Feasibility and acceptability of a culturally adapted psychological first aid training intervention (Preparing Me) to support the mental health and well-being of front-line healthcare workers in China: a feasibility randomized controlled trial

ABSTRACT

Background: Psychological first aid (PFA) training helps to prepare healthcare workers (HCWs) to manage trauma and stress during healthcare emergencies, yet evidence regarding its effectiveness and implementation is lacking.

Method: A two-arm feasibility randomized controlled trial design was conducted in a Chinese tertiary hospital. Participants were randomly allocated to receive either a culturally adapted PFA training (the intervention arm) or psychoeducation (the control arm). Feasibility indicators and selected outcomes were collected.

Results: In total, 215 workers who expressed an interest in participating in the trial were screened for eligibility, resulting in 96 eligible participants being randomly allocated to the intervention arm (n = 48) and control arm (n = 48). There was a higher retention rate for the face-to-face PFA training session than for the four online group PFA sessions. Participants rated the PFA training as very helpful (86%), with a satisfaction rate of 74.25%, and 47% reported being able to apply their PFA skills in responding to public health emergencies or providing front-line clinical care. Positive outcome changes were observed in PFA knowledge, skills, attitudes, resilience, self-efficacy, compassion satisfaction, and post-traumatic growth. Their scores on depression, anxiety, stress, and burnout measures all declined. Most of these changes were sustained over 3 months (p < .05). Repeated measures analysis of variance found statistically significant interaction effects on depression (F_2,232 = 2.874, p = .046, $\eta p^2$ = .031) and burnout (F_2,211 = 3.729, p = .018, $\eta p^2$ = .037), indicating a greater reduction in symptoms of depression and burnout with PFA compared to psychoeducation training.

Conclusion: This culturally adapted PFA training intervention was highly acceptable among Chinese HCWs and was feasible in a front-line care setting. Preliminary findings indicated positive changes for the PFA training intervention on knowledge, skills, attitudes, resilience, self-efficacy, compassion satisfaction, and post-traumatic growth, especially a reduction of depression and burnout. Further modifications are recommended and a fully powered evaluation of PFA training is warranted.

HIGHLIGHTS

• Psychological first aid (PFA) training was culturally adapted and evaluated to help prepare healthcare workers to manage trauma and stress during healthcare emergencies.

• This culturally adapted PFA training was highly acceptable among Chinese healthcare workers and was feasible in a front-line care setting.

• Preliminary findings show positive changes for the PFA training intervention on knowledge, skills, attitudes, resilience, self-efficacy, compassion satisfaction, and post-traumatic growth, especially a reduction of depression and burnout.

Antecedentes: La capacitación en primeros auxilios psicológicos (PFA, por sus siglas en inglés) ayuda a preparar a los trabajadores de la salud (TS) para manejar el trauma y el estrés durante emergencias de atención médica, pero falta evidencia sobre su efectividad e implementación.

Métodos: Se llevó a cabo un diseño de ensayo controlado aleatorizado de viabilidad de dos brazos en un hospital terciario chino. Los participantes fueron asignados aleatoriamente para recibir una capacitación en PFA adaptada culturalmente (el brazo de intervención) o para recibir psicoeducación (el brazo de control). Se recopilaron indicadores de viabilidad y resultados seleccionados.

Resultados: Se evaluó la elegibilidad de 215 trabajadores que expresaron interés en participar en el ensayo, lo que dio como resultado que 96 participantes elegibles fueran asignados aleatoriamente al brazo de intervención (n = 48) y al brazo de control (n = 48). Hubo una tasa de retención más alta para la sesión de capacitación presencial de PFA que para cuatro sesiones grupales de PFA en línea. Los participantes calificaron la capacitación de PFA como muy útil (86%), con una tasa de satisfacción del 74,25%, y el 47% de ellos informaron que podían aplicar sus habilidades de PFA para responder a emergencias de salud pública o brindar atención clínica de primera línea. Se observaron cambios positivos en sus conocimientos, habilidades, actitudes, resiliencia, autoeficacia, satisfacción por compasión y crecimiento postraumático de la PFA. Sus puntuaciones en las medidas de depresión, ansiedad, estrés y agotamiento disminuyeron. Es alentador que la mayoría de estos cambios se mantuvieran durante tres meses (p < ,05). El análisis ANOVA de medidas repetidas encontró un efecto de interacción estadística sobre la depresión (F_2,232 = 2,874, p = ,046, $\eta p^2$ = ,031) y burnout (F_2,211 = 3,729, p = ,018, $\eta p^2$ = ,037), indicando una mayor reducción de los síntomas de depresión y burnout en la PFA en comparación con la formación en psicoeducación.

Conclusión: Esta intervención de capacitación en PFA adaptada culturalmente fue muy aceptable entre los trabajadores de la salud chinos y fue factible en un entorno de atención de primera línea. Los hallazgos preliminares indicaron cambios positivos para la intervención de capacitación de PFA en conocimientos, habilidades, actitudes, resiliencia, autoeficacia, satisfacción por compasión y crecimiento postraumático, especialmente una reducción de la depresión y el agotamiento. Se recomiendan modificaciones adicionales y se justifica una evaluación completa de la capacitación de PFA.

KEYWORDS:

• Psychological first aid training
• front-line healthcare workers
• mental health and well-being
• feasibility randomized controlled trial

PALABRAS CLAVE:

• Capacitación en primeros auxilios psicológicos
• trabajadores sanitarios de primera línea
• salud mental y bienestar
• ensayo controlado aleatorizado de viabilidad

1. Background

Healthcare workers (HCWs) are frequently exposed to potentially traumatic events, both directly, such as during the coronavirus disease 2019 (COVID-19) pandemic, and indirectly through experiencing workplace violence and witnessing patient suffering. These intense work environments increase the risk of HCWs experiencing emotional and psychological demands due to having multiple exposures to trauma (d'Ettorre et al., 2021; Orrù et al., 2021). The COVID-19 pandemic has further exacerbated these challenges, leading to a greater likelihood of long-term consequences, such as HCWs experiencing post-traumatic stress, depressive symptoms, and professional burnout (Dutheil et al., 2021; Scott et al., 2023). This not only affects individual HCWs but also has led to organizational implications such as decreased levels of patient safety and increased staff turnover (Benfante et al., 2020; Patel et al., 2018). Unlike military personnel and disaster responders, HCWs, particularly those outside emergency settings, often lack specific training to effectively manage trauma and stress, emphasizing the need for interventions.

Psychological first aid (PFA) is an early intervention to reduce acute distress and facilitate adaptation after trauma exposure (Brymer et al., 2006; Everly Jr & Lating, 2022). Originally introduced in the 1940s (Blain et al., 1945), PFA has been developed as an initial intervention which is distinct from psychological debriefing to avoid potentially unhelpful elements, that is, emotional catharsis (Raphael, 1986; Shultz & Forbes, 2014). PFA is underpinned by five essential elements to prevent continued trauma: (1) promoting a sense of safety, (2) calm, (3) self and collective efficacy, (4) connectedness, and (5) hope (Hobfoll et al., 2007). Typically, PFA in its current form involves providing humanistic and practical support while emphasizing the provision of information, comfort, practical assistance, and making referrals as needed (Brymer et al., 2006; Everly Jr & Lating, 2022; World Health Organization, 2011). Unlike traditional psychotherapy, PFA can be taught to individuals with limited clinical expertise, making it a widely accessible first-line psychosocial support approach (Shultz & Forbes, 2014). PFA training has been advocated by international and professional organizations to equip first responders to intervene and provide early mental health support and facilitate stepped care for recovery at the individual, organizational, and public health levels, particularly for humanitarian aid and disaster relief situations (Ni et al., 2023; Shah et al., 2020). Previous scoping reviews have reported positive outcomes associated with PFA training, including improved mental and behavioural health preparedness, enhanced resilience, and increased self-efficacy, providing beneficial support to occupational groups frequently exposed to traumatic events (Wang et al., 2021). In recent times, anecdotal evidence reveals a growing interest in providing PFA training to high-risk occupational groups, such as police, emergency workers, and social care practitioners (Tessier et al., 2022; Geoffrion et al., 2023; Lalani & Drolet, 2018). Some medical schools and residency programmes have even started incorporating PFA to address distress behaviours. However, the effectiveness of implementing PFA training to these groups and in these settings remains unclear.

