Simulation training for invasive cardiovascular procedures: the Heart-SIMS-1 randomized trial

Abstract

Backgrounds

Interventional cardiology training has a long learning curve, with potential procedural risks to patients and clinicians. We aimed to assess whether mentored simulation-based training with 3D-printed models can improve the skills of beginners in coronary diagnostic procedures in a pilot randomized trial.

Methods

Twenty-nine final-year medical students recruited from a single University were lectured on the fundamentals of invasive coronary angiography (ICA) for one-hour, and then randomized to conventional or simulation training. Conventional training (n = 15) consisted of watching a 20-minute video demonstrating ICA steps performed in a 3D-printed coronary simulator. The simulation training group (n = 14) were offered, in pairs, the same content in a 20-minute hands-on session using a 3D-printed simulator. The co-primary endpoint was efficacy and safety of performing a simulated ICA in the angiography suite. Efficacy and safety were graded using a 13-point procedural checklist (0-100%) and the identification of five procedural “red flags” items, respectively. The secondary endpoint was theoretical knowledge (multiple-choice test).

Results

All participants completed the protocol. In both components of the co-primary endpoint, the simulation group scored higher: efficacy score of 91.5 ± 3.8% vs. 64.6 ± 8.3% (mean difference 95% CI [20.8, 30.8]) and safety score 100.0% (100.0-100.0%) vs. 62.5 (20.8–79.2%) (median difference 95% CI [20.8, 79.2]), p < 0.001. The median number of “red flags” were 2 (1–4) in conventional and 0 (0–0) in simulation training (p < 0.001). Also, simulation group obtained a higher score in the theoretical knowledge test: 85.7 ± 9.0% vs. 76.8 ± 12.7%, p = 0.039.

Conclusion

Mentored simulation-based training using 3D-printed simulators significantly improved theoretical knowledge and basic procedural skills of ICA. These results suggest that simulation-based training should be pursued for improving patient safety and technical proficiency.

Trial registration

NCT06224101.

Graphical abstract

Final-year medical students randomised to oficient in performing simulated invasive coronary angiography (efficacy and safety) and exhibited superior theoretical knowledge, compared to traditional teaching.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12909-026-08831-6.

Introduction

Classical medical teaching usually consists of formal theoretical sessions (grand lectures) and practical training in smaller groups with patients. Conventionally, for invasive procedures, an apprenticeship model with the principle of repetitive experiential learning though practice has been utilised for training [1]. However, this approach can be limited by time and accessibility to training.

Simulation training may help to overcome some of these barriers. The success of simulation training in the aviation industry has inspired its application in other fields, including healthcare, whereby simulation of real-world scenarios, particularly for invasive procedures, may also become indispensable in training individuals to effectively and safely manage complex situations [2, 3]. Effective clinical skills training requires educational strategies that support learners in progressing from basic competence to more advanced performance through guided practice, according to the Zone of Proximal Development (ZPD) and scaffolding concept.

Invasive coronary angiography (ICA) is a common cardiac procedure that requires extensive training with a long learning curve to achieve both technical proficiency and procedural safety.

Limited data exists evaluating the benefit of simulation training for ICA [4–11]. Most studies have focused on senior cardiology residents and interventional fellows, assessing whether virtual simulation training can translate into improved performance in the catheterization laboratory. When comparing fellows pre-exposed to virtual simulators (n = 12) versus controls (n = 20), Prenner and co-authors reported shorter procedure time (23.98 min vs. 24.94 min, p = 0.03) and significant less radiation exposure (56,348 mGycm² vs. 66,120 mGycm², p < 0.001) in the former [4]. Popovic and colleagues randomized 20 cardiology residents to simulation-based training or standard practice (1:1). The procedure time was shorter, the radiation dose lower, and the global procedure skill was higher in the simulation group (p < 0.002) [5] Similarly, virtual simulation-based training translated to superior performance scores, assessed by fluoroscopy duration and total procedure time, in a Swedish study including 16 senior cardiology residents [8]. Wolfram Voelker et al. randomized 18 cardiology fellows to virtual reality simulation training (n = 9) or controls (n = 9). The assessment was performed in a simulated procedure using a pulsatile coronary flow 3D-printed model in a catheterization laboratory, showing superior skill score of the simulation group [9]. Fischer and co-authors recruited 118 medical students and randomized them to virtual simulation ICA training or traditional power-point lectures. The former group showed superior performance scores in a multiple-choice test (59.5 vs. 43.7, p < 0.01) [11].

