Single-session intervention with and without video support to prevent the worsening of emotional distress among healthcare workers during the SARS-CoV-2 pandemic: a randomised clinical trial

Abstract

Background

The SARS-CoV-2 pandemic was a major stressful event that significantly affected healthcare providers.

Objective

To evaluate the efficacy of a single-session intervention (SSI), with and without weekly personalised prerecorded videos, in preventing emotional distress worsening.

Methods

Nationwide randomised clinical trial conducted in Brazil from 19 May 2020 to 31 December 2021. We included healthcare professionals with anxiety, depression or irritability scores below a T-score of 70, measured by the Patient-Reported Outcomes Measurement Information System. Participants were randomised to SSI or SSI plus weekly personalised prerecorded videos for 4 weeks (SSI-ET). The primary outcome was the proportion of participants with a significant escalation of symptoms (ie, T-score above 70).

Findings

From 3328 assessed for eligibility, 1112 participants were randomised (SSI=549; SSI-ET=563). The cumulative proportion of incident cases at 6 months was 17.5% (95% CI: 13.1% to 21.6%) for SSI and 15% (95% CI: 10.9% to 18.8%) for SSI-ET, with no difference between groups (HR=0.81, 95% CI: 0.83 to 1.79). Both groups showed significant symptom reductions at 1 month follow-up, maintained at 3 and 6 months (Cohen’s d=0.68–1.08).

Conclusions

Adding asynchronous videos to SSI provided no additional benefit. While symptom improvements occurred over time, the lack of a control group limits conclusions about intervention efficacy.

Clinical implications

Given its simplicity and scalability, the synchronous component, which includes core mental health support elements, such as active listening, validation, and empathic care, may represent a feasible way to assist healthcare professionals in crisis settings, even though further research is needed to determine its specific impact.

Trial registration number

NCT04632082.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• The SARS-CoV-2 pandemic significantly affected the mental health of healthcare workers due to high infection rates, increased workload and psychological demands of patient care, highlighting the necessity for preventive measures to address increasing emotional distress.

• There is no previous evidence concerning the efficacy of brief psychological interventions on preventing high levels of emotional distress, as measured by anxiety, depression or irritability, on professionals exposed to crisis situations.

WHAT THIS STUDY ADDS

• In this nationwide randomised clinical trial, we found no benefit of adding asynchronous personalised video components of psychoeducation. Although a substantial symptom improvement occurred over the 6-month follow-up, the absence of a control group prevents attributing this improvement directly to the synchronous components. Accordingly, further controlled research is necessary to determine whether these elements may effectively prevent symptom worsening.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The implications of effective single-session interventions extend beyond the immediate pandemic context, as they could change how policymakers approach psychological support for health professionals during crisis contexts by allowing more accessible preventive models.

• Brief, transdiagnostic and less specialised interventions may be promising public health measures to be further investigated in future studies, considering their scalability and cost-effectiveness, factors with critical implications, especially in settings with limited resources.

Background

The SARS-CoV-2 pandemic has profoundly impacted the mental health of healthcare workers,¹ underscoring the need for preventive interventions. Brazil is among the countries that have been severely affected by the COVID-19 pandemic, with a staggering 38.7 million cases and a mortality rate of 338 deaths per 100 000 inhabitants during the pandemic. The total number of COVID-19 deaths in Brazil has surpassed 711 000, making it the country with the second-highest death toll worldwide.² Frontline healthcare workers had faced higher risks of adverse mental health outcomes due to increased infection rates, workload and the psychological strain of patient care.¹ Therefore, early preventive strategies were crucial for mitigating the immediate effects of stress and preventing symptom escalation.^{3 4}

During crisis situations, individuals often exhibit symptoms of emotional distress, such as depression, anxiety and irritability; however, most of them neither develop psychiatric syndromes nor require long-term treatment.⁵ Additionally, psychological debriefing, once a common intervention, has been critically re-evaluated and recent guidelines now advise against offering this technique to trauma-exposed individuals.^{6 7}

In this context, brief interventions have been widely recognised as the primary choice for initial care, despite limited evidence supporting their effectiveness.⁸ However, the potential benefits of integrating such general techniques with routine outcome assessments and risk stratification, a foundation for enhanced psychoeducation, remain largely unexplored. Consequently, the effectiveness of personalised psychoeducation in augmenting simple interventions for individuals in high-exposure crisis scenarios is poorly understood. Also, despite the increasing evidence supporting the efficacy of single-session psychological interventions for the treatment of various conditions, such as depression and anxiety,⁹ their efficacy for the prevention of mental disorders remains unassessed.