A review examining current PFA training initiatives reported issues regarding its limited generalizability, a wide variability in PFA training programme content, and a lack of culturally competent curricula (Ni et al., 2023). The inconsistency in how PFA training has been delivered, and the lack of clear learning objectives and outcomes, have implications for the long-term sustainability of PFA training (Wang et al., 2021). This may explain emerging reflections from PFA providers, who describe not having a clear understanding of PFA, and experiencing challenges in being able to identify the mental healthcare needs of PFA recipients. They report having difficulties maintaining PFA fidelity, such as finding the right balance between offering reassurances and avoiding making false promises that are not in line with the endorsed ‘do no harm’ principles within the PFA approach (Horn et al., 2019). The COVID-19 pandemic has prompted adaptations of PFA through digital training models to help HCWs to manage both patient distress and their own stress (Ni et al., 2023); however, the effectiveness of these adaptations remains to be established. Therefore, there is a need for research on well-adapted, rigorously evaluated PFA training programmes that are tailored to the needs of HCWs, and to establish their effectiveness and how best to sustainably implement them in front-line healthcare organizations.

1.1. ‘Preparing Me’ project

The ‘Preparing Me’ project sought to pilot a culturally adapted PFA training programme for use by Chinese HCWs. Implementation science theories were used to inform decisions around how to culturally adapt and evaluate PFA training, through the selection of sensitive outcome measures. Examples of cultural differences included: (1) Chinese HCWs have a dual role, both providing routine care and responding to disaster and public health emergencies, unlike most PFA training initiatives implemented in humanitarian aid and disaster relief settings; and (2) the cultural norms, values, and language in eastern China are distinct from Western and English-speaking contexts. Adapting PFA training to this HCW population has the potential not only to contribute towards adding an understanding of PFA in front-line care contexts, but also to strengthen the evidence base for PFA implementation.

Therefore, a systematic cultural adaptation has been conducted with the aim of enhancing the appropriateness and effectiveness of a PFA training programme to align with the working context of Chinese HCWs, called READ-Y. This adaptation process was informed by the ADAPT guidance and cultural adaptation frameworks on how to culturally adapt interventions, and included conducting formative studies with multi-stakeholder involvement (Moore et al., 2021; Bernal et al., 2009; Heim & Kohrt, 2019; Sangraula et al., 2021). Three cohorts of stakeholders, including HCWs, mental health experts, and clinical education experts, were actively engaged in providing contextual insights into Chinese culture, the working environment, and nature of trauma to which HCWs were exposed, as well as more detailed descriptions of the specific contexts in which PFA would be delivered. These efforts, alongside a multi-stakeholder consultation, led to the tailoring of the READ-Y PFA training programme, which was originally based on the John Hopkins Guide to PFA, published in the USA (Everly Jr & Lating, 2022). This culturally adapted PFA training has undergone iterative refinement, manualization, and the development of accompanying training resources, before being implemented in a feasibility randomized controlled trial (RCT).

1.2. Aims and objectives

PFA training programmes are examples of complex interventions, which have multiple components. The intended outcomes from PFA training are influenced both by the wider cultural context and by how those who deliver and receive the PFA training react to it. The Medical Research Council (MRC) framework for developing and evaluating complex healthcare interventions suggests conducting feasibility testing before a full-scale evaluation (Skivington et al., 2021). Guided by the MRC framework, this feasibility trial aims to ensure that PFA training intervention can be delivered as intended, provides insight into trial parameters, such as recruitment and retention rates, and provides preliminary estimates of potential outcome changes. This article presents the results of the feasibility RCT, while the findings from the cultural adaptation of PFA study and an embedded process evaluation will be reported elsewhere.

The objectives of the feasibility trial were:

1. to determine the feasibility and acceptability of the adapted Chinese PFA training intervention in relation to response, recruitment, and retention rates

2. to examine potential changes on the preparedness and related psychosocial measures, including behaviours, resilience, and professional quality of life, of HCWs who received the PFA training, compared to those who received a psychoeducation training session, which was regularly delivered to HCWs in this setting

3. to test the trial procedures to decide whether a definitive full-scale RCT is warranted.

2. Method

2.1. Study design

A parallel-arm, feasibility RCT design was conducted from 16 December 2021 to 27 March 2022.

2.2. Setting

The feasibility trial was conducted in the Second Xiangya Hospital, Central South University, Changsha, Hunan Province, China. This tertiary hospital has 4600 beds and 135 clinical wards serving over 100 million people in Hunan province and the surrounding provinces, including Jiangxi province. The study site was selected for two reasons. First, it is a designated national emergency response centre, with a remit of responding to all public health emergencies and an organizational interest in improving healthcare emergency preparedness. Secondly, as a key national centre for healthcare, education, and clinical research in China, it has high levels of complex patient cases and a commitment to addressing staff well-being needs.

2.3. Sample

Participants were front-line HCWs who work clinically providing direct care to patients, and included doctors, midwives, nurses, and pharmacists.

The sampling frame consisted of: (1) HCWs from emergency response disciplines; namely, the emergency department, infectious disease department, paediatric department, and intensive care units; and (2) HCWs who potentially could be temporarily rotated to provide an emergency response.¹

The inclusion criteria were: (1) being a certified HCW who worked clinically providing direct patient care; and (2) being employed in an emergency response department, or being potentially deployed as part of an emergency response team.

The exclusion criteria were: (1) not being available to complete the full training programme (e.g. on sick leave or other planned commitments); and (2) having prior knowledge in related mental health areas (e.g. completed 3 h training in mental healthcare or practised in emergency mental healthcare).

2.4. Sample size and power

Recommendations on how to decide the sample size for a feasibility study were considered. These recommendations included that: (1) the sample size should be 10% of the projected sample size for the definitive study that will be informed by the feasibility study; or (2) a sample size of between 24 and 50 may be reasonable in a feasibility study (Billingham et al., 2013). Given that the actual sample size for a definitive study is yet to be determined, this study followed the guidance set out in recommendation (2), taking account of possible dropout. Thus, this study aimed to recruit a sample of 60–80 front-line HCWs, to compensate for a possible dropout rate of 35%, as experienced by the only previous trial of PFA training (Sijbrandij et al., 2020).

2.5. Recruitment

The study was publicized to potential participants using two approaches. First, gatekeepers (the directors from the nursing and medical departments) helped with disseminating the project information to clinical ward managers and asked them to cascade this down to their staff in shift briefings. Each clinical ward represents front-line clinical care units which have an average of 30 staff members per unit team, consisting of 10 doctors and 20 nurses, and serving 47 beds in one care unit. Secondly, the announcement of the project was shared via word of mouth, advertisements in social media, or posters displayed around the hospital. There were no financial incentives or pressure from senior leadership to take part in the study. Any interested participants were asked to contact the research team directly.

2.6. Screening and informed consent

The research team asked all potentially eligible front-line HCW screening questions to determine whether they met the eligibility criteria. They were given an opportunity to ask questions and reminded that their decision to participate was entirely voluntary.

Eligible participants who agreed to participate were asked to provide written informed consent. Participants were informed that their the decision to participate or withdraw from the study at any time would not have any adverse consequences on their employment status, and any data collected would be kept confidential and no identifiable results would be shared with the hospital staff or anyone else outside the research team.

2.7. Randomization, allocation concealment, and blinding

The randomization algorithm was computer generated and used to stratify the sample of trial participants using randomly sized blocks of 12. Randomization was performed by an independent trial statistician. Participants were informed of their allocation group via a text message/telephone call, and each participant was allocated a unique and anonymous identification number.

Participants and research members (trial statistician, outcome assessor, and study supervisors/coauthors) remained blind to participation allocation status. The three trainers were not blinded as this was not feasible because delivering the training required them to have in-depth knowledge and involvement in adapting the PFA training programme.

Several measures were taken to avoid contamination. These included: (1) participants were asked not to disclose their allocation status to others participating in the trial during the intervention period; (2) the intervention and control groups were kept separate during the delivery of the training and supervision sessions by assigning to two distanced classrooms and different group supervision channels; and (3) different trainers delivered the PFA and psychoeducation training sessions.

