Study design and procedure
This was an open-label, parallel-arm randomised controlled trial. We invited midwives, nurses, and public health nurses who had roles in perinatal care to participate in a one-day e-learning course of perinatal mental health education. The participants were randomly divided into two groups: Groups A (intervention) and B (control). The contents of the e-learning consisted of two parts: perinatal mental health assessment and empathic communication. They viewed the same educational contents, but the timing of measurement was different. The perinatal mental health assessment part was learned in the morning (about two hours) and the empathic communication part in the afternoon (about two hours). The same paper-and-pencil multiple-choice test (empathic communication skills and assessment knowledge), video-viewing test (empathic communication skills), and an attitude test were conducted before the morning session (the pre-tests) in both Groups A and B. Similar tests were performed in the afternoon (the post-tests). The post-tests were performed before viewing the empathic communication part for Group B participants and after the viewing for Group A participants.
We compared the participants’ level of empathic communication skills between Group A and Group B. The sessions took place in Oita (April 2019), Yokohama (April 2019), and Sendai (June 2019). Participants who were included in the study submitted their signed consent form to the researcher.
Participants
Participants were midwives, nurses, and public health nurses who had roles in perinatal care. In Japan, in addition to midwives, obstetric nurses, and public health nurses also provide psychological support to women during the perinatal period. Therefore, such professionals other than midwives were included in this study. Participants were excluded if they had taken the same e-learning course before the study or were aged over 80, since the rate of internet use drops sharply for people over 80 years old, they are considered as difficult target population for e-learning education in Japan [16]. We recruited participants by sending a study recruitment letter and study consent form to midwife associations, hospitals, clinics, and birth centres in Oita Prefecture, Yokohama City, and Sendai City during periods of February 2019 through May 2019.
Randomisation and masking
Randomisation was performed with a random chart with a block size of four. A consecutive number of sealed opaque envelopes with a randomisation letter (A or B) were prepared by the researcher. Envelopes were opened by research assistants when the participants came in for the seminar, who were then allocated to one of the two groups. Due of the characteristics of intervention, masking was not applicable in this trial.
Educational content
The present educational course was developed by the T. and F. Kitamura Foundation of Studies and Skill Advancement in Mental Health (Tokyo). The material was recently developed for clinical midwives, as well as other perinatal health professionals, to acquire knowledge and gain skills of empathic communication in perinatal mental health issues. Both the perinatal mental health assessment and empathic communication sections consist of eight 15-minute video scripts. The perinatal mental health assessment section described (a) diagnostic criteria and the importance of mental health assessment, (b) the epidemiology of mental disorders commonly seen in the perinatal period, (c) mood disorders, (d) anxiety disorders, (e) psychotic disorders and puerperal psychosis, (f) child abuse and intimate partner violence, (g) perinatal bonding disorders, and (h) perinatal grief. The empathic communication section described (a) preparation for a mental health interview, (b) basic interview skills, (c) active listening, (d) probe techniques, (e) presentation of a case of acute insomnia after childbirth, and (f) catharsis.
Outcomes
The primary outcome was improvement of skills of empathic communication rated using a video-viewing test of a simulated patient interviewed by a midwife (SEC-V). Secondary outcomes were improvement in the skills of empathic communication rated using paper-and-pencil multiple-choice tests (SEC-P), knowledge of perinatal mental health assessment rated using paper-and-pencil multiple-choice tests (KPMHA), and attitudes towards support for women in perinatal periods.
Measurement
Demographic information and other variables: The participants were asked their age, years of clinical experience, educational background, and job title. These questions were asked in the pre-test.
Empathic communication skills: We used two types of measurements about the participants’ competency to empathically communicate with pregnant women. One was a paper-and-pencil multiple-choice test with five questions to evaluate the skills of empathic communication (SEC-P). Here, a short case vignette was presented with four choices for the best empathic response. One point was given for the correct answer whereas the other answers received zero points. The total score ranged from 0 to 5.
The second was a video-viewing test of a simulated patient interviewed by a midwife to evaluate the skills of empathic communication (SEC-V). A short (approximately 5 minutes long) video script (for example see Appendix) was shown and was followed by the question ‘If you were in charge of the care of this woman, what would you say to her? Give your response verbatim’. The degree of empathy of each narrative was rated objectively by evaluators (ES, AH, and YU) using the Empathic Understanding in Interpersonal Processes scale (EUIP [17]). The EUIP has a single item which is rated using a 5-point scale from not empathic at all (1) to very empathic (5). Low scores on the EUIP indicate that the participant’s utterance is irrelevant to or even antagonising the patient’s emotion while higher scores on the EUIP indicate that their utterance is in accordance with the patient’s emotion or even leading the patient into deeper insight. This evaluation was performed in the pre- and post-tests using different video scripts. In order to check interrater reliability of the EUIP assessment, the three raters independently rated the verbatim responses of 20% of the participants. Interrater agreement was κ = .80. In the remaining 80% of cases’ responses were rated by the three raters separately.
Knowledge of perinatal mental health assessment (KPMHA): To assess the knowledge of perinatal mental health, a pencil-and-paper multiple-choice test with 10 questions, each with a choice of four answers, was administered. Questions were created by experts in perinatal mental health. One point was given for the correct answer whereas the other answers received zero points. The score ranged from 0 to 10. To ensure an equivalent level of difficulty for pre- and post-tests, pilot-tests of 82 questions were conducted among six midwives who were not eligible participants to calculate the rate of correct answers, and 30 questions (15 each for pre- and post-tests) of the same level of difficulty were chosen. The final versions were asked in the pre- and post-tests.
Attitude for psychological support: We used the Counsellor Response Form (CRF [18]) to measure the participant’s active attitude towards providing psychological support to the case described in the video script. The CRF has a single item and is rated using a 9-point scale, ‘Do not want support her’ (1) to ‘Want to support her’ (9). This was asked in the pre- and post-tests.
Ethical considerations
The study protocol was approved by the Institutional Review Board of Tokyo Healthcare University (Kyou30-35C) and registered with UMIN-CTR in Japan (UMIN000036052; dated 1, March 2019). This clinical trial adhered to the Code of Ethics of the World Medical Association and the clinical research ethical guidelines for human subjects established on 27 April 2015, by the Ministry of Education, Culture, Sports, Science and Technology, and the Ministry of Health, Labour and Welfare, Japan.
Sample size and power
There have been few previous studies of this kind. Referencing a study using the EUIP [19], based on 95% power to detect a significant difference (α = 0.05, two-sided), 164 participants were required.
Statistical analysis
We used two-way analysis of variances (ANOVAs) to examine the main effects of the group (intervention vs control) and the timing (pre- vs post-tests) and their interactive terms on the scores of SEC-V, SEC-P, KPMHA and attitude. All statistical analyses were conducted using SPSS version 26.0.