, and 40.5 years, respectively. Eleven medical students (9.6%) were aged 28 years or older, and seven residents (12.3%) were aged 30 years or older. Among medical doctors, the mean ages of faculty, those in project faculty positions, and others were 46.8, 43.2, and 33.6 years, respectively (Table 1, 2). The mean age for each specialty is shown in Table 2.
The mean length of clinical experience among medical doctors was 15.4 years (standard deviation 8.8 years, range 2-40 years). Doctors in a faculty position had a mean of 21.8 years of experience and longer experience than non-faculty doctors (Table 1).
Scores of the DS and 4 SASs
Cronbach’s alpha for the DS, SAS-2, SAS-3, SAS-4, and SAS-h were 0.70 (15 items), 0.56 (11 items), 0.57 (six items), 0.53 (three items), and 0.60 (six items), respectively.
Mean DS scores of medical students, residents, and medical doctors were 68.8 (SD 7.6), 72.2 (SD 7.2), and 73.8 (SD 8.2), respectively, and scores were higher in respondents with longer clinical experience (Figure 1). SAS-4 and SAS-h scores were high in medical doctors, especially doctors with experience of 30 years or longer. Mean DS scores of medical doctors in faculty positions, project faculty positions, and others were 75.5 (SD 9.4), 74.1 (SD 7.0), and 72.1 (SD 6.9), respectively, and faculty doctors showed higher scores on the DS, SAS-4 and SAS-h, and lower scores on the SAS-2 than doctors in other positions (Table 2).
Medical students and doctors with less than 6 years of experience had significantly higher score on the SAS-3 than the SAS-4 based on the t-test (p<0.05), indicating that respondents in these groups had a higher inclusion preference than independence preference and were around stage 3. Average scores on the SAS-4 of doctors with 6 to 12 years of experience were higher than those on the SAS-3, indicating that this group had a higher independence preference and were around stage 4. Doctors with 24 years of experience or longer, especially those with experience of 30 years or longer, had high DS, SAS-4 and SAS-h scores, indicating doctors in these groups were at stage 4 or 5.
Factors related to scale scores
Univariate analysis of variance for all respondents of the DS and the four SASs using gender and age indicated that age was a significant variable of DS (F(39,321)=2.47, p<0.001) and SAS-h (F(39,322)=2.70, p<0.001), but gender was not significant for any scale scores (p>0.05).
Univariate analysis of variance for medical students’ DS and the four SASs using gender and age indicated that age was a significant variable of DS (F(12,96)=1.99, p=0.33) and SAS-2 (F(12,95)=2.52, p<0.001), but gender was not significant for any scale scores (p>0.05).
Univariate analysis of variance for residents’ DS and the four SASs using gender and age indicated that both factors were not significant variables for any scale scores (p>0.05).
Because age and length of clinical experience are considered to be correlated, length of clinical experience was used for analysis of medical doctors.
Univariate analysis of variance for medical doctors’ DS and the four SASs using gender, position, specialty, and length of experience (<6, <12, <18, <24, <30, <36, ≥36 years) as independent variables indicated that length of experience was the significant variable of DS (F(6,136)=3.06, p=0.008), SAS-3 (F(6,139)=3.08, p=0.007), and SAS-h (F(6,138)=2.25, p=0.042) scores. Interaction of position and specialty (F(5,139)=2.73, p=0.022) and position, specialty, and length of experience (F(3,139)=3.30, p=0.022) were significant variables for SAS-3 score. Gender was not a significant variable for any scale scores (p>0.05).
Significant events and experiences that affected their present attitudes as medical doctors
The length of listed topics ranged from one Japanese word (e.g. RINSHO-JISSHU; clinical clerkship) to short sentences with 80 Japanese characters. At the end of categorization of topic themes, all topics were classified as either professional related experience or other personal experience. Topics of professional related experience were labeled as clinical experience, significant clinical event, patient, role model, mentor, or professional education or license. Each topic had at least one label. If the topic related to the experience of a significant clinical event described with the term ’patient’ as a key word, it was given the labels of ‘clinical experience’, ‘significant clinical event’, and ‘patient’. Personal experiences were labeled as self/own family/friend experience, learning abroad or at other institutes, reading/lectures/news, research experience, and others.
