This study was designed as a prospective, longitudinal, randomized controlled trial and conducted in accordance with the CONSORT (Consolidated Standards of Reporting Trials) guidelines (Multimedia Appendix 1) [17]. Participants were randomly and equally assigned to either a simulation-based education group or a traditional educational method group. Data were collected between April 2022 and January 2023.

Participation was voluntary, and all students provided informed consent before their inclusion in this study. The study adhered to the ethical principles of the Declaration of Helsinki [18] and was approved by the institutional review board of the Department of Surgery, Faculty of Medicine and Health Sciences, University of Salamanca (protocol code 001-2018; July 13, 2018). Measures were implemented to protect participant privacy and confidentiality, including the anonymization of all collected data and secure storage of records. No compensation or incentives were provided to students.

This study used convenience sampling by recruiting participants who were easily accessible in a university. While this approach allows for efficient data gathering, it is crucial to recognize that it may limit the generalizability of the study due to potential selection biases. Therefore, an intention-to-treat analysis was performed to address this concern.

A total of 34 students from the Faculty of Medicine at the University of Salamanca (medical degree program) participated in this study. A sample was selected including those from the fourth, fifth, and sixth years of study. Both data collectors and analysts were unaware of the assigned interventions. The sample consisted of 32% (n=11) male and 68% (n=23) female individuals distributed in 2 independent groups. The inclusion criteria were as follows: (1) students enrolled in the medical degree program at the University of Salamanca; (2) students in the fourth, fifth, or sixth year; (3) students who voluntarily provided informed consent; (4) students willing to participate in all activities and assessments of the study; and (5) students with sufficient cognitive and physical capabilities to participate in the teaching interventions (simulation based).

All educational sessions were conducted by 2 instructors with extensive experience (>10 years) in ultrasound-guided RA following standardized teaching protocols. One of the researchers delivered both the traditional and simulation-based interventions, whereas the second researcher acted as an independent evaluator and remained blinded to group allocation. To minimize bias, group assignment was randomized using a table generated by an independent researcher in Microsoft Excel. A double-blind methodology was implemented; neither the participants nor the evaluator had knowledge of the group allocation or performance outcomes until the end of the study.

This study investigated the simulation of interscalene, supraclavicular, and sciatic nerve blocks, including the external and internal popliteal divisions of the nerves, as they appear as Plan A and Plan B blocks in the Regional Anaesthesia UK guidelines. An Esaote MyLab Alpha ultrasound device, identical to that used to generate simulation ecosystem images, was used in the experiment. Two individuals with plexuses that were confirmed to have adequate ultrasound visibility before testing took part in the study. All participants were scheduled for an assessment 2 days after their simulation training with the HUSP simulator. The assessment required participants to identify plexuses in 3 live individuals. To ensure impartiality in student group allocations, a blind investigator who was not involved in providing the theoretical lesson or guiding the simulation ecosystem practice administered the tests. The time taken by each student to locate the plexus was recorded, and successful identification was noted (maximum identification time: 120 seconds). Any attempt exceeding 120 seconds was considered unsuccessful in locating the nerve plexus.

The primary outcome variable was the success rate of ultrasound-guided peripheral nerve identification, recorded as a binary outcome (success or failure). The secondary outcome was the time (in seconds) required to complete each nerve localization procedure. The procedures were analyzed across 4 anatomical approaches: interscalene, supraclavicular, popliteal, and sciatic nerve division. The academic year of the students (fourth, fifth, or sixth year) was also included.

All data are expressed as the mean and SD. Descriptive statistics were used to summarize the participants’ demographic characteristics (age, sex, and academic year) and procedure times (in seconds). Categorical variables were reported as absolute frequencies and percentages. Data normality was assessed by visually inspecting histograms. Additionally, the Shapiro-Wilk test was performed to formally evaluate the normality of the data. Given the pilot nature of this study and the limited number of participants per group, an intention-to-treat analysis was performed [19]. The primary outcome, success in nerve identification, was evaluated using the chi-square or Fisher exact tests to compare the proportions between the traditional and simulation-based teaching groups for each anatomical approach. For the secondary outcome (procedure time), mixed linear models were applied to assess the mean differences between groups, including the variable “academic year” as a covariate. This allowed for the evaluation of potential confounding and interaction effects on the results of the study. Interaction terms (group × time) were tested to explore whether the effectiveness of the intervention varied over time [19]. Unstandardized coefficients, 95% CIs, and P values were reported for all regression models. A significance level of P<.05 was considered statistically significant. All statistical analyses were conducted using the JASP software (version 0.16.3).

A cohort of 34 students from the Faculty of Medicine at the University of Salamanca (medical degree program) was recruited for this study. All students provided voluntary consent to participate (Figure 4).

The cohort comprised students in the fourth, fifth, and sixth years of the medical degree program. Comprehensive demographic data, including course year, sex, and age, are presented in Table 1. Furthermore, detailed information stratified by course year and duration of each ultrasound procedure is provided in Table 2.

The results revealed no significant differences in successful nerve localization across the various approaches between the traditional teaching and simulation groups. Specifically, in the interscalene approach, 100% (17/17) of the students in the traditional teaching group successfully located the brachial plexus compared to 82% (14/17) in the simulation group, where 18% (3/17) of the students did not succeed. However, no statistically significant difference was observed (P=.07). Similarly, in the supraclavicular approach, both groups demonstrated comparable success, with 18% (3/17) of the students in each group unable to locate the relevant structure. In popliteal sciatic nerve localization, slightly fewer students in the simulation group (1/17, 6%) failed than in the traditional teaching group (2/17, 12%); however, this difference was not statistically significant (P=.07). Finally, for sciatic nerve division, all students in the simulation group successfully identified the division, whereas 12% (2/17) of the students from the traditional group, one from the fourth year and one from the fifth year, were unable to do so.