2.8. Intervention arm: the adapted READ-Y Chinese PFA training programme

The intervention arm received the culturally adapted READ-Y PFA training programme. This cultural adaptation of the PFA training was underpinning by an in-depth qualitative exploration of the cultural context of trauma exposure and coping in China. Chinese cultural norms differ significantly from the Western context, particularly in terms of emotional expression, views of crises, and collectivism. Compared to the original Western, disaster relief, and health contexts, three key cultural differences were incorporated into the PFA training: (1) Chinese people have a greater tendency to repress emotions and are not used to expressing signs of vulnerability; (2) Chinese people view crises as both dangers and opportunities for growth, leading to self-reliance over seeking external help for themselves; (3) and collectivist cultures encourage an ethos of self-sacrifice and prioritizing others’ well-being, often leading to self-neglect. In addition to adjusting for these cultural norms, the PFA adaptation ensured that the training used actual examples of natural disasters that had occurred in recent memory for Chinese HCWs, and the use of types of disasters (such as earthquakes and public health emergencies, rather than hurricanes) that would occur more commonly in China than in the USA.

These distinctions informed the selection of modifications which were made to culturally adapt the PFA training, originally developed in the USA, for use by Chinese front-line HCWs. For example, an additional self-care component ‘Yourself’ was added to the READ-Y PFA training package, as a means of emphasizing the need to focus upon the facilitation of self-healing and providing guidance for seeking help. For example, participants were encouraged to focus upon self-awareness and emotional expression, using word narrative games, animation, and role play in case scenarios.

To align with the five essential elements of trauma recovery principles, specific techniques and case scenarios were adapted specifically for the Chinese context. The PFA training programme was restructured by combining the second step of Assessment (conducted through listening to the survivor's story) with the third step of Prioritization (psychological triage) into one step called ‘Evaluation’. Combined with self-care, emphasized in the training curriculum, the PFA training programme was structured according to the renamed ‘READ-Y PFA’ programme: R (Rapport), E (Evaluation), A (Aid), D (Disposition), Y (Care for Yourself for others), adopting the Chinese name ‘基础心理支持培训’ (Basic Psychosocial Support Training, in English). The specific modifications of the original five-session, online RAPID PFA training course can be seen in Table 1 of the Supplementary file. In terms of delivery format, this adapted training programme followed a blended simulation approach with standard patients. It included a 1 day training divided into five 75 min, in-person group sessions, complemented by four online group supervision sessions. The training was facilitated by peer trainers rather than senior psychiatrists. The final structure of this culturally adapted PFA training programme is detailed in Table 2 in the Supplementary file.

Table 1. Recruitment and retention rates and feasibility of data collection tools.

Recruitment and retention	Number
Approach (within two weeks: 28 November to 8 December 2021)
Direct method: two gatekeepers disseminating project information via regular morning meetings	Nurses: 103 Doctors: 24
Indirect methods: posters, social media, consultation	Nurses: 84 Doctors: 4
Total: 215
Ineligible:	112
Out of hospital (other province)	12
Clinical ward managers/senior roles	20
Receive related training	34
Unavailable schedule (sick leave/routine duty)	36
Eligible participants	103
Refused consent	7
Consenting participants	96
Randomization	Intervention	Control
	48	48
Retention
In-person training sessions attended in a day	48	47
Online group supervision session attended (n); mean duration (min)	33	30
Group 1 (12 January/19 January 2022)	9 (65.45)	18 (75.1)
Group 2 (12 February 2022)	7 (72.3)
Group 3 (22 February/26 February 2022)	9 (61.23)	12 (68.6)
Group 4 (15 March 2022)	8 (59.78)
Time dedicated to practice (mean ± SD)
T1	2.89 ± 0.64	3.15 ± 0.60
T2	2.68 ± 0.49	3.13 ± 0.64
T3	2.44 ± 0.51	2.29 ± 0.57
Test after all sessions
T1	44	47
T2	44	47
T3	40	40
Feasibility of measurement tools
Time taken to fill in questionnaires (min), mean
T0 (Baseline)	13.91	15.75
T1	12.54	11.76
T2	11.65	9.11
T3	10.68	7.63
Missing data from questionnaires
T0 (Baseline)	4	1
T1	1	1
T2	2	1
T3	6	5

Table 2. Baseline characteristics of trial participants in the two arms.

	PFA training	Psychoeducation	T	p
Mean age (years)			0.150^b	.72
21–30	33 (76.9)	36 (73.2)
> 30	11 (23.1)	11 (26.8)
Gender			0.000^c	.21
Female	39 (89.7)	43 (90.2)
Male	5 (10.3)	4 (9.8)
Profession			0.000^c	.16
Nurse	38 (86.4)	43 (89.6)
Doctor	6 (13.6)	4 (10.4)
Education			0.803^b	.42
Undergraduate	31 (69.2)	36 (78.0)
Postgraduate	13 (30.8)	11 (22.0)
Marital status			0.230^b	.36
Married	27 (59.0)	25 (53.7)
Single	17 (41.0)	22 (46.3)
Professional experience (years)			1.091^d	.29
1–3	6 (15.4)	8 (14.6)
3–5	12 (25.6)	12 (22.0)
5–10	26 (59.0)	27 (61.0)
Professional title			1.009^c	.53
Resident	3 (7.7)	6 (14.6)
Junior	25 (59.0)	28 (56.1)
Senior	16 (33.3)	13 (29.3)
Weekly working hours			0.515^b	.28
< 35	10 (20.5)	15 (26.8)
35–44	27 (64.1)	27 (61.0)
> 44	7 (15.4)	5 (12.2)
Previous public emergency response experience			1.414^b	.42
No	38 (79.5)	35 (72.3)
Deal with acutely distressed patients			2.580^c	.06
Often	40 (94.9)	39 (80.5)
Sometimes	4 (5.1)	8 (19.5)
Any form of self-care training			4.490^d	.23
Seldom	39 (82)	39 (83)
Aware of mental health and well-being			1.373^d	.59
Seldom	31 (66.9)	26 (65.9)
Any form of trauma and stress training^a			0.001^d	.98
No	47 (98)	46 (97)
KSA	60.37 ± 18.89	65.17 ± 18.10	55.85^d	.85
IES-R	38.82 ± 15.33	35.57 ± 18.24	39.47^d	.14
DASS Depression	8.17 ± 7.68	8.30 ± 8.61	17.46^d	.10
DASS Anxiety	6.61 ± 7.79	6.00 ± 7.19	16.25^d	.37
DASS Stress	11.67 ± 10.57	10.90 ± 9.36	13.10^d	.06
BRS	3.22 ± 0.61	3.43 ± 0.58	31.61^d	.71
GSE	23.71 ± 5.91	25.71 ± 7.33	46.69^d	.15
SCSQ Positive coping	2.07 ± 0.43	2.11 ± 0.45	40.61^d	.51
SCSQ Negative coping	1.46 ± 0.56	1.38 ± 0.47	32.23^d	.50
ProQOL Burnout	25.55 ± 6.06	23.20 ± 4.85	43.93^d	.32
ProQOL Compassion satisfaction	33.97 ± 8.02	35.70 ± 7.23	43.91^d	.42
ProQOL Secondary traumatic stress	25.74 ± 5.86	23.49 ± 5.43	37.23^d	.83
PTGI	68.42 ± 14.80	68.02 ± 15.96	42.27^d	.79

Note: Data are shown as n (%) or mean ± SD.

^aParticipants had previously learned about mental health issues associated with trauma and stress during their routine training, i.e. 1 h lecture. However, they have not received any formal training related to the topic area.

^bPearson chi-square; ^ccontinuity correction; ^dFisher’s exact test.

PFA = psychological first aid; KSA = Knowledge, Skills, and Attitude; IES-R = Impact of Events Scale – Revised; DASS = Depression, Anxiety, and Stress Scale; BRS = Brief Resilience Scale; GSE = General Self-Efficacy Scale; SCSQ = Simplified Coping Style Questionnaire; ProQOL = Professional Quality of Life; PTGI = Post-Traumatic Growth Inventory.

2.8.1. Delivery

Four groups of 11–12 trainees received training at a mutually convenient hospital classroom on four consecutive days, on 16–19 December 2021. Online group sessions were provided on a web-based digital learning platform (https://www.xiaoetong.com), accessible through password-protected accounts. This contained short, text-based information regarding the key concepts of PFA skills, followed by time to reflect on the barriers and challenges to applying these PFA skills in front-line clinical practice.

In the initial study protocol (Wang et al., 2022), the plan was to conduct two follow-up group sessions after the training. However, the number of online group sessions was doubled to four and held at 3 week intervals because trainees from the regional COVID-19 response team had requested this, and also because the online nature of the sessions allowed for greater flexibility.