Objective

This study aimed to estimate the impact of adding personalised psychoeducational components to a single synchronous session aimed at preventing the worsening of emotional distress among healthcare workers during the SARS-CoV-2 pandemic. Through a nationwide two-arm randomised clinical trial in Brazil, this study compared a single-session intervention (SSI) with an enhanced version incorporating weekly personalised prerecorded videos for 4 weeks (SSI enhanced telepsychoeducation, SSI-ET).

Methods

Study design

The TelePSI trial was a nationwide, investigator-initiated, pragmatic, unblinded, randomised controlled trial conducted in Brazil and evaluated the effectiveness of remote SSI and SSI-ET interventions. The trial was designed and supervised by a steering committee of members from academia and international experts in emotional disorders (DSP) and psychotherapy for emotional disorders (PC).

This article adhered to the principles outlined in the Declaration of Helsinki and guidelines established in the approved protocol and is reported as recommended by the Consolidated Standards of Reporting Trials (CONSORT) 2010 guideline (online supplemental table S1).

Participants

Participants were recruited nationwide via helplines, e-mails, social media and traditional media (eg, national newspapers and TV commercials). All assessments and study procedures were conducted online and all participants provided oral informed consent.

The inclusion criteria were the following: (1) professionals or interns from health services seeking help for emotional distress; (2) T-score below 70 in Patient-Reported Outcomes Measurement Information System (PROMIS) scales of anxiety, depression or irritability (here combined labelled as ‘emotional distress’).¹⁰ In the protocol defined a priori, a z-score on the PROMIS score below 1.5 was included as an inclusion criterion; nevertheless, our recruitment soon revealed that a substantial proportion of participants would have been excluded using this criterion. Thus, this criterion was not used during participants' enrolment.

The exclusion criterion was suicidal ideation assessed using a specific question from the Patient Health Questionnaire (PHQ-9).¹¹ Individuals who rated 1, 2 or 3 were immediately referred to a psychiatrist to ensure participants' safety.¹²

Randomisation and masking

The participants were randomly allocated in a 1:1 ratio to SSI or SSI-ET, and this occurred at the intervention and therapist levels after the participants completed the questionnaires and was centrally conducted by independent researchers. An automated computer-generated random number table was used to perform randomisation. Due to the nature of the interventions, both participants and therapists were aware of the assigned intervention, but participants were unaware of which intervention was of interest to the researchers.

Procedures

SSI or SSI-ET sessions started no later than 1 week after randomisation. All sessions were conducted using Google Meet. At the end of the treatment (week 4), we sent online self-report questionnaires to participants via WhatsApp. Follow-up assessments were conducted 3 and 6 months after treatment. Participants received detailed reports of their answers via email. In cases in which respondents did not respond to the questionnaires, the research team sent up to six reminders to encourage the completion of assessments.

Interventions

Single-session intervention (SSI)

The SSI protocol consisted of a one-session intervention based on supportive principles delivered by video call. The general principle was to assist participants in stressful situations using components of psychological first aid. Therapists were oriented to respect participants' safety, dignity and rights, adapt the intervention to cultural issues and encourage participants to develop autonomy and self-efficacy. Moreover, therapists were instructed to listen carefully and empathetically to participants and to create a welcoming, gentle and non-judgmental environment, where participants felt comfortable and confident expressing their thoughts, concerns and emotions. Therapists helped participants resolve problems without causing harm, strengthened their support networks and improved the ability to feel safe, connected, calm and hopeful, while ensuring social, physical and emotional support. Improving participants' coping skills, promoting greater self-confidence, enhancing satisfaction and reducing anxiety, depression and emotional pain were the main targets of the intervention. The principles of observing, listening and connecting guided the entire session.