Specifically, there were no statistically significant differences based on the educational interventions received. The reported values correspond to unstandardized coefficients (β), which represent the effect of each intervention in the original units of the outcome variable along with their P values and 95% CIs. Both educational interventions had positive impacts: the simulator group achieved β of 106.72 (95% CI 10.06-203.37; P=.03), whereas the traditional teaching group achieved β of 102.80 (95% CI 5.03-200.57; P=.40; Figure 5A). The academic year covariate in the regression model did not interact with the effect of the educational interventions (β=−10.43, 95% CI −29.79 to 8.91; P=.28).

The analysis revealed no statistically significant differences based on the educational methods used. Both intervention groups (traditional education and simulator-based education) exhibited positive effects. The simulator-based group achieved β of 79.38 (95% CI 24.78-164.73; P=.01), whereas the traditional teaching group achieved β of 74.36 (95% CI 9.03-145.23; P=.03; Figure 5B). In the regression model, the academic year covariate did not interact with the effect of the educational interventions (β=−8.69, 95% CI −22.39 to 4.52; P=.14).

The analysis demonstrated no statistically significant differences based on the educational methods. The simulator-based group attained β of 98.93 (95% CI 32.54-188.50; P=.07), whereas the traditional teaching group achieved β of 105.46 (95% CI 27.59-195.32; P=.01; Figure 5C). In the regression model, the academic year covariate did not interact with the effect of the educational interventions (β=−12.63, 95% CI −28.91 to 1.33; P=.08).

The analysis demonstrated statistically significant differences based on the educational method used. The simulator-based group attained β of 60.72 (95% CI −3.63 to 125.07; P=.06), whereas the traditional teaching group achieved β of 74.67 (95% CI 9.57-139.77; P=.02; Figure 5D). Consequently, the traditional educational intervention demonstrated a statistically significant effect, whereas the simulator-based intervention did not reach statistical significance. In the regression model, the academic year covariate did not interact with the effect of the educational interventions (β=−10.22, 95% CI −23.11 to 2.66; P=.08).

This study evaluated the effectiveness of a simulation-based educational tool (HUSP) compared to traditional teaching for the acquisition of RA competencies in undergraduate medical students. Overall, no statistically significant differences were observed between the 2 groups regarding successful nerve localization across most approaches. In the interscalene approach, 100% (17/17) of the students in the traditional group successfully identified the brachial plexus compared with 82% (14/17) in the simulation group. Similar results were observed for the supraclavicular and popliteal approaches. In the sciatic nerve division approach, all students in the simulation group succeeded, whereas 12% (2/17) of the students in the traditional group failed, reflecting a statistically significant advantage of simulation-based training over traditional teaching in this specific case. No significant interaction was observed between the academic year and intervention type.

These findings align with those of previous studies supporting the use of simulation technologies for teaching ultrasound-guided procedures [20-22]. Simulation allows for safe and standardized practice of complex procedures and has been shown to improve knowledge retention and practical skill acquisition among medical students [23]. In this study, the HUSP simulator enabled students to reach level 2 of the Kirkpatrick evaluation model, focusing on the acquisition of knowledge and skills following educational intervention. Specifically, students learned to identify the relevant anatomical landmarks for nerve block procedures using ultrasound. Achieving higher levels of clinical competence requires complementary training involving needle use, as found in cadaver laboratories or high-fidelity simulators. During this study, an additional software module integrating needle simulation was developed, enhancing the potential of the HUSP simulator to support advanced skill acquisition.

Chen et al [2] demonstrated the effectiveness of simulation-based learning and emphasized the need for clarification of optimal simulation modalities. Other studies have reported that using live models or fresh cadaver limbs in elective courses can improve anatomical understanding and provide a valuable clinical context [24]. The HUSP simulator offers an effective alternative, especially when access to live models or cadavers is limited in the clinical setting. Its customizable design allows educators to create tailored simulation exercises, which is a pedagogical advantage. The recent integration of needle simulation hardware has further improved clinical applicability. In terms of cost-effectiveness, the simulator is affordable; the hardware is low cost and compatible with any PC, with the software license being the main investment [25]. These characteristics make it a viable educational solution, particularly in resource-constrained settings [26].

A key strength of this study is its contribution to improving ultrasound education during the preclinical stage by addressing gaps identified in previous research [27]. Despite the growing interest in integrating ultrasound into undergraduate curricula, previous studies have faced limitations such as restricted access to resources, faculty availability, and methodological rigor [28]. The development of a simulator incorporating needle simulation represents a significant step toward comprehensive interventional training for the HUSP system. This study has several limitations. First, the relatively small sample size, dictated by the fixed academic cohort, may limit the generalizability of the findings, a common challenge in similar educational studies [15]. Second, the lack of blinding and the dual role of some authors as both instructors and researchers could have introduced bias. Furthermore, an intention-to-treat analysis was performed, ensuring that all participants were analyzed in their originally assigned groups regardless of protocol adherence, thereby minimizing bias and better reflecting real-world educational conditions.

Simulation-based learning is an effective tool for supporting the acquisition of competencies in RA and offers a practical, safe, and scalable alternative to traditional educational methods. Beyond individual skill acquisition, integrating simulators such as HUSP into medical curricula could standardize training, reduce variability in learner performance, and enhance patient safety by preparing students for clinical practice in a controlled environment. To strengthen the evidence and broaden its applicability, further multicenter studies with larger and more diverse cohorts are warranted, which may inform best practices and guide the implementation of simulation-based approaches in medical education programs.