The two PFA trainers were mental health specialists, who worked in the psychiatry department and held master’s degrees in Clinical Psychology and Nursing. They both had more than 10 years of mental health and counselling experience and were sufficiently experienced to be able to deliver the PFA training and supervision sessions. Training entailed expert-led sessions on PFA, followed by online supervision to address any concerns and challenges encountered during applying the newly learnt PFA knowledge and skills in clinical practice.

2.9. Control arm: psychoeducation

Participants who were allocated to the control arm received psychoeducation sessions. These sessions consisted of the ‘usual or standard training’ that was regularly offered to front-line HCWs by the hospital. This consisted of a two-session psychoeducation training module, which involved two 75 min didactic lectures on understanding empathy and communication skills, and two post-training group discussion sessions. The psychoeducation training sessions were delivered to four groups of trainees over four consecutive days. Although the psychoeducation training was delivered by different trainers from those in the PFA sessions to avoid contamination, these trainers had a similar background to the PFA trainers. The trial also had a back-up trainer from the hospital’s emergency psychiatry department in case cover was needed.

2.10. Fidelity

The fidelity was monitored through recording all training and supervision sessions in the PFA intervention arm. Team supervision sessions were conducted once every 2 weeks in the field office, and selected audio recordings of the online supervision sessions were assessed by the first author. At the end of the supervision session, feedback on the overall session was discussed within research teams on the challenges and barriers reported during trainee supervision sessions, and a psychiatrist was invited to give professional supervision for all trainers.

2.11. Research support

Two locally based research assistants provided facilitation support for this study. One research assistant carried out data collection by disseminating a QR code and sending a reminder message to prompt trainees to complete questionnaires and attend training. The other research assistant facilitated the role-play during the PFA training session and online support supervision session. To ensure intervention fidelity, the research assistant facilitators had regular supervision from the lead researcher, who used a checklist and prompt-sheet to structure the sessions. During the process, both research assistants maintained a dual role of facilitator and observer, and took field notes.

2.12. Data collection procedures

Outcome data collection was conducted between 8 December 2021 and 30 March 2022. The outcome data were collected at four time-points: (1) before the training (T0); (2) 2 weeks post-training (T1); (3) 2 months post-training (T2); and (4) 3 months post training (T3) for both groups via encrypted, online survey software (Wenjuan Star). Participants were reminded to complete the follow-up assessments by WeChat messages, with a QR code to enter and fill in questionnaires. If they did not respond, they received a standardized scripted telephone call from a research assistant. Automated SMS or WeChat text message reminders were sent to participants every 3 days over the next 2 weeks or until the follow-up assessment was completed on the study website.

To minimize the risk of unmasking, the following procedures were followed: (1) the outcome assessments were carried out by an independent research assistant and disseminated QR code at the exact follow-up time; and (2) the intervention team and outcome evaluation team were based in two different physical department locations.

2.13. Feasibility and acceptability indicators (objective 1)

To determine the feasibility and acceptability of the adapted Chinese PFA training intervention in relation to response, recruitment, and retention rates, the following data were collected:

• Recruitment and retention: (1) number of eligible participants approached and consenting to take part and randomized; (2) number of participants who successfully complete training; and (3) number of participants attending online group sessions.

• Training attendance (adherence): (1) number of sessions attended; (2) time dedicated to practice; and (3) test after all sessions.

• Feasibility of measurement tools: (1) time taken to fill in questionnaires; and (2) missing data from questionnaires.

• Acceptability (usage and helpfulness):

  	-	PFA Training Follow-up Questionnaire: An adapted PFA training follow-up questionnaire from McCabe et al. (2014) was used, which consisted of two domains. The first domain included two self-reported items to collect trainees’ skill practice time: ‘Have you practised skills in dealing with public health emergencies during the past one month?’ and ‘Have you practised skills in clinical front-line setting during the past one month?’ The second domain included eight self-reported items of their views on the helpfulness of PFA; for example. using questions on whether their confidence in providing psychosocial support had increased.
  	-	Satisfaction survey: A self-reported, seven-item questionnaire was developed, with questions about their satisfaction with the training content and instructions. The satisfaction survey was rated on a five-point Likert scale, from 1 (Strongly satisfied) to 5 (Strongly dissatisfied).

2.14. Preliminary effectiveness outcome (objective 2)

To examine preliminary estimates of efficacy on the preparedness and related psychosocial measures, including behaviours, resilience, and professional quality of life of HCWs, the following measures were used:

• Preparedness:

  -

  Knowledge, Skills, and Attitude (KSA): An adapted PFA KSA survey form based on McCabe et al. (2014), was administered, which consisted of knowledge (seven items), skills (seven items), and attitudes (five items). In this study, the value of Cronbach’s alpha were α = .970, α = .957, and α = .702 respectively.

• Psychological measures:

  	-	Impact of Events Scale – Revised (IES-R): To determine the frequency of post-traumatic symptoms related to a recently experienced traumatic event, the IES-R was used, with a 22-item scale rated on a five-point Likert scale from 0 (not at all) to 4 (extremely) and total score ranging from 0 to 88, with higher scores indicating greater severity (Weiss, 2007). Cronbach's alpha for the questionnaire was α = .944. However, after pilot testing, this measure was only assessed at baseline owing to overlapping assessment of traumatic stress with two other measures (DASS-21 and ProQOL) described below.
  	-	21-Item Depression, Anxiety, and Stress Scale (DASS-21): The DASS-21 has 21 items and three subscales; namely, symptoms of depression, anxiety, and stress (Lovibond & Lovibond, 1995). Responses were scored on total scores multiplied by 2, with higher scores indicating greater severity. Cronbach's alpha values for each of the three subscales were: DASS-Depression α = .820, DASS-Anxiety α = .789, and DASS-Stress α = .798.
  	-	Brief Resilience Scale (BRS): The BRS was used to assess resilience, with six items scored on five-point Likert scales from 1 (totally disagree) to 5 (totally agree) (Fung, 2020). Total scores range from 1 to 5, with higher scores indicating a greater ability of recovery from psychological distress. Cronbach’s alpha was α = .970.
  	-	General Self-Efficacy Scale (GSE): The GSE was used to measure general self-efficacy, with 10 items scored on 4-point Likert scales from 1 to 4, with higher scores indicating a high level of self-efficacy (Chen et al., 2001). Cronbach’s alpha in this study was α = .938.
  	-	Simplified Coping Style Questionnaire (SCSQ): The SCSQ was used to assess the participants’ attitudes and actions that they would take in the face of life events. This 20-item self-report four-point Likert scale, from 0 (never) to 3 (very often), includes two dimensions: active coping (12 items) and passive coping (eight items), with greater scores indicating being more inclined to adopt the coping style (Xie, 1998). Cronbach’s alpha for the positive coping subscale was α = .825 and for negative coping was α = .802.
  	-	Professional Quality of Life (ProQOL): The ProQOL is a five-point Likert scale to assess three domains of perceived quality of life in relation to working as a helper, and includes both the positive and negative aspects of this work: compassion satisfaction (CS), burnout (BO), and secondary traumatic stress (STS), with the last two domains representing compassion fatigue (Shen et al., 2015). Total scores range from 10 to 50, with greater scores indicating higher levels on each subscale. Cronbach’s alpha values for the three subscales in this study were: ProQOL-CS α = .600, ProQOL-BO α = .930, and ProQOL-STS α = .804.
  	-	Post-Traumatic Growth Inventory (PTGI): The PTGI was used to assess positive responses to traumatic events (Tedeschi & Calhoun, 1996), with 21 items including five subscales with a Likert range of 0–5 (no transition at all to a lot of transition), with higher scores for post-traumatic growth indicating better growth. Cronbach’s alpha was α = .931.

At the baseline time-point, data on participants’ demographic characteristics (e.g. age, gender, education, and professional title) and work characteristics (e.g. years of working experience, previous experience of emergency response, frequency of supporting distressed patients, awareness of mental health and well-being and self-care, and trauma training) were collected.

Multiple outcome measures were selected in this feasibility trial, as one of the goals of a feasibility RCT is to identify appropriate tools and fine-tune hypotheses for the definitive RCT (Eldridge et al., 2016). A small pilot study to test the data collection procedures was conducted before conducting the main the feasibility trial. This pilot study was conducted with five front-line HCWs, who did not participant in the main feasibility trial.