Empathetic listening was the essential component of the session and the main technique adopted in this modality. Therapists carefully listened to participants' concerns, validated their emotions and efforts, demonstrated attentiveness and employed verbal and non-verbal empathy components as the foundation of empathetic listening. Other interviewing techniques were encouraged: (1) facilitation (refers to the posture, actions or words that encourage the participant to communicate); (2) echo (repetition of the participant’s words encouraging him/her to express feelings); (3) empathic responses; (4) validation (the therapist should look for what can be validated in the participant’s speech); (5) tranquilisation; (6) summarising the participant’s narrative; (7) highlighting transitions in the sessions; and (8) adapting the language for the participant.

The psychoeducation session had three main objectives: (1) offering a safe environment for participants to express their emotions, thoughts and experiences (10–30 min); (2) reviewing participants' self-rated symptomatic assessments completed during project registration, assisting them in areas of greatest concern (5–15 min); and (3) collaboratively planning lifestyle changes to promote positive coping strategies and to mitigate mental health risk factors. To facilitate a collaborative approach throughout the session, participants' symptom levels and their global assessments of risk and protective behaviours were presented using visually appealing graphics during the session.

SSI with enhanced psychoeducation (SSI-ET)

The SSI-ET incorporated the same components of the SSI group in the synchronous SSI but also dedicated 10 min of the session to planning eight brief videos that were sent to participants weekly over the subsequent 4 weeks (two videos per week). These psychoeducational videos were carefully selected by therapists based on participants' responses to PROMIS scales, risk and protective behaviours, and expressed needs, while also considering the best timing (ie, during daytime hours) and sequence for sending the videos. The videos addressed topics such as fear of contagion, distinguishing typical versus excessive anxiety and depression, anger versus irritability, burnout, acute stress reactions, sleep hygiene, healthy eating habits, exercise, excessive alcohol and drug use, overexposure to news and social media, caring for children, caring for the elderly and social support. These videos were developed by the researchers and are freely available on the TelePSI website (https://telepsi.hcpa.edu.br). Therapists remained available for 4 weeks to address messages sent by participants through a phone chat (WhatsApp). Although theoretically possible, interactions with therapists via this platform were primarily limited to simple ‘thank you’ notes or operational matters (eg, resolving scheduling conflicts). As such, these interactions were not systematically measured or emphasised as a significant component of the intervention. Moreover, therapists were explicitly instructed to minimise direct contact with participants through this platform, and any additional communication was considered negligible.

Outcomes

The primary outcome was the proportion of participants with a T-score ≥70 on the PROMIS scales of anxiety, depression and/or irritability at 6 months collectively measuring emotional distress. Secondary outcomes were the proportion of participants with a T score ≥60 on any of the emotional distress scales at 1, 3 and 6 months; and mean score changes between groups in anxiety, depression and irritability. Other secondary outcomes were life satisfaction, measured with the Satisfaction with Life Scale, which reports its results with higher scores indicating higher satisfaction with life, and service satisfaction assessed through the net-promoter score (NPS), calculated by asking participants a single question regarding the likelihood of recommending each therapy to someone on a scale from 0 to 10. The respondents were classified as promoters (9–10), passives (7–8) or detractors (0–6). The NPS was derived by subtracting the percentage of detractors from promoters by coding detractors as −1, passives as 0 and promoters as 1. We also asked participants to rate their overall satisfaction using a single question ranging from 1 ('Extremely dissatisfied') to 7 ('Extremely satisfied'). Measures of anxiety, depression, irritability and life satisfaction were measured at months 1, 3 and 6, and service satisfaction was evaluated at month 1.

Participants’ safety was ensured by tracking suicide attempts, psychiatric hospitalisation and a short questionnaire developed by the research team to track the side effects of interventions with the following items: ‘My mental health is worse because of therapy’, ‘I developed a dependency on therapy to get on with my life’, ‘Therapy made me feel guilty about my problems’, ‘Therapy reactivated bad thoughts that worsened my mental health’, ‘I got involved in unnecessary conflicts with other people because of therapy’, ‘Therapy led me to think there is no solution for my problems’, ‘Therapy made me feel unable to face my problems’ and ‘Overall, therapy has done me more harm than good’. Participants responded to each statement using a 5-point Likert scale, and the responses were subsequently categorised into those who indicated the presence of undesirable side effects (‘Completely Agree’ or ‘Agree’) and those who did not (‘Completely Disagree’, ‘Disagree’ or ‘Neither agree nor disagree’).