Although these studies suggest that simulation training in ICA is a promising novel learning tool, all of them have employed virtual simulators, which are limited by their higher costs, lack of tactile feedback and most importantly lack of compatibility with interventional equipment used in vivo. In contrast, custom-made three dimensional (3D) printed simulators fully compatible with standard catheterization laboratory tools allow diagnostic and interventional coronary procedures to be replicated in the catheterization laboratory [12, 13]. There are no randomized trials exploring the role of 3D-printed models in interventional cardiology training for beginners.

We designed The Simulation Training for Invasive Cardiovascular Procedures - Heart-SIMS-1 pilot randomized trial to assess whether mentored simulation-based-training with 3D-printed simulator models can improve the procedural skills and safety performance of beginners in ICA compared to conventional training, while monitoring for potential harms such as procedural errors or complications.

Methods

Trial design and population

Final-year medical students were recruited between December 2023 and January 2024 from a single site at the Faculty of Medicine of the University of Coimbra, Portugal, and invited to participate in the study after providing signed informed consent (supplementary data). Exclusion criteria were pregnancy and prior experience in interventional cardiology.

We designed a parallel-group, randomized controlled trial with a 1:1 allocation ratio, comparing simulation-based training with conventional training after a run-in phase. The trial was designed with a superiority framework to determine whether simulation-based training improves procedural skills and safety performance in novice ICA operators. (Fig. 1). No changes were made to the trial after commencement, including outcomes or analyses, and all were pre-specified. Outcomes of benefits and harms were assessed immediately upon completion of the intervention, and no additional follow-up period was conducted.

Fig. 1.

Heart-SIMS-1 workflow

The trial protocol was drafted in accordance with Good Clinical Practice guidelines and the principles outlined in the Declaration of Helsinki, registered at ClinicalTrials.gov (NCT06224101; https://clinicaltrials.gov/study/NCT06224101) on January 14, 2024 and approved by local authorities: the Ethics Committee of the Faculty of Medicine of the University of Coimbra (OBS.SF.052-2023). It was funded by the non-profit organization Associação para o Desenvolvimento e Investigação em Cardiologia de Coimbra [The Coimbra Association for Development and Research in Cardiology].

No patients or members of the public were involved in the design, conduct, or reporting of the trial.

We used the CONSORT reporting guideline [14] to draft this manuscript, and the CONSORT reporting checklist [15] when editing, included in supplement.

Simulator

The custom-made 3D-printed simulator SimulHeart^® (3D CardioSolutions, Coimbra, Portugal) was used in Heart-SIMS-1 (Fig. 2, panels A and B). It consists of a clear acrylic tank filled with water in which standard 3D-printed vascular anatomical structures are inserted (aorta and coronaries). The whole vascular structure was connected to a simulated right radial access, and to a pulsatile pumping system that simulates the arterial pressure wave. It enables the use of standard diagnostic and interventional devices with realistic haptics feedback. It was used with a video camera connected to a monitor during training sessions and using ionizing radiation imaging in the catheter laboratory for final assessment. The SimulHeart^® was recently validated in terms of content and appearance in simulated coronary procedures [13].

Fig. 2.

Panels (A) and (B): Setup for simulation training, including the simulator SimulHeart^® (red arrow), with 3D printed aorta, coronary arteries, and left ventricle. Panel (C): Evaluation setting: simulator under sterile drapes at the angiography suite (red arrow) and the Candidate for evaluation (Ca) and Second Operator (SO)

Training sessions

The training session on ICA was performed at an academic simulation facility: Centro de Simulação Biomédica dos Hospitais da Universidade de Coimbra (Coimbra, Portugal). All participants attended a theoretical lecture covering the basic concepts of coronary anatomy, clinical indications, procedural steps, equipment, and safety considerations of ICA. The session was prepared by an interventional cardiologist and had a 60-minute duration.

