Data was systematically collected from a total of 118 medical students, required prior to graduation, meeting the criteria for participating in an OSCE designed to evaluate palliative care competencies at the University of Cádiz in May 2023. The OSCE included a series of 25 assessment items focused on various clinical skills and communication abilities essential in palliative care.

One of the ten stations in the OSCE was specifically designed to evaluate end-of-life care competencies. This station was structured to simulate a real-world scenario and included three key participants: two trained actors (one portraying a patient with a terminal illness and the other a concerned family member) and a qualified evaluator who observed the interaction without intervening. The scenario was carefully scripted to ensure standardization, with the actors adhering to predefined roles that emphasized specific challenges related to end-of-life care, such as managing intense emotions, navigating ethical dilemmas, and making informed decisions about palliative sedation.

Students were tasked with engaging in a structured clinical encounter, which included taking a medical history, assessing the patient’s clinical and emotional status, and addressing the concerns of the family member. During the interaction, students were expected to demonstrate diagnostic proficiency, empathetic communication, and professionalism. For example, the patient actor might express physical discomfort or anxiety about their prognosis, while the family member actor could pose questions about advance directives or request clarity on the patient’s care plan.

The evaluator used a checklist of 25 predefined assessment items to rate the student’s performance across domains such as diagnostic accuracy, empathy, professionalism, and ethical decision-making. This standardized approach minimized subjectivity and ensured consistency across all evaluations. To replicate the unpredictability of real clinical encounters, students were not informed of the specific scoring criteria beforehand. This design provided a realistic assessment of their preparedness to handle complex, emotionally charged scenarios in end-of-life care.

Student performance in the end-of-life care OSCE station was analyzed using basic descriptive statistics, including mean, median, interquartile range, minimum, and maximum scores, to capture central tendencies and distribution across the evaluated competencies. A histogram was created to visually represent score distribution, helping to identify common performance trends.

Each item’s compliance rate was calculated to determine the proportion of students who met specific predefined criteria in the OSCE station. This metric provided a direct measure of how successfully students demonstrated the required competencies, such as diagnosing core pathologies or navigating ethical decision-making. High compliance rates indicated areas where students generally performed well, reflecting the effectiveness of current training in those domains. Conversely, low compliance rates highlighted specific competencies where additional curricular emphasis may be needed.

The discriminative ability of each item was assessed using discrimination coefficients, which measure how effectively an item distinguishes between students with higher and lower overall performance. Items with high discrimination coefficients, such as those assessing communication and consent for palliative sedation, were critical for identifying nuanced competencies and advanced decision-making skills. Items with low discrimination coefficients, where nearly all students answered correctly, typically reflected foundational knowledge or skills that were uniformly mastered and thus required less focused discussion.

Item difficulty was evaluated using ease rates, defined as the proportion of students who successfully completed each item. This measure identified tasks that were inherently more challenging, such as managing advance directives or identifying signs of imminent end-of-life conditions. Together, these metrics—compliance rate, discrimination coefficients, and ease rates—provided a robust framework for interpreting student performance, enabling targeted improvements in training that address specific gaps while reinforcing existing strengths.

All data analyses were performed using SPSS 22.0 for Windows (SPSS Inc., Chicago, IL, USA).

In our evaluation, items related to medical history, examination, communication skills, ethical-legal aspects, and professionalism were analyzed (Table 1).