Therapists' fidelity to interventions was assessed by a supervisor who reviewed randomly selected recorded sessions. The supervisor used a scale developed by our team to evaluate the quality and fidelity of each session. The assessment of quality involved seven distinct criteria: (1) establishing a therapeutic alliance with the patient; (2) structuring and organising the session; (3) compassion, empathy, genuineness and warmth; (4) effective use of non-specific therapeutic skills such as clarification, reflection and validation; (5) confidence and self-assurance; (6) reassurance; and (7) proficiency in instilling hope. Each item was rated on a scale of 1–5, with higher scores indicating greater session fidelity.

Choice of the primary outcome

The primary outcome measures were PROMIS anxiety, depression, and irritability scales collectively measuring emotional distress. PROMIS is a self-rated assessment tool developed by the National Institutes of Health (NIH) to assess patients' perspectives on their health status and quality of life. These scales, covering emotional states over the previous 7 days, comprised eight items for anxiety, eight items for depression, and five items for irritability. Responses are recorded on a 5-point Likert scale, with higher scores indicating increased emotional distress. All PROMIS scales demonstrate high internal consistency and reliability (above .70), being suitable both for screening individuals with subclinical symptoms and for estimating treatment outcomes in patients diagnosed with internalising disorders.¹⁰ We used PROMIS T-scores with an average of 50 and a SD of 10. Therefore, a T-score of 70 (ie, our inclusion criterion) represents 2SD above the sample mean. This measure was chosen as the primary outcome because it is designed to be patient-friendly and uses fewer items than most scales designed to evaluate the same symptom domains, thereby decreasing the respondent burden while estimating a wide range of symptoms of emotional distress. Additionally, this instrument is freely available and has been translated into several languages, facilitating the replication of our methods in future studies.

Statistical analysis

The study was designed to have 90% power to detect a 5% difference between groups in the primary outcome (alpha 0.05). Estimating a 20% loss during follow-up, this analysis indicated a sample size of 1100 participants. Analysis was performed according to intention-to-treat (ITT) criteria.

The cumulative proportion of incident cases at 6 months was assessed using survival analysis with all available data. Differences between groups and hazard ratios were estimated using Cox proportional hazards. Continuous outcomes were analysed using linear mixed models with all available data from baseline, 1, 3 and 6 month follow-up. Comparisons of NPS, treatment satisfaction and adverse events between groups were tested using a Multiple Imputation Welch Two Sample t-test and pooled results from logistic regressions accounting for missing data. All analyses were performed using R. Survival analysis, linear mixed models and equivalence tests were conducted using survival, lme4, mice, MKmisc and Effects packages.13,15 Cohen’s d was used as a measure of effect size, with values of 0.2, 0.5 and 0.8 interpreted as small, medium and large, respectively.¹⁶

In order to test for the absence of minimal clinically relevant differences between treatment arms, we have also included, a posteriori, an equivalence test (two one-sided tests equivalence testing) between interventions considering −7% as the smallest effect size of interest, as defined a priori for the statistical power analysis.¹⁷ This represents approximately half of the differences found for the other prevention trials for the same time period.¹⁸

Due to the crisis context of the SARS-CoV-2 pandemic, this trial was prospectively registered at ClinicalTrials.gov (NCT04632082) on 11 November 2020, during the recruitment process; nevertheless, less than 6% of participants had already been included by the time of registration, no preliminary statistical analyses were performed and only minor changes in the previously defined methodological procedures were implemented during the entire trial. Rather than excluding participants based on a clinical assessment of suicide, we excluded them based on affirmative responses to the PHQ-9 suicide item. This decision stemmed from the concern that randomising participants after the psychiatric assessment might introduce bias in a trial designed to compare distinct interventions against a single session. Also, a z-score on the PROMIS below 1.5 was initially defined as an inclusion criterion; however, our recruitment process soon revealed that a substantial proportion of participants would have been excluded using this criterion, and this criterion was not implemented.

