Introduction

People With Mental Illness (PWMI) can be subjected to stigma and discrimination from health professionals working in different fields [1–3]. Experiences of discrimination have been reported by 31% of service users in physical-health care settings and 44% of service users in mental-health care settings [4]. A review by Henderson et al conducted in 2014 [1] found that professionals suffering from burnout, professionals working in mental health and early career professionals are the ones most in need of interventions to decrease their stigmatization of patients.

A plethora of research investigated the effect of different approaches to reduce the stigma of mental illness. Interventions were mainly divided into three categories: education, social contact and social activism [5]. In a scoping review on interventions aiming to decrease health students ‘stigma towards patients with schizophrenia by Chen & all conducted in 2024 [6], interventions ranged from theoretical lessons, movies, role-playing, reading and discussion to drawing, focus group discussions, contact with PWMI through internships and augmented reality. Duration of interventions varied from a 45-minute voice simulation experience to a 13-week program. Overall, diverse interventions have shown potential to shift attitudes toward schizophrenia, though their effectiveness varies depending on the approach and context. Among all types of interventions, a combination of education and social contact, was shown to be the most effective approaches for specific target groups such as medical students [6, 7]. However, these studies were subject to many methodological limitations, mainly the lack of assessment of their sustainability on the long-term, the focus on High-Income Countries (HIC) [8], and the considerable heterogeneity in interventions, controls and outcome measurements across studies [9–12]. Furthermore, few studies have specifically evaluated stigma-reduction interventions among medical students using measures adapted to this population and the majority of studies focused on assessing knowledge and attitudes while neglecting behavioural outcomes [7, 9, 13].A recent review of the literature by Heim et al in 2019 [9], focusing on mental-health related stigma reduction interventions among medical and nursing students in LMIC, reported significant improvements in outcome measures, although it was difficult to determine the most effective approach due to the heterogeneity of studies [9]. Despite growing global attention to stigma toward people with mental illness, important research gaps remain in low- and middle-income countries (LMICs) [9]. Most evidence on stigma-reduction interventions among healthcare professionals and medical students has been generated in high-income countries, where mental health services, educational resources, and cultural perceptions of mental illness differ substantially from those in LMICs. Consequently, the transferability of these interventions to LMIC settings remains uncertain. In many LMICs, including countries in North Africa and the Middle East, mental illness continues to be strongly influenced by cultural beliefs, social norms, and structural barriers within health systems, which may shape both the manifestation of stigma and the way future healthcare providers interact with people with mental illness [14]. Furthermore, mental health training within medical curricula in LMICs often remains limited, with fewer opportunities for structured contact with people with mental illness and for training that explicitly addresses stigma and discrimination in clinical practice. In Tunisia specifically, research on mental health stigma has largely focused on public attitudes or patient experiences, while little attention has been given to stigma among medical students or to the evaluation of structured anti-stigma interventions within medical education [15]. A study by Saguem (2023) [16] aimed to evaluate the effectiveness of a four-session educational intervention for reducing stigma of mental illness targeting family medicine trainees in Tunisia. While this was a controlled study with a 2 month follow up measure, no behavioral outcomes have been assessed. As a result, there is little empirical evidence on how future physicians in Tunisia perceive mental illness, how prepared they are to respond to discrimination experienced by patients, and whether targeted educational interventions can effectively reduce stigma in this context. Addressing this gap is essential to inform context-appropriate strategies for integrating stigma-reduction approaches into medical training and ultimately improving the quality of mental health care in Tunisia.

In the light of the gaps in the literature and the field implementation challenges, the READ (Responding to Experienced and Anticipated Discrimination) anti-stigma training was developed to teach medical students how to respond constructively to experienced and anticipated discrimination among patients, as well as to reduce stigma. It was implemented at thirteen sites across ten countries including Tunisia, and aimed to assess both knowledge, attitudes, and behaviour changes among medical students [17].

This paper aims to evaluate the effectiveness of the READ training in Tunisia on mental health stigma related knowledge, attitudes, and skills in responding to patients anticipated and experienced discrimination, in a large sample of Tunisian medical students.