2.15. Statistical analysis

The feasibility data are reported using CONsolidated Standards of Reporting Trials (CONSORT) guidelines (Moher et al., 2001), including a trial flowchart (Figure 1). Descriptive data in Table 1 are displayed with retention and mean time spent on the measurement tool to further indicate the procedure.

Figure 1. CONsolidated Standards of Reporting Trials (CONSORT) flowchart diagram.

To examine the preliminary effects, the primary analysis was intention-to-treat (ITT) analysis, adjusted for baseline mean scores. Among 96 participants who were allocated, one person was excluded because of being unavailable to attend training (working a night shift) and four participants were excluded from the final data analysis because they did not complete baseline assessments. The missing data were screened using listwise deletion. The baseline characteristics of participants in the two arms were compared using the chi-squared test for proportions and an independent sample t-test for continuous variables. The missing data at follow-up were imputed according to the maximum likelihood estimation option. As the latest missing data technique, this improves the accuracy and power of the analysis relative to other missing data handling methods (Schafer & Graham, 2002). The final data set for the ITT analysis included 91 participants, 44 in the PFA training intervention arm and 47 in the psychoeducation control arm. The pre-protocol analysis, in which the group comparison includes only those who completed the originally allocated treatment, was also conducted (Hernán & Robins, 2017).

For the outcome measures, normality distributions via the Shapiro–Wilk test and outliers were assessed. Since the outcomes data in this study presented a homogeneity variance and normal distribution, a repeated-measure analysis of variance (ANOVA) for within-group and between-group comparisons was performed. The between-subjects factor was group (PFA training intervention vs psychoeducation) and the within-subjects factor was time (baseline vs three follow-ups). A Bonferroni–Holm correction was applied to adjust for repeated comparisons. An estimate of the intervention effects, in terms of effect sizes, partial eta-squared (η²), is reported. The significance level was < 0.05 (two-sided). All analyses were carried out using SPSS version 28 (IBM Corp., Armonk, NY, USA).

3. Results

3.1. Feasibility: recruitment

In total, 215 front-line HCWs contacted the research team expressing an interest in participating in the trial, resulting in the targeted sample size being exceeded. Of the 215 participants, 103 workers heard about the trial from a clinical ward manager, and 112 workers heard about the trial via social media, within only 1 week.

The 215 interested participants represented a range of specialities, including internal medicine, surgery, emergency medicine, oncology, geriatrics, and ophthalmology, and were from 87 different clinical wards (68.5% of 135 wards) in the hospital. During the screening stage, 103 were assessed as eligible and willing to participate, with a 48% recruitment rate from those who expressed an interest in participating. Over half of the potential participants (n = 112, 52%) were assessed as ineligible, for the following reasons: being from other geographical provinces (n = 12, 10.7%), having senior management roles that cannot prioritize providing direct care (n = 20, 17.8%), having previously completed mental health training (n = 34, 30.3%), and not being available to participate in the training owing to prior commitments (n = 36, 32%) (Table 1).

Of the 103 eligible participants, seven people (6.7%) refused to sign the consent form, as they reported that they could not accept the possibility of being allocated to the control arm. This resulted in 96 participants (93.2%) giving consent and being enrolled into the trial, with 48 participants being randomly allocated to the PFA intervention arm and 48 participants being randomly allocated to the control arm. Figure 1 presents the participant flow in a CONSORT flow diagram.

3.1.1. Sample characteristics

Most of the 96 participants were aged < 30 years, married, and female (82 female and nine male); nine participants were male workers and mostly doctors with postgraduate degrees. More than half of the sample had 5–10 years of professional working experience, worked for 35–44 h a week, and had no previous public emergency response experience, yet frequently had to support acutely distressed patients. Most of them had received no training in self-care awareness, trauma, or stress. As presented in Table 3, there were no significant differences in participants’ baseline characteristics between the two groups or mean outcome measure scores. No harmful or adverse events were recorded for any participants throughout the duration of this feasibility study.

Table 3. Outcome mean changes from baseline to the three follow-ups.

Outcome	PFA training				Psychoeducation
	T0 (n = 44)	T1 (n = 43)	T2 (n = 42)	T3 (n = 38)	T0 (n = 47)	T1 (n = 46)	T2 (n = 46)	T3 (n = 42)
KSA	57.47 ± 17.35	77.56 ± 9.80	77.64 ± 10.01	79.25 ± 9.65	65.10 ± 17.88	74.23 ± 10.03	73.21 ± 10.51	75.23 ± 11.50
DASS Depression	8.17 ± 7.68	6.06 ± 6.18	4.61 ± 5.29	4.39 ± 6.41	8.30 ± 8.61	5.90 ± 7.30	7.50 ± 11.78	8.60 ± 10.18
DASS Anxiety	6.61 ± 7.79	3.89 ± 5.01	3.33 ± 5.19	4.50 ± 5.90	6.00 ± 7.19	4.25 ± 5.55	4.90 ± 8.07	6.20 ± 11.20
DASS Stress	11.67 ± 10.57	8.11 ± 8.47	6.67 ± 8.55	5.78 ± 7.00	10.90 ± 9.36	6.30 ± 7.89	7.20 ± 9.33	8.35 ± 11.93
BRS	3.22 ± 0.61	3.45 ± 0.38	3.57 ± 0.46	3.59 ± 0.5	3.43 ± 0.58	3.58 ± 0.49	3.52 ± 0.55	3.50 ± 0.57
GSE	23.71 ± 5.91	24.95 ± 4.13	26.32 ± 5.54	25.84 ± 5.50	25.71 ± 7.33	25.62 ± 6.20	26.62 ± 7.22	25.82 ± 6.48
SCSQ Positive coping	2.07 ± 0.43	2.14 ± 0.46	2.22 ± 0.51	2.22 ± 0.48	2.11 ± 0.45	2.29 ± 0.60	2.14 ± 0.46	2.19 ± 0.61
SCSQ Negative coping	1.53 ± 0.55	1.37 ± 0.49	1.37 ± 0.46	1.49 ± 0.57	1.31 ± 0.45	1.28 ± 0.58	1.34 ± 0.49	1.37 ± 0.60
ProQOL Burnout	25.55 ± 6.06	23.61 ± 5.34	22.57 ± 5.69	22.67 ± 5.17	23.20 ± 4.85	23.66 ± 5.08	23.06 ± 5.22	23.51 ± 5.47
ProQOL Compassion satisfaction	33.97 ± 8.02	36.11 ± 6.62	36.52 ± 6.49	37.93 ± 5.90	35.70 ± 7.23	35.90 ± 6.42	36.00 ± 6.42	36.61 ± 6.85
ProQOL Secondary traumatic stress	25.74 ± 5.86	24.00 ± 5.67	23.59 ± 5.38	25.65 ± 5.41	23.49 ± 5.43	24.31 ± 6.57	24.19 ± 5.97	25.08 ± 5.11
PTGI	68.42 ± 14.80	74.61 ± 11.34	77.43 ± 13.23	77.53 ± 11.21	68.02 ± 15.96	75.38 ± 14.37	75.09 ± 14.86	75.65 ± 15.90

Note: Data are shown as mean ± SD.

PFA = psychological first aid; KSA = Knowledge, Skills, and Attitude; DASS = Depression, Anxiety, and Stress Scale; BRS = Brief Resilience Scale; GSE = General Self-Efficacy Scale; SCSQ = Simplified Coping Style Questionnaire; ProQOL = Professional Quality of Life; PTGI = Post-Traumatic Growth Inventory.

3.2. Feasibility: retention rates

Retention rates are reported for both the in-person training sessions and the online follow up sessions. Of those allocated to the in-person training, 48 individuals (100%) completed the in-person PFA training session, while 47 (97.1%) attended control training (one person withdrew because of sickness). Of the 95 participants who attended the in-person training, all of them (100%) completed training, showing a high retention rate for in-person training, as seen in Table 2.

Regarding the online group supervision sessions attended, a moderate retention rate was obtained, with 33 individuals (68.75%) attending in the PFA training arm and 30 individuals (63.82%) in the control arm. The mean duration time for online supervision sessions was 71 min; however, an average 30% of trainees left in the middle of each session owing to a bad network connection, urgent duty call, or family issues.