Findings

Of the 3328 individuals who sought to participate in this project from 19 May 2020 to 31 December 2021, 3086 consented to participate in the study, and 490 were excluded and referred to a psychiatrist due to concerns about suicide risk. We enrolled 1112 participants who met the inclusion criteria and subsequently allocated them to SSI (n=549) or SSI-ET (n=563) (figure 1). The majority of the sample consisted of women (86.4%), and the mean age of the sample was 37.2 (SD=9.46). Regarding symptom severity at baseline, a high heterogeneity was observed in participants' symptom levels, with distinct distributions across different symptom domains. While most individuals exhibited moderate symptom levels across all domains, this predominance was substantially higher for anxiety symptoms (84.1%) compared with depression (39.8%) and irritability (42.0%). Each group is further described in table 1.

Figure 1. Consolidated Standards of Reporting Trials flow diagram.

Table 1. Sample description among groups.

–	SSI(n=549)	SSI-ET(n=563)
Sex at birth (female)	475 (86.7%)	485 (86.1%)
Emotionaldistressscales
PROMIS anxiety	–	–
None to slight	27 (4.9%)	15 (2.7%)
Mild	64 (11.7%)	71 (12.6%)
Moderate	458 (83.4%)	477 (84.7%)
PROMIS depression	–	–
None to slight	143 (26%)	140 (24.9%)
Mild	193 (35.2%)	193 (34.3%)
Moderate	213 (38.8%)	230 (40.9%)
PROMIS irritability	–	–
None to slight	214 (39%)	204 (36.2%)
Mild	105 (19.1%)	121 (21.5%)
Moderate	230 (41.9%)	238 (42.3%)
Age (years), mean (SD)	37.5 (9.06)	36.9 (9.84)
PROMIS anxiety (T-score), mean (SD)	64.2 (4.80)	64.5 (4.45)
PROMIS depression (T-score), mean (SD)	57.9 (5.68)	58.3 (5.84)
PROMIS irritability (T-score), mean (SD)	57.4 (7.77)	57.4 (7.71)
Satisfaction with life (sum), mean (SD)	20.4 (5.91)	19.9 (5.68)

PROMISPatient-Reported Outcomes Measurement Information SystemSSIsingle-session interventionSSI-ETsingle-session intervention with enhanced psychoeducation

Primary outcome

The estimated cumulative proportion of incident severe cases at 1 month was 6.9% (95% CI 9.4% to 4.4%) in the SSI group and 4.8% (95% CI 6.9% to 2.7%) in the SSI-ET group, at 3 months it was 12.4% (95% CI 15.8% to 8.9%) in the SSI group and 9.9% (95% CI 12.9% to 6.8%) in the SSI-ET group, and at 6 months it was 17.5% (95% CI 13.1% to 21.6%) in the SSI group and 15% (95% CI 10.9% to 18.8%) in the SSI-ET group with no difference between groups (HR=0.81, 95% CI 0.83 to 1.79, p=0.291) (figure 2).

Figure 2. Survival analysis. SSI, single-session intervention; SSI-ET, SSI with enhanced psychoeducation.

Secondary outcomes

Incidence of subthreshold levels of emotional distress (T-scores above 60). This analysis was performed using the sample that did not have a score of 60 or above at baseline (n=143). The estimated cumulative proportion of incident severe cases at 1 month was 20.8% (95% CI 31.5% to 8.5%) in the SSI group and 19.4% (95% CI 17% to 21.2%) in the SSI-ET group, at 3 months it was 25.2% (95% CI 36.6% to 11.8%) in the SSI group and 25.7% (95% CI 37.3% to 11.9%) in the SSI-ET group, and at 6 months it was 32.7% (95% CI 45.1% to 17.5) in the SSI group and 28.9% (95% CI 41.3% to 13.9%) in the SSI-ET group, with no difference between groups (HR=0.81, 95% CI 0.58 to 2.53, p=0.587).