Subsequently, students were randomized using a computer-generated random sequence implemented in Python code in a 1:1 ratio to two training methodologies: Group 1 - conventional training and Group 2 - simulation training. No stratification or blocking was applied, and allocation was not restricted. The random allocation sequence was stored in a secure, password-protected file. Personnel who enrolled participants did not have access to the sequence and assigned interventions only at the time of allocation, ensuring that the sequence remained concealed until interventions were assigned.

Group 1 (15 students) watched a video recording of a training session on ICA utilizing SimulHeart^®. The video consisted of interventional cardiologists describing and performing all the steps necessary to perform trans-radial diagnostic coronary angiography of left and right coronary arteries. In addition to the technical steps, key safety points were emphasized, and the sequence was repeated three times. Total video duration was 20 min.

Participants randomized to Group 2 (14 students) were trained with the SimulHeart^® in groups of 2 participants per session (Fig. 2). The simulation training session took place outside of the catheterization laboratory, without radiation exposure. Two interventional cardiologists trained the students following the exact same steps covered in the video session using the same catheters and wires as a standard ICA procedure: the diagnostic catheters consisted of a Judkins Left JL3.5 and a Judkins Right JR4 and the wire was a J-tip 0.035 150 cm wire. The hands-on simulation sessions were limited to 20 min for each pair of students.

Evaluation

Following the training session, students’ knowledge was assessed with a theoretical exam and a practical evaluation. The theoretical test assessed fundamental knowledge of anatomy, clinical indications, procedural steps, and safety considerations through a 25-question multiple-choice test, graded 0 to 100% (supplementary data).

Practical assessment was conducted in a catheterization laboratory at Unidade de Intervenção Cardiovascular, Serviço de Cardiologia da Unidade Local de Saúde de Coimbra (Coimbra, Portugal). The simulator SimulHeart^® was installed under sterile drapes in the angiography table to mimic a real procedure (Fig. 2) and a sheath was pre-inserted in the simulated radial artery. Students were therefore blinded to the type of simulator being used for the assessment and had only direct visualization of the radial sheath. Each student was instructed to perform diagnostic left and right coronary angiography within an allocated procedural time of 20 min. An interventional cardiologist blinded to the randomization group acted as second operator controlling the fluoroscopy and aiding with catheter and wire handling prior to being inserted in the simulator. If the student was unable to complete each of the prespecified procedural steps in less than two minutes, the second operator would take control and complete that step moving forward to the next step under evaluation. Fluoroscopy protocol was optimized to the minimum dose possible, at four frames per second, and all students were wearing appropriate protective equipment including lead body aprons and thyroid shields. An investigator blinded to the randomization and training sessions allocation rated participants’ performance according to a pre-specified checklist of efficacy and safety endpoints (supplementary data).

Endpoints

The two co-primary endpoints were the efficacy and safety of performing ICA. Efficacy score (percentage) was graded by evaluating 13 steps in ICA performance, adapted from Chaer R. et al. [16]. For each step, the investigator graded the student with 0 (absent), 1 (incomplete) or 2 points (step completed) (ex: advancing the wire to the aortic root, engaging the right coronary artery; the full list may be consulted in the provided supplementary data). Safety score (percentage) was calculated according to the identification of five procedural “red flags” (inadequate purge, advancing catheter without wire, failure to assess the pressure curves, procedural duration exceeding 20 min, advancing the wire without disengaging the catheter), which were used to systematically define and assess potential harms during the procedure. Score was higher in the absence of safety concerns.

The secondary endpoint was the result of the theoretical test (0-100).

Additionally, data on procedural duration, fluoroscopy time and radiation dose were collected.

Statistical analysis

Continuous variables were reported using mean ± standard deviation or median (interquartile range), according to the normality of the distribution. Categorical data were presented by frequency and proportion. The difference between groups in primary and secondary endpoints was analysed with T-student or Mann-Whitney U tests. Proportions were compared using Fisher’s exact test. To convey the precision of estimated effects, the 95% confidence intervals (CI) of the mean difference were provided. For non-parametric comparisons, the Hodges–Lehmann median difference with 95% CI was reported, as the Mann–Whitney U test itself does not estimate a mean difference by default. No knowledge pre-testing was performed before randomization, because medical students had minimal prior knowledge on ICA. There was no sample size calculation for this pilot study. No interim analyses or stopping guidelines were planned for this trial. IBM Statistical Package for the Social Sciences, v29.0 (SPSS) was employed for statistical analysis. A post-hoc power analysis was conducted using the using G*Power 3.1.9.2 software (Erdfelder, Faul, & Buchner, 1996).