Materials and methods

Results and discussion

From April 2017 to December 2019, 123 final-year medical students undertook their psychiatric rotation at the department “Psychiatry A” of Razi Hospital. Among these students, eight did not consent to participate in the study (Fig 1). Our sample was, therefore, composed of 115 students (44 controls and 71 intervention). Most students were female (74.78% vs 25.22%) which corresponds to the current gender distribution at the Faculty of Medicine of Tunis. They were aged between 21 and 24 years old (mean age: 23 years old + /- 0.72) (Table 1). Five students did not attend either the first or the second presentation, however, they were then debriefed about its contents and still included in the READ study (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics.

https://doi.org/10.1371/journal.pone.0352158.t001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Study flow chart.

https://doi.org/10.1371/journal.pone.0352158.g001

There was no evidence of an association between the explanatory variable (intervention or control) and gender (chi-square test, p = 0.69) or age (t-student test, p = 0.26) (Table 1).

Table 2 shows the MAKS and MICA2 mean total scores and Table 3 the OSCE categories percentages at baseline for the control and intervention arms.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Mean scores (MICA2 and MAKS), and percentages (OSCE) for each study arm at baseline and follow-up.

https://doi.org/10.1371/journal.pone.0352158.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Percentages (OSCE) for each study arm at baseline and follow-up.

https://doi.org/10.1371/journal.pone.0352158.t003

At follow up and compared to the baseline, both intervention and control groups had higher MAKS mean total scores (from 20.93; SD = 3.03 to 22.70; SD = −2.21 for the control group, and from 21.63; SD = 2.19 to 23.84; SD = −2.62 for the intervention group) and lower MICA2 mean total scores (from 44.90; SD = −8.97 to 41.40; SD = −8.50 for the control group; and from 43.73; SD = −7.76 to 33.52; SD = −9.56 for the intervention group) (Table 2).

The percentages of participants that scored” pass” at OSCE at the follow-up increased for both groups (from 27.27% to 47.73% for the control group; and from 22.54% to 56.34% for the intervention group) (Table 3).

In our sample, OSCE data were missing for 73 participants (63.5% of the total sample, N = 115). Analyses were conducted on complete cases (n = 42). No imputation was performed. Given the extent of missing data, findings should be interpreted with caution due to potential bias. We didn’t find differences in age (p = 0.73) and gender (p = 0.53) between the groups with and without OSCE scores. There were no missing data from MAKS and MICA2 scores.

Male students were more likely to be scored “pass” at OSCE than female students (p = 0.003).

There was no evidence of a difference between MAKS mean total scores of the intervention and control group after adjusting for age, gender, and baseline MAKS mean total scores (Adjusted coefficient:0.70; p = 0.11; 95% CI: −0.16 to 1.6) (Table 2).

On the other hand, after the READ training, the adjusted coefficient of MICA2 mean total scores in the intervention group was lower than the adjusted coefficient of the control group after adjusting for age, gender, and MICA2 baseline scores (Mean difference (MD): −7.88 p < 0.001; 95% CI: −10.23 to −3.96) (Table 2).

After the READ training, the intervention group was 4.45 times more likely to be scored “pass” at OSCE compared to the control group (p = 0.046; 95% CI: 1.03 to 19.26) after adjusting for age, gender, and OSCE baseline scores (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Associations between the READ intervention and MAKS MICA2 and OSCE.

https://doi.org/10.1371/journal.pone.0352158.t004

Discussion

READ-Tunisia is the first study investigating the effectiveness of anti-stigma training in reducing the stigma of medical students towards PWMI and including both knowledge, attitudes, and behaviours as its measured outcomes in the Eastern Mediterranean region. Complying to the standards of the international multisite READ study, it used a rigorous methodology through the implementation of a manualized anti-stigma training based on theoretical foundations and medical education groundwork [17]. The READ training proved its effectiveness in decreasing students’ stigmatizing attitudes and improving their skills while it had no effect on their stigma mental health-related knowledge beyond the impact of the rotation.