The baseline outcome measures were completed by 44 individuals (91.6%) in the PFA training arm, taking an average of 13.91 min to fill in the assessments. Four individuals (8.3%) completed the PFA training but did not complete the baseline assessment (owing to forgetting to scan the questionnaire QR code). Analyses found that participants who did not complete the follow-up questionnaires (n = 4) did not significantly differ from completers (n = 91) in baseline outcome measures. The follow-up questionnaires were completed at three time-points, as follows: 89 (93.6%) filled in the T1 follow-up, 88 (92.6%) filled in the T2 follow-up, and 80 (84.2%) filled in the T3 follow-up. At the end of the follow-up, six individuals (12.5%) in the PFA training arm did not complete the assessments.

3.3. Acceptability: usage, helpfulness, and satisfaction with the PFA training

Usage: Overall, just under 50% of the trainees applied PFA skills either in disaster or in their routine clinical care settings, indicating a low usage of PFA skills (Table 2). Although usage decreased over time, there was a slightly higher rate of participants reporting that they had used their PFA skills in routine front-line clinical care settings. For usage in public health emergencies, 6% of PFA trainees reported that they had applied their PFA skills more than three times, 46% of PFA trainees had applied their skills to practice once or twice, and 48% of them had never used their PFA skills; whereas for usage in the front-line clinic care setting, 8% of PFA trainees had applied their PFA skills more than three times, 34% of them once or twice, and 58% had never used them.

Helpfulness: The participants, overall, found the PFA training to be helpful, with an average score of 22 out of 24 (SD = 2.02), being helpful in two areas: (1) becoming better listeners (86% of participants agreed or strongly agreed), and (2) increasing their empathy (84% of participants agreed or strongly agreed). Moreover 72–78% of participants agreed or strongly agreed that they had been more willing to support acutely distressed individuals, more capable of establishing rapport with others, more actively engaged with the emergency response, and more confident in providing basic psychological support. Only 65% of them agreed or strongly agreed on being able to make referrals and being able to differentiate between normal and abnormal behavioural reactions, and 34% of them reported that they were unclear or not confident enough to make referrals or distinguish between different behavioural reactions.

Satisfaction: In total, 35 of 48 PFA trainees (74%) strongly agreed that they were satisfied with the training content, instructional methods, and simulation scenario; whereas they rated a lower satisfaction with the online group sessions.

3.4. Fidelity assessment

The PFA training was completed in five sessions within a day, with only two trainers and no specialized equipment needed other than access to online materials, PowerPoint, and simulation exercises. The mean duration of each single session ranged from 70 to 85 min. The facilitator observed how the training was delivered and marked a high level of adherence to the standard training manual (see Table 3 in the Supplementary file).

3.5. Mean changes to outcome variables

Table 3 shows the mean scores with standard deviations for completers in both arms at baseline and the three follow-ups. To show the statistical differences in these mean outcome changes for both arms compared to baseline, a further outcome comparison assessed mean differences based on pairwise comparison in repeated-measures analysis of variance (ANOVA) according to both ITT and the completer analysis (see Table 4 in the supplementary file).

Table 4. Preliminary effects of the adapted psychological first aid (PFA) training intervention (versus psychoeducation).

Measure	Group	T0	T1	T2	T3	F time	p	$\eta p^2$	F group	p	$\eta p^2$	F group*time	p	$\eta p^2$
KSA
	PFA training	57.47 ± 17.35	77.56 ± 9.80^a	77.64 ± 10.01^a	79.25 ± 9.65^a	33.750***	.001	.275	0.900	.345	.010	6.378*	.038	.075
	Psycho-education	65.10 ± 17.88	74.23 ± 10.03^a	73.21 ± 10.51	75.23 ± 11.50^a
DASS
Depression	PFA training	8.17 ± 7.68	6.06 ± 6.18^a	4.61 ± 5.29^a	4.39 ± 6.41^a	10.018***	.001	.101	0.166	.684	.002	2.874*	.046	.031
	Psycho-education	8.30 ± 8.61	5.90 ± 7.30	7.50 ± 11.78	8.60 ± 10.18
Anxiety	PFA training	6.61 ± 7.79	3.89 ± 5.01^a	3.33 ± 5.19^a	4.50 ± 5.90	6.379***	.001	.067	0.273	.603	.003	1.453	.235	.016
	Psycho-education	6.00 ± 7.19	4.25 ± 5.55	4.90 ± 8.07	6.20 ± 11.20
Stress	PFA training	11.67 ± 10.57	8.11 ± 8.47^a	6.67 ± 8.55^a	5.78 ± 7.00^a	11.549***	.001	.115	0.183	.670	.002	1.995	.123	.022
	Psycho-education	10.90 ± 9.36	6.30 ± 7.89^a	7.20 ± 9.33^a	8.35 ± 11.93
BRS
	PFA training	3.22 ± 0.61	3.45 ± 0.38^a	3.57 ± 0.46^a	3.59 ± 0.5	5.054**	.004	.054	0.972	.327	.011	1.775	.164	.020
	Psycho-education	3.43 ± 0.58	3.58 ± 0.49	3.52 ± 0.55	3.50 ± 0.57
GSE
	PFA training	23.71 ± 5.91	24.95 ± 4.13	26.32 ± 5.54	25.84 ± 5.50	3.091*	.035	.034	0.487	.487	.005	1.093	.349	.012
	Psycho-education	25.71 ± 7.33	25.62 ± 6.20	26.62 ± 7.22	25.82 ± 6.48
SCSQ Positive coping
	PFA training	2.07 ± 0.43	2.14 ± 0.46	2.22 ± 0.51	2.22 ± 0.48	2.318	.083	.025	0.064	.801	.001	1.662	.181	.018
	Psycho-education	2.11 ± 0.45	2.29 ± 0.60^a	2.14 ± 0.46	2.19 ± 0.61
ProQOL
Burnout	PFA training	25.55 ± 6.06	23.61 ± 5.34	22.57 ± 5.69^a	22.67 ± 5.17^a	3.402*	.028	.037	0.069	.793	.001	3.729*	.018	.037
	Psycho-education	23.20 ± 4.85	23.66 ± 5.08	23.06 ± 5.22	23.51 ± 5.47
Compassion satisfaction	PFA training	33.97 ± 8.02	36.11 ± 6.62^a	36.52 ± 6.49^a	37.93 ± 5.90^a	5.627**	.002	.059	0.005	.944	.000	2.345	.086	.026
	Psycho-education	35.70 ± 7.23	35.90 ± 6.42	36.00 ± 6.42	36.61 ± 6.85
Secondary traumatic stress	PFA training	25.74 ± 5.86	24.00 ± 5.67	23.59 ± 5.38	25.65 ± 5.41	2.306	.077	.025	0.260	.611	.003	2.282	.080	.025
	Psycho-education	23.49 ± 5.43	24.31 ± 6.57	24.19 ± 5.97	25.08 ± 5.11
PTGI
	PFA training	68.42 ± 14.80	74.61 ± 11.34^a	77.43 ± 13.23^a	77.53 ± 11.21^a	15.554***	.001	.149	0.161	.689	.002	0.508	.657	.006
	Psycho-education	68.02 ± 15.96	75.38 ± 14.37^a	75.09 ± 14.86^a	75.65 ± 15.90^a

Note: Data are shown as mean ± SD.

The statistical test used is repeated measures analysis of variance. Adjustment for multiple comparisons: Bonferroni. Mauchly’s test p < .01. ^a Compared to baseline, p < 0.05. *p < .05; **p < .01; ***p < .001.

All results are reported using Greenhouse–Geisser results as Mauchly’s test p < .01, except for ProQOL secondary traumatic stress, reported as multivariate test results due to p = .107 in Mauchly’s test.

KSA = Knowledge, Skills, and Attitude; DASS = Depression, Anxiety, and Stress Scale; BRS = Brief Resilience Scale; GSE = General Self-Efficacy Scale; SCSQ = Simplified Coping Style Questionnaire; ProQOL = Professional Quality of Life; PTGI = Post-Traumatic Growth Inventory.

3.5.1. KSA measure

Regarding changes to KSA, both arms achieved better KSA scores post-training. For the PFA training arm, the mean KSA score increased significantly at T1 [p < .001, 95% confidence interval (CI) −23.93 to −9.50], T2 (p < .001, 95% CI −23.74 to −8.46), and T3 (p < .001, 95% CI −23.75 to −8.46), indicating a significantly increased KSA mean score over time. In contrast, in the control arm, the mean KSA score increased slightly at the three follow-ups, with statistical significance at T1 and T3, yet no significant difference at T2 (p = .06, 95% CI −14.629 to 0.153).