Results

Twenty-nine final-year medical students (median age 23 years old, 58.6% female) were included in the study, 15 in Group 1 (conventional training) and 14 in Group 2 (simulation-based training). All participants completed the theoretical and practical evaluation phase. All randomized participants were included in the analysis according to the group to which they were originally assigned. There was no crossover between groups, no missing data, and no participants were excluded from the analysis.

Both elements of the co-primary endpoint were superior in the simulation group: efficacy score 91.5 ± 3.8% vs. 64.6 ± 8.3% (mean difference 95% CI [20.8, 30.8]) and safety score 100.0% (100.0-100.0%) vs. 62.5% (20.8–79.2%) (median difference 95% CI [20.8, 79.2]), p < 0.001 for both. For the secondary endpoint of theoretical test results, the simulation group also scored higher: 85.7 ± 9.0% vs. 76.8 ± 12.7% (mean difference 95% CI [0.5, 17.3], p = 0.039). There were no significant differences in procedural duration, fluoroscopy time or radiation dose between the two groups (Table 1; Fig. 3).

Table 1.

– Baseline characteristics and evaluation data

	All (n = 29)	Group 1 Conventional training (n = 15)	Group 2 Simulation training (n = 14)	[95% CI] / p-value
Age	23 (23–24)	23 (23–24)	23 (23–24)	[-1, 1]0.983
Female gender	17 (58.6%)	6 (40%)	11 (78.6%)	[1.7–28.4]0.060
Co-primary endpoint
Efficacy score (%)	78.1 ± 14.6	65.6 ± 8.4	91.5 ± 3.7	[20.8, 30.8] < 0.001
Safety score (%)	79.2 (52.1–100.0)	62.5 (20.8–79.2)	100.0 (100.0-100.0)	[20.8,79.2] < 0.001
Secondary endpoint
Theoretical test results (%)	81.1 ± 11.7	76.8 ± 12.7	85.7 ± 9.0	[0.5, 17.3] 0.039
Exploratory endpoints
Procedure time (min)	14.9 ± 3.3	15.5 ± 2.8	14.2 ± 3.7	[-3.9, 1.1] 0.272
Fluoroscopy time (min)	7.2 ± 2.3	7.4 ± 2.6	7.0 ± 2.1	[-2.2, 1.4] 0.666
Radiation dose
DAP (microGy.m2)	281.0 (149.0-593.5)	281.0 (166.0-581.0)	301.5 (126.5-639.8)	[-150.0, 150.0] 0.780
Air kerma (mGy)	30.0 (14.5–51.0)	30.0 (16.0–50.0)	29.0 (12.8–55.0)	[-10.0, 12.0] 0.813

CI mean difference confidence interval, DAP Dose Area Product

Fig. 3.

Co-primary endpoints (efficacy and safety) and secondary endpoint (theoretical assessment) according to trial arm

Table 2 displays relevant procedural details, namely the ability to cannulate the coronary arteries and the presence of safety concerns. Group 2 showed higher ability to cannulate the right and left coronary arteries. Group 1 (conventional training) had a higher proportion of all safety red flags: procedural duration over 20 min (73% vs. 7%), inadequate catheter purge (27% vs. 0%), failure to assess pressure curves (73% vs. 7%), advancing catheter without wire (47% vs. 0%) and advancing wire without disengaging catheter (53% vs. 0%). Figure 4 illustrates the number of red flags according to trial arm allocation, that was significantly higher in Group 1 (p < 0.001). Figure 5 depicts the efficacy score per subject.

Table 2.