Relevant time effects. Linear mixed models, using all available data, showed prominent time effects for all outcomes, indicating significant differences from baseline to the 1, 3 and 6 month follow-up assessments (online supplemental table S2). The mean changes were significant in each treatment arm. Standardised symptom reduction for anxiety symptoms from baseline to 1 month was Cohen’s d=0.68 (p<0.001) for SSI and Cohen’s d=0.77 (p<0.001) for SSI-ET. Standardised symptom reduction for depressive symptoms from baseline to 1 month was Cohen’s d=0.86 (p<0.001) for SSI and Cohen’s d=1.06 (p<0.001) for SSI-ET. Finally, standardised symptom reduction for irritability symptoms from baseline to 1 month was Cohen’s d=0.68 (p<0.001) for SSI and Cohen’s d=0.89 (p<0.001) for SSI-ET. All differences were maintained at the 3 and 6 month follow-ups with similar magnitudes (figure 3).

Figure 3. Estimated scores on all secondary outcomes as a function of each timepoint (baseline, 1 month, 3 months and 6 months follow-up). PROMIS, Patient-Reported Outcomes Measurement Information System; SSI, single-session intervention; SSI-ET, SSI with enhanced psychoeducation.

Time-by-group interactions. We found significant time-by-group interaction for the PROMIS anxiety scores (online supplemental table S2), meaning that at 1 month SSI was significantly worse than SSI-ET in reducing PROMIS anxiety scores (mean difference=1.12, 95% CI 0.14 to 2.11, p=0.024; Cohen’s d=0.14). The difference was no longer significant at the 3 month or 6 month follow-ups (Cohen’s d range 0.05–0.09, p values >0.05). No time-by-group interactions were observed for irritability or depressive symptoms.

NPS and overall satisfaction. There were no significant differences in the probability of participants recommending the treatments. The estimated NPS for SSI was 53% and that for SSI-ET was 54.6% (mean difference=2.6%, t=0.339, df=27, p value=0.737). There were also no differences between the overall treatment satisfaction with scores of 5.6 for SSI and 5.7 for SSI-ET (mean difference=0.12, t=1.125, df=103, p-value=0.213).

Therapists' fidelity. The supervisor observed 10 psychoeducation sessions without videos and 9 sessions with videos. The average fidelity score for therapists of SSI was 4.94 (SD 0.1), while for therapists facilitating the SSI-ET the mean score was 4.78 (SD 0.32), with no difference between groups (p>0.05).

Attrition. The flow diagram in figure 1 illustrates the number of individuals assessed at 1, 3, and 6 month follow-up, demonstrating considerable attrition rates. There was a significant between-group difference in the proportion of participants who did not receive the allocation intervention (18.9% in SSI vs 13.3% in SSI-ET, χ²=6.16, df=1, p value=0.013). A sensitivity analysis using modified ITT (excluding those who did not receive allocated intervention) revealed similar results for all outcomes (primary outcome: HR=0.807, p=0.269). There were no significant differences between losses during follow-up (χ²=0.958, df=3, p value=0.8114).

Post-hoc equivalence test. Equivalence test between treatment arms was within the equivalence bounds of −7% and 7% for the primary outcome: SSI vs SSI-ET (Z=−1.81, p=0.035).

Safety assessment. Three patients attempted suicide during the 1 month follow-up (SSI: 2/349, 0.6% vs SSI-ET: 1/367, 0.3%, χ²=0.39, p=0.534), one during the 3 months follow-up (SSI: 1/288, 0.3% vs SSI-ET: 0/292, 0%, χ²=1.02, p=0.314) and three at 6 months follow-up (SSI: 1/235, 0.4% vs SSI-ET: 2/250, 0.8%, χ²=0.28, p=0.599). Nine patients were hospitalised due to mental health reasons during the 1 month follow-up (SSI: 4/349, 1.1% vs SSI-ET: 5/367, 1.4%, χ²=0.07, p=0.795), seven at the 3 months follow-up (SSI: 2/288, 0.7% vs SSI-ET: 5/292, 1.7%, χ²=1.26, p=0.262) and four at 6 months follow-up (SSI: 1/235, 0.4% vs SSI-ET: 3/250, 1.2%, χ²=0.89, p=0.346). No differences were observed between the SSI and SSI-ET groups in terms of undesirable side effects related to the intervention (online supplemental table S3).