This is in accordance with findings from the review done by Thornicroft et al in 2016 that revealed a consistent pattern across the literature of attitudinal and behavioural changes and less evidence for knowledge improvement in the immediate aftermath of anti-stigma trainings [7]. A 2022 systematic review carried by Gervas et al [24] found few cases of knowledge gain in interventions based on methods such as teaching practices, social contact, direct contact, contact with recovered patients, educational emails, filmed contact or in vivo exposure. The absence of significant changes in knowledge gain in our study might be due to the small sample size. Indeed, in the large international multisite READ study, anti-stigma training was associated with positive changes in knowledge, attitudes, skills, and patient perceived empathy among medical students [25]. However, another INDIGO site study done in the Czech Republic with smaller sample size than our study demonstrated reductions in stigma-related attitudes and improvements in mental illness knowledge [26]. Another explanation might be the good pre-acquired knowledge about mental illness during medical education among our sample. Indeed, medical students undertaking their studies in the Faculty of Medicine of Tunis are introduced to mental illness during their second year of medical studies where they undertake a two-week psychiatric rotation at the hospital, along with teaching about the psychopathology of mental illness at the faculty. In their fifth year, they are taught a whole module about the different psychiatric disorders, usually before or during the psychiatric rotation. However, these teachings focus on psychopathological and clinical aspects of psychiatric disorders and do not aim to increase awareness about the stigma of mental illness and its consequences on patients and their families. More than that, didactic methods are predominantly based on traditional lecturing where students passively acquire knowledge about mental illness without learning how to behave with patients.

Instead, READ training allocates most of its time to interactive and self-reflective discussions and role plays where students are either playing the role of a patient with anticipated and experienced discrimination or the role of a health professional/friend with the task of responding to discrimination. In both cases, students are confronted with their potential stigmatizing attitudes and work towards acquiring the necessary anti-stigma skills.

Our findings contribute to the growing body of evidence supporting the effectiveness of structured anti-stigma interventions implemented during psychiatric training in LMIC contexts. However, it is also important to consider whether such interventions lead to sustained long-term improvements. In a systematic review by Gervás et al [24], the authors reported that many direct-contact interventions did not demonstrate long-term benefits, highlighting the need to better understand factors influencing the durability of anti-stigma effects. In particular, the duration and intensity of the intervention may play an important role in determining its impact. Longer or repeated anti-stigma training sessions may allow for deeper reflection and consolidation of learning, potentially leading to greater and more sustained improvements in knowledge, attitudes, and behavioural outcomes. Future studies should therefore explore whether integrating repeated sessions or longitudinal training components within medical curricula could enhance the sustainability and magnitude of anti-stigma intervention effects.

Previous studies investigating the impact of psychiatric rotations on students and undergraduates in the health field found mixed results, with some suggesting a worsening of misconceptions and stereotypes about PWMI [7, 9]. A Greek study found that final-year medical students held more stigmatizing attitudes towards patients with schizophrenia at the end of their psychiatric rotation [27]. An Egyptian interventional study including 300 fifth-year medical students from Ain Shams University found no change in their MICA scores after three weeks of training [28]. In both studies, students were given training on common psychiatric disorders, interviewing skills, psychotherapy, and pharmacotherapy. The outcomes of the studies were attributed to insufficient duration of the psychiatric rotation. However, evidence regarding the relationship between the duration of psychiatric training and stigma reduction remains inconclusive. While some studies suggest that three-week rotations may be insufficient to meaningfully influence students’ perceptions [29], others have reported improvements in knowledge, beliefs, and attitudes following rotations lasting more than four weeks [30–32], and even stronger effects with longer educational programs, such as 13-week interventions that significantly reduced social distance toward PWMI [30]. Variations in findings could be explained by other factors such as differences in the studies’ designs, the implementation strategies, the used outcome measures, but also the conditions of the clinical rotations [9]. In LMICs like Tunisia and Egypt where access to mental health services is difficult and health coverage for psychiatric disorders is insufficient, medical students undertaking their psychiatric rotations in centralized psychiatric hospitals tend to witness only patients with severe symptoms. Consecutively, students may acquire biased negative perceptions about mental illness [33].

Acknowledgments

We are grateful to the Indigo Network, a collaboration of research colleagues in over 30 countries worldwide committed to developing knowledge about mental illness-related stigma and discrimination, both in terms of their origins and their eradication. It was coordinated by the Centre for Global Mental Health, Institute of Psychiatry, Psychology and Neuroscience at King’s College London.