3.5.2. Psychological measures

There were significant psychological outcome changes for the PFA training arm. First, compared to baseline, the PFA training arm produced a mean score decrease on measures of depression, anxiety, stress, and burnout, with statistically significant changes over time; however, no statistically significant decline was found on the measure of secondary traumatic stress. Secondly, mean scores on resilience, self-efficacy, compassion satisfaction, and post-traumatic growth increased over time compared to baseline, with statistically significant differences found for all variables. These findings suggest that PFA trainees experienced positive improvements in resilience, compassion satisfaction, and post-traumatic growth, as well as reductions in depression, anxiety, stress, and burnout over time.

In the control arm, compared to baseline, mean score increases on only positive coping and post-traumatic growth were observed after receiving psychoeducation. Despite obtaining lower scores on depression and anxiety, and higher scores on resilience, self-efficacy, positive coping, and compassion satisfaction, no statistically significant differences were found.

3.6. Preliminary effect of the intervention versus control

Both ITT and the completer analyses were performed to detect any differences in preliminary effects between the two arms, and similar findings were found (Table 4). Regarding the preliminary effect of the PFA training relative to psychoeducation, when controlling for baseline differences, a significant main effect of time was found for most variables except for SCSQ positive coping and ProQOL secondary traumatic stress; however, no main effect of group was found for any variables. Three group-by-time interaction effects were found for KSA, DASS depression, and ProQOL burnout.

Regarding these significant main effects of time, a large time effect size was found for KSA (F_1,149 = 33.75, p < .001, $\eta p^2$ = .27), DASS stress (F_3,237 = 12.88, p < .001, $\eta p^2$ = .14), and PTGI (F_2,204 = 15.55, p < .001, $\eta p^2$ = .15). A medium time effect size was found for DASS depression (F_2,189 = 10.66, p < .001, $\eta p^2$ = .12), DASS anxiety (F_2,200 = 7.23, p < .001, $\eta p^2$ = .08), ProQOL (F_2,189 = 5.53, p = .003, $\eta p^2$ = .06), BRS (F_2,192 = 5.16, p = .002, $\eta p^2$ = .06), GSE (F_2,205 = 3.54, p = .02, $\eta p^2$ = .04), and ProQOL compassion satisfaction (F_2,198 = 8.17, p < .001, $\eta p^2$ = .09).

No significant main effects of group were found between two groups for any outcome variables. However, significant time-by-group interaction (group*time) effects were found for three variables (KSA, DASS depression, and ProQOL burnout); therefore, a single effect analysis was performed to examine the difference between groups.

Regarding the KSA, a moderate interaction effect size was found (F_1,137 = 6.37, p < .001, $\eta p^2$ = .07). As shown in Figure 2, KSA improved more in the PFA training arm than in the control arm at T1 (mean difference=−7.632, p = .05, 95% CI −15.436 to 0.171) and T3 (mean difference = 4.427, p = 0.05, 95% CI −0.13 to 8.98), suggesting a better KSA improvement after receiving the PFA training.

Figure 2. Interaction effect for mean Knowledge, Skills, and Attitude (KSA) scores from baseline to the three follow-ups, by arm.

Regarding DASS depression, a moderate interaction effect size was found (F_2,232 = 2.874, $\eta p^2$ = .031, p = .046). As shown in Figure 3, DASS depression in the PFA training arm decreased over time at T1 (mean difference = 7.632, p = .05, 95% CI −15.436 to 0.171) and T3 (mean difference = 4.427, p = .05, 95% CI −0.13 to 8.98), whereas the control group experienced a decline and then a slight increase. Also, there was a significant group difference at T2 (mean difference = 4.43, p < .04, 95% CI 0.05 to 8.82). These results indicate that the PFA training arm experienced a greater reduction in depression over time.

Figure 3. Interaction effect for mean 21-Item Depression, Anxiety, and Stress Scale (DASS-21) depression scores from baseline to the three follow-ups, by arm.

Regarding ProQOL burnout, a significant medium interaction effect size was discovered (F_2,211 = 3.729, $\eta p^2$ = .037, p = .018). As shown in Figure 4, the mean score for burnout in the PFA training arm decreased sharply over time, whereas the control arm fluctuated during 3 months, suggesting that the PFA training arm experienced a significant decrease in burnout compared with the control arm.

Figure 4. Interaction effect for mean Professional Quality of Life (ProQOL) burnout scores from baseline to the three follow-ups, by arm.

4. Discussion

This feasibility trial examined the feasibility and acceptability of delivering a culturally adapted PFA training intervention to front-line HCWs who work clinically, providing direct patient care, and are part of a public health/disaster emergency response team in China.

4.1. A feasible and acceptable PFA training intervention for Chinese front-line HCWs

In relation to the first study objective, the findings from this feasibility trial show the feasibility and acceptability of delivering culturally adapted PFA training to HCWs who are part of a public health/disaster emergency response team. Several key indicators support this. First, a rapid recruitment process that exceeded the target sample size signifies a widespread interest among front-line HCWs and the wider hospital organization in providing PFA training. This is also due in part to the relevance of PFA training, especially following the COVID-19 lockdown in China. As a tertiary hospital designated for emergency response, the COVID-19 pandemic heightened awareness of the importance of emergency response preparedness, including mental health preparedness, both at the organizational level and among individual HCWs, which contributed towards the positive reception of the PFA training, delivered in December 2021 to March 2022. This also reflects a broad global trend towards recognizing the importance of developing early prevention strategies for at-risk occupational groups and trauma-exposed organizations (Chen et al., 2020; Malik et al., 2021; Chandler et al., 2022; Francis et al., 2020).

Secondly, the high completion rates for both face-to-face training and online sessions over a 3 month period are noteworthy. These are in contrast to previous studies on PFA training among healthcare professionals, which often reported high dropout rates or had limited longitudinal follow-up assessments (Blake et al., 2021). Over half of the trainees (6% + 46%) had used their PFA skills in responding to a public health emergency 3 months after training. In addition, the positive feedback and high satisfaction ratings received for this PFA training, which includes a blended learning approach, case simulations, and follow-up supervision, reinforce the necessity for practical, real-world, applicable training. These high levels of attendance and satisfaction can be attributed to the cultural adaptations that were integrated in this approach, addressing the specific needs and preferences of HCWs, and making the training more engaging and relevant to their working conditions. It is recognized by previous studies that health professionals tend to favour blended learning for their professional development, as it allows for continuous learning and a more comprehensive understanding of knowledge and skills, and has a positive impact on attitudes and behaviour (Byungura et al., 2022). Despite these clear benefits, efforts to implement blended learning for PFA training programmes have often been hampered by funding constraints and limited resources.

Thirdly, using a trial design for evaluating the PFA training intervention in a healthcare setting had no reported harmful or adverse effects and was well received. This is evident through the close adherence to the adapted training manual, the absence of significant feasibility issues during delivery, and the completion of online evaluation surveys, featuring multiple measures at various time-points. This adds valuable implementation insights for PFA training in the context of front-line care and Eastern collectivist cultures. Given the increasing emphasis on enhancing PFA training delivery, this implementation evidence represents a practical, much-needed, and promising option to address the pressing needs of HCWs and organizations (Wang et al., 2021).

4.2. Preliminary estimate effect of this PFA training intervention

In relation to the second study objective, the results suggest that this adapted PFA training has preliminary positive effects on knowledge and skills acquisition, as well as measures of participant mental health and psychological well-being.

The study yielded positive results regarding PFA knowledge, skills, attitude, self-efficacy, and resilience. These are the primary outcome of interests for PFA training initiatives, and the effect of PFA training is well established, as evidenced by previous research on its positive impact on knowledge acquisition, resilience, and self-efficacy among primary HCWs (Sijbrandij et al., 2020), vocational nursing students (Zhang et al., 2022), social practitioners (Lalani & Drolet, 2018), and nursing home staff (Schoultz et al., 2022). In addition to this, the improvements in compassion satisfaction and post-traumatic growth are promising, adding to our knowledge of the potential benefits of PFA training. These findings suggest that PFA training may serve as an alternative approach to address compassion fatigue, a concerning issue that is being extensively researched among HCWs (Nolte et al., 2017).