All topics were also categorized according to timing of experience: before medical school, during medical school, during residency program, during clinical practice (including experiences not specified during medical school or residency program), or timing not identified by description.
Among 406 respondents, 182 respondents (50 medical students, 25 residents, 107 medical doctors) listed at least one topic that affected their present attitude as a medical doctor (Table 3). The total number of topics was 478 (2.63 topics/respondent). Table 4 shows the number of respondents who listed each category.
One hundred sixty-seven respondents listed 233 clinical experiences, such as clinical clerkships and experiences during their residency training program. Among them, 96 significant clinical events, such as faced child birth, patient death, critical care, emergency, clinical difficulties, and high responsibility, were listed by 94 respondents. The term ‘patient’ was used by 76 respondents. For example, ‘death of my patient’, and ‘my patient expressed his gratitude with tears on my last day of the two-week clinical clerkship rotation’. Presence of medical doctors as role models(77 topics by 63 respondents), support by a mentor (26 topics by 21 respondents), and lectures in medical school courses and acquisition of a sub-specialty license (nine topics by nine respondents) were listed, and they were categorized as professional related experience.
Besides professional related experience, 94 respondents listed 135 personal experiences, and 86 of them were topics related to events in their own, their family members’, or their friends’ lives, such as ‘my experience of childbirth and childcare’, ‘family member’s death’, and ‘marriage’, listed by 72 respondents.
As for the timing, 199 topics were experienced during residency and clinical practice, and 116 topics were experienced when respondents were medical students. Medical doctors listed 16 topics that happened when they were medical students and 31 topics that happened when they were residents.
To clarify the relationship between experience and PI, respondents were classified into Kegan’s stage 2, 3, 4, 5, or transitional phases between stages using SAS scores. Staging criteria were created based on Kegan’s stage definition and the actual respondent’s score distribution. Criteria are shown in Table 4b, and 211 respondents (52.0%) were classified. A higher number of respondents in suspected stage 5 and the 4-5 transition phase listed significant clinical events and patient-related experiences as professional experiences than did respondents at stage 4 or lower.
Significant experience and PIF scale scores
To clarify the relationship between respondents’ past experiences and their current PIF scale scores, respondents were divided into high and low score groups for each PIF scale using the mean score of medical doctors.
The number of respondents who listed topics of each category in the high (≥74) and low (<74) DS score groupsis shown in Table 5A. Fisher’s exact test indicated that the number of respondents who did and did not list topics of patient, role model, or professional related experiences in high and low DS scores were significantly different.
Tables 5B, C, D, E show the number of respondents in the high and low score groups for SAS-2, 3, 4, and h who listed each topic.
The number of respondents who had a mentor in the high and low score groups for SAS-2, had clinical experiences and professional related experiences in the high and low score groups for SAS-3, experienced significant clinical events in the high and low score groups for SAS-4, and had patient, role model, and professional related experiences in the high and low score groups of SAS-h were significantly different.
For medical students, frequencies of high and low scores on the DS were significantly different whether respondents listed experience with patients or not. Significant clinical events for SAS-2 score, clinical experience for SAS-3 score, personal experiences for SAS-4 score, and reading/lectures/news for SAS-h score were significant.
For residents, no specific category or group was significant for PIF scale scores.
For medical doctors with less than 12 years of clinical practice, professional related experiences were significant for SAS-4 score. For medical doctors with 12 to 24 years of clinical practice, having a role model was significant for SAS-3 score. For doctors with 24 years or more of clinical practice, clinical experience was significant for SAS-3, and professional related experiences were significant for SAS-h score.