– Procedural details

	All (n = 29)	Conventional training (n = 15)	Simulation training (n = 14)	95% CI / p-value
RCA cannulation	10 (34.5%)	2 (13.3%)	8 (57.1%)	[0.01–0.72] 0.021
LCA cannulation	9 (31%)	2 (13.3%)	7 (50%)	[0.03–0.95] 0.05
Both	3 (10.3%)	1 (6.7%)	2 (14.3%)	[0.03–5.33] 0.59
Procedural safety concerns
Duration > 20 min	12 (41.4%)	11 (73.3%)	1 (7.1%)	[0.01–0.29] < 0.001
Inadequate purge	4 (13.8%)	4 (26.7%)	0 (0%)	[1.01–1.85] 0.1
Failure to assess pressure curves	12 (41.4%)	11 (73.3%)	1 (7.1%)	[0.00-0.29] < 0.001
Advancing catheter without wire	7 (24.1%)	7 (46.7%)	0 (0%)	[1.17–3.01] 0.006
Advancing wire without disengaging catheter	8 (27.6%)	8 (53.3%)	0 (0%)	[1.28–3.68] 0.002
Number of safety concerns	1 (0–2)	2 (1–4)	0 (0–0)	[1–4] < 0.001

RCA right coronary artery, LCA left coronary artery, CI mean difference confidence interval

Fig. 4.

Number of safety red flags by trial arm, p < 0.001 for the difference. Procedural “red flags”: duration exceeding 20 min; inadequate purge; failure to assess pressure curves; advancing catheter without wire and advancing wire without disengaging catheter

Fig. 5.

Efficacy score for each subject according to trial arm. Black bars: Group 1, control; blue bars: Group 2, simulation

The post-hoc power analysis yielded a power (1-ß) of 100% and 90% for the co-primary endpoints, respectively, and 55% for secondary endpoint.

Discussion

In this study we have completed the first randomized controlled evaluation of mentored simulation-based training using high-fidelity 3D-printed in coronary diagnostic procedures. The key findings of this study are that beginners exposed to simulation training: (1) were more proficient in performing procedural steps at the angiography suite, (2) had superior safety scores and (3) had an improvement in theoretical knowledge when compared to traditional teaching methods (Graphical abstract).

It has previously been shown that virtual simulation training reduces procedural time [4–8], lowers radiation [4–7] and contrast dosage [7], and improves overall performance [5, 6, 9, 10].

Whilst high-fidelity virtual reality simulators offer excellent evidence-based training opportunities their utilization is limited by high costs and low availability outside academic centres. Additive manufacturing technology presents an opportunity to overcome these limitations, enabling cost-effective production of realistic models [13]. In fact, 3D printed simulators offer several advantages: inexpensive production and maintenance, modular design that enable training in virtually any vascular anatomy, radiation-free usage and haptic sensations resembling real-life procedures [17]. Furthermore, despite having been primarily utilised to plan complex patient-specific interventional procedures [12, 18–20], they may easily be adapted to facilitate mentored-based simulation sessions for educational purposes. In this publication, we present the results of a coronary model simulation for ICA, but owing to the modular design they can also be used for coronary intervention or structural procedures.

The simulation training session in Heart-SIMS-1 was conducted in a radiation-free environment, using a custom-made 3D printed coronary model and matching the duration of conventional training. Contrary to other studies with a prolonged simulation programme, there was only one short and comprehensive training session. This was because we were interested in assessing the value of focused simulation training in basic coronary angiography in naïve subjects. In fact, simulation training can only play a wider role in cardiovascular training if it can compete with current practice in a time efficient manner.

There is a wealth of data in the value of simulation for teaching healthcare professionals across several fields of knowledge. Simulation sessions potentiate the learning process by enhancing procedural skills, knowledge retention, and students’ motivation.

Simulation-based training may improve the hand-eye coordination and therefore the procedural skills of novices [9]. It has the potential to reduce the early part of the learning curve, mitigating risks and facilitating learning with increased safety for patients [21]. In our study, there was a superior efficacy in completing the procedure in the simulation group, likely signalling increased coordination and muscle memory. This advantage of simulation-based training has long been recognised in the surgical field, in a sense that it fosters the development of the “pre-trained novice,” someone with automated psychomotor skills and spatial judgments, enabling a heightened focus on interventional strategy and prevention of complications [22]. In fact, the simulation group in Heart-SIMS-1 had a striking superiority in safety, with marked reduction in the number of “red flag” manoeuvres. This was surprising, as both the theoretical lecture and