The study also investigated the impact of PFA training on anxiety, stress, depression, and burnout, and demonstrated positive changes. This result is encouraging for a potential reduction in these outcomes being sustained over a 3 month period. This is a novel finding since, compared to the literature that exists on PFA training for reducing anxiety and stress, its efficacy in reducing depressive symptoms and burnout remains underexplored. Furthermore, compared to the psychoeducation currently implemented in front-line care settings, PFA training shows greater efficacy in reducing symptoms of depression and burnout. These findings need to be replicated in a further evaluation in a large-scale trial. Given the high prevalence of burnout among health professionals globally and the mixed results yielded from current burnout intervention efforts (Costa & Pinto, 2017; West et al., 2016), this finding adds empirical evidence that echoes the anecdotal evidence on educating junior doctors in PFA to enable them to manage acute stress (Chivers, 2023).

There are several potential explanations for these findings. First, PFA training shares similarities with wellness training programmes, which have been researched as effective interventions for reducing burnout among healthcare professionals (Lu & Ratnapalan, 2023). Both types of training involve longitudinal support and the acquisition of skills related to resilience, stress-coping mechanisms, and self-care. Secondly, substantial evidence demonstrates that burnout and well-being are interconnected constructs, and prevention interventions addressing both personal and workplace factors are needed (Prentice et al., 2023). PFA training empowers participants by enhancing their problem-solving and stress-management skills, enabling them to cope with challenging work environments. This improves their personal well-being and enhances patient care by fulfilling professional values, a key factor in preventing burnout (Hoffman & Bonney, 2018). Thirdly, the inclusion of online group sessions and peer facilitators in the PFA training programme enhances peer support and may reinforce its potential advantages, as peer support and collective benefits have been demonstrated in mindfulness communication programmes (Hoffman & Bonney, 2018). Lastly, previous studies among emergency workers have highlighted the benefits of PFA in reducing stigma and overcoming barriers to seeking help (Tessier et al., 2022). Receiving PFA training may encourage healthcare professionals to proactively recognize and address trauma and stress-related issues, thereby better managing their own well-being within the demanding medical culture, which often emphasizes perfectionism, denial of personal vulnerability, and delayed gratification (Hoffman & Bonney, 2018).

To gain a deeper understanding of the mechanisms behind PFA training and its impact, future research could consider trainees’ experiences and perceived valued outcomes. In addition, investigating implementation factors to determine the optimal intensity and duration of PFA training, as well as the role of supervision group sessions, is essential. While previous studies have suggested a need for more intensive support, such as supervision to achieve these results, limited documentation exists in this regard (Wang et al., 2021).

4.3. Future modification

Several modifications to the trial design and training intervention are warranted for future research studies. First, some participants who would have liked to take part were unable to because of their geographic location. Given the continued interest expressed by study participants, adjustments to the timing and structure of the programme may be necessary, allowing for greater flexibility. Balancing the accessibility and potential impact of PFA training is crucial. Combining the structured and guided nature of face-to-face instruction with the flexibility and self-directed aspects of online learning may create a comprehensive and well-rounded training solution. This approach is important because relying solely on either in-person or online training may have limitations, such as gaps in curriculum content or mismatches with trainees’ learning skills and expectations in relation to the learning environment (Everly Jr & Lating, 2022). However it is also important to understand the current suitability of the content, and the mix of supervision, group learning formats, intensity, and duration, to maximize the sustainability of the delivery of these PFA training programmes.

The sample for this trial was predominantly female nurses, while only a small proportion (10%) of male doctors participated, indicating an underrepresentation of this specific group. Male doctors are often considered a hidden or silent group that is challenging to reach, with high levels of burnout, and greater efforts should be made to address the feasibility challenges associated with engaging male doctors in such training programmes (Weiss, 2007). This professional group could potentially be engaged through incorporating the training into doctors’ residential training schemes, clearly outlining the expected outcomes of the interventions, and increasing the flexibility and availability of supervision timeslots to attract male doctors and facilitate their participation in the training.

Furthermore, there was a relatively low attendance for online group sessions. This could be addressed by offering shorter follow-up group sessions within certain time frames to motivate participants to attend. Moreover, the study reported a low rate of confidence in making referrals and differentiating between normal and abnormal behavioural reactions, implying a need for reinforcement in the assessment and psychological triage aspects of the training. In addition, relying solely on self-reported practice changes may not be sufficient, as it provides limited insight into potential harms and lacks objective assessment in real practice (Everly Jr & Lating, 2022). Future efforts to thoroughly examine skill application with observation in real practice is essential to ensure that PFA is delivered adequately and safely in real-world healthcare settings, thus contributing towards the ongoing improvement and refinement of PFA training to maximize its overall effectiveness and impact.

4.4. Strengths and limitations

This study has several strengths. First, it employs a robust study design, being the first rigorous evaluation using an RCT to assess the feasibility, acceptability, and preliminary effectiveness of a culturally adapted PFA training intervention among HCWs in front-line care settings. Secondly, the inclusion of an active control group ensures that the results are relevant and practical within routine care contexts. Thirdly, the study sample included healthcare professionals from various specialities, including internal medicine, emergency care, critical care, and geriatric medicine, working in general hospitals, as well as involving emergency response teams. This sample diversity increases the generalizability of the findings beyond samples primarily composed of disaster rescue workers, if replicated in a larger scale clinical trial. Moreover, the multiple outcome measures with repeated evaluations for 3 months suggest a broad outcome of interest to be selected for PFA training in any future work. Lastly, the level of detail reported on the training content and implementation strategy for this adapted PFA training programme is relatively unusual in the PFA literature.

However, certain limitations are also evident. First, while the feasibility trial findings show promise, they should be interpreted cautiously as the sample size is not adequately powered to draw definitive conclusions. Secondly, the reliance on self-report measures for outcome evaluation and skill usage introduces a potential response bias. Thirdly, the follow-up period for outcome assessment is limited in its ability to reflect longer term and organizational outcomes among PFA trainees, which could be valuable for future organizational uptake. Lastly, certain pragmatic design choices, such as the use of a less structured psychoeducational approach in the control arm without an assessment for provider fidelity, introduce variability to the study procedures.

5. Conclusion

The culturally adapted PFA training intervention tailored to Chinese HCWs indicated a high level of acceptability and feasibility for implementation in front-line care settings. This led to positive changes with improvements in PFA knowledge, skills, and attitudes, resilience, self-efficacy, compassion satisfaction, and post-traumatic growth, as well as a greater reduction in symptoms of depression and burnout, compared to psychoeducation. Future evaluation is warranted to validate these promising results.

Open Scholarship

This article has earned the Center for Open Science badges for Open Data, Open Materials and Preregistered. The data and materials are openly accessible at 1) Chinese clinical trial registration website: www.medresman.org; 2) trial protocol: doi.org/10.3389/fpsyt.2021.809679 and 3) trial protocol: doi.org/10.3389/fpsyt.2021.809679.

Trial registration

This trial has been approved by the Institution Review Board from Central South University (XTXL20200610) and by the Psychiatry, Nursing and Midwifery Research Ethics Committee at King’s College London, UK (LRS/DP-21/22-23161). It has also been processed for registration at the Chinese Clinical Trial Registry (ChiCTR2300071402).

Author contributions

I. N., M. L., and L. W. were responsible for the research concept development, methodology, and trial registration. L. W. drafted the manuscript. I. N. and M. L. contributed to the text and critically revised the manuscript. Y. L and T. X. obtained part of the funding and support from the local authorities, and undertook hospital population engagement. X. L., C. J., and Z. Z. contributed to recruitment, data collection, and research facilitation. T. L., J. W., and L. Z. contributed to training development, delivery, and supervision. All authors approved the final version of the manuscript.

Acknowledgements

The authors would like to express their gratitude to all healthcare professionals and their clinical ward managers on the front-line who took part in this study and generously shared their valuable experiences. The authors also appreciate the invaluable administrative support of the Nursing Department, Medical Department, Academic Affairs Department, Clinical Skill Training Center, Medical Simulation Center, and Department of Psychiatry at the Second Xiangya Hospital.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

All materials (data, code, and supporting information) are available from ResMan (www.medresman.org), archived at the time of submission on ResMan (DOI: 10.5281/zenodo.3968301).

Additional information

Funding

This study is supported by the Health Research Project from Health Commission of Hunan Province [20190365] and the innovative education project of Central South University [2018CXKZ06].

Notes

1 For example, during the COVID-19 pandemic response, more than 10 teams with 600 front-line workers from different disciplines were called out and dispatched by the national and provincial health commission to rescue Wuhan and other provinces.