Discussion

In this nationwide two-arm randomised clinical trial conducted in Brazil, we investigated the efficacy of SSI compared with SSI enhanced with personalised psychoeducational videos (SSI-ET) to prevent the worsening of emotional distress among healthcare workers during the SARS-CoV-2 pandemic. We found no significant advantage of the video-enhanced SSIs on preventing emotional distress exacerbation. Although symptom improvement was observed in both groups, the absence of a control group prevents attributing this improvement directly to either intervention. As these findings suggest the potential feasibility of an SSI as a standalone intervention in a high-stress cohort, they reinforce the importance of further exploring the efficacy of brief interventions in randomised controlled trials.

Despite the widespread use of Psychological First Aid, empirical evidence regarding its effectiveness remains limited.¹⁹ Our study suggests that the core principles of Psychological First Aid may play an important role in the prevention of internalising symptoms and that the additional psychoeducational components with therapist contact do not significantly enhance intervention efficacy. The role of SSIs in psychological interventions is increasing across various conditions, including depression and anxiety.⁹ Additionally, our findings indicate that brief interventions based on these principles may be relevant not only for populations experiencing severe distress but also for those with moderate symptom levels. However, without a control group, we cannot definitively attribute symptom improvement to the interventions themselves, and these prevention strategies should therefore be further investigated in controlled studies. Although the baseline incidence of depression in this population without intervention is unknown, previous studies have reported rates ranging from approximately 25% to 30% in control interventions at 24 weeks follow-up, which is substantially higher than 15%–17.5% we found here.¹⁸ Even though these studies are not directly comparable, they provide some benchmarks for what would be expected in a 6 months period for individuals with subthreshold symptoms.

Regarding adverse events, the incidence was low for all measures, except for the concern of dependence (above 5%). This concern was raised by the participants even though the therapist had very limited contact. The contact with the therapist on sending the videos did not increase patient satisfaction, which again reinforced the importance of synchronous contact with the therapist. Safety assessment during the follow-up period found that the majority of suicide attempts occurred during the initial month of follow-up and reinforce the need of assessment for suicidality after therapy.

This study has important limitations. First, we had a significant attrition rate and reliance on self-reports in a virtual environment, which may have impacted the representativeness and comparability of the findings with the current literature. Despite these challenges, methodological rigour was maintained through statistical approaches that accounted for all available data and used validated patient-reported outcomes. Brazil was among the countries most affected by the severe economic and health crisis caused by the COVID-19 pandemic, achieving the second-highest death toll globally. This raised several important ethical questions during the approval of this project by the ethics committee. In this extreme context, the project was primarily focused on assisting individuals seeking mental health support during the pandemic through our contact channels rather than recruiting participants primarily for research purposes. Accordingly, including these individuals in a control group without an active component was not deemed acceptable by the Ministry of Health, precluding analyses that could distinguish the effects of active interventions from placebo responses or the effect of regression to the mean. Thus, it remains uncertain how much of the observed symptom improvement was truly caused by the interventions. Nevertheless, the favourable satisfaction levels with the received intervention suggest that these approaches may hold promise and should be further investigated in future research. Moreover, although we believe that this study provides a significant contribution to the field of brief psychological interventions in crisis contexts, its execution in a challenging and unique setting (ie, in a single country severely affected by a worldwide pandemic) may limit the generalisability of these findings to other crisis scenarios. Although it is theoretically possible that the additional contact with the SSI-ET group therapists, required for sending the personalised psychoeducation videos, could represent an additional component of this intervention, therapists were explicitly instructed to minimise direct contact with participants, preferably limiting it to what was strictly necessary for sending the videos. While the additional communication was considered negligible due to these imposed restrictions, the extent of this contact was not systematically assessed, limiting the capacity to precisely infer the potential impact of this effect. Lastly, the videos were free to access on our website, therefore, we cannot guarantee that contamination did not occur in the SSI arm.

Clinical implications

Even though adding asynchronous videos to the SSI provided no additional benefit, this study highlights the potential feasibility of SSIs as a scalable approach to supporting healthcare workers during health crises, emphasising the role of the synchronous component of the intervention. While symptom improvement was observed, the absence of a control group prevents attributing these changes directly to the intervention. Nevertheless, this improvement, along with the favourable satisfaction levels reported by participants, underscores the need for further investigation of brief and transdiagnostic interventions in future controlled studies, as accessible and cost-effective mental health strategies are essential in high-stress settings.

Data availability statement

Data are available upon reasonable request.