Our narrative assessment confirmed the positive impact of cultural safety training based on our co-designed curriculum  for Colombian medical students and interns. The most significant changes aligned with all seven elements of the CASCADA partial order of intermediate outcomes in behaviour change. The stories provided by trial participants highlighted changes in knowledge, attitudes, subjective norms, and action related to cultural safety.
Our findings fit with a review of cultural safety training changing student knowledge, attitude, self-confidence, and behaviour when interacting with Indigenous populations . The review suggested that cultural safety training in medical education enhances respect for and acceptance of traditional and cultural health practices, it improves relationships between health professionals and patients from non-dominant cultures, and promotes healthier outcomes . Another review reported changes in knowledge, attitudes, confidence, perceptions, collaboration, empathy, communication, behaviour, and practice . A recent scoping review identified culturally safe strategies to improve rural Indigenous palliative care . The authors described involvement of patients in decision making, self-reflection of care providers, and recognition of how culture shapes health care. A rapid review of diabetes care in Indigenous populations of Canada, Australia, New Zealand and the United States  identified positive effects of culturally safe interventions on clinical outcomes of patients, enhanced patient satisfaction and access to health care, and increased care provider confidence in providing care. These results support the idea of Arthur Kleinman that considering sociocultural aspects of care could enhance patient outcomes .
Our findings suggest that our training could encourage changes in practice of medical students and interns. While the literature reviews are entirely focused on Indigenous health, we provided evidence of the usefulness and relevance of cultural safety training among non-Indigenous populations that use traditional medicine, a widespread phenomenon in Latin American countries. Our results are relevant for medical educators interested in enhancing intercultural skills of medical trainees.
Our RCT explored the effectiveness of an innovative teaching strategy based on the transformative learning approach, which uses education that is interactive, participatory, and based on challenges . Although we provided transformative learning through our game jam, the control group also received elements of transformative learning as this was requested by the directives at La Sabana University. Mezirow proposed transformative learning as a way to confront ethnocentrism . Learners change beliefs about themselves, about others, and about practices, to make them more inclusive, open, and emotionally able to behavioural change . Health-related students involved in transformative learning reported more confidence in caring for patients from non-dominant cultures [39,40,41]. Similarly, a recent game jam promoted self-discovery, reflections on identity, and support of the cultural identity of the Sami people in Finland . Our results support the effectiveness of transformative learning for cultural safety training.
The reported areas of change in the stories reflected the content of our co-designed curriculum [24, 25] and included acknowledging culturally unsafe actions and their consequences, examining the students’ own attitudes, beliefs, and values, and how they shape their professional practice, willingness to listen and learn from their patients about traditional practices, and skills to discuss with patients to reach an agreement on their treatment, thus improving the doctor-patient relationship. Our results provide some evidence that cultural safety training based on our co-designed curriculum  may promote positive outcomes for both health professionals and patients.
We previously reported a positive perception of traditional and cultural health practices among Colombian medical students who participated in a five-month community-based cultural safety training program [13, 14]. The present study had a very substantial non-response rate, suggests positive outcomes for medical trainees after a much briefer (8 h) intervention. With the content overload in contemporary medical training and little time to include new subjects, our findings will be relevant to medical educators interested in cultural safety training.
A common limitation of medical education research based on self-reported data is social desirability bias , where participants feel pressured to report what they think the researcher wants to hear. We tried to reduce this bias by collecting anonymous data, making clear for the participants that their stories would not have any influence on their academic performance, and by suggesting that stories of negative changes or no change at all were also welcomed. There are reports of reduced desirability bias in web-based surveys, which we used in our study . Despite this, we are cautious in interpreting our results. Only a self-selected minority of the students submitted a story, so this report emphasises stories from students interested in cultural safety. The stories describe how cultural safety training works when it does work, not the extent to which it worked.
To enhance trustworthiness of the data, the MSC stories were anonymized and we did not collect any additional information beyond responses to the MSC questions. A drawback of this approach was that we did not know if the students who reported their stories participated in the intervention or control group; both groups received cultural safety training. We did not know the sociodemographic characteristics of the subsample who provided the stories of change. Future research might consider changes reported by, for example, younger students and older students.
We are aware the narrative evaluation is at best an indirect assessment of change clinical practice. Future cultural safety research could use more objective measures, like direct observation of trainees in clinical settings and patient-related outcomes, ideally measured on or reported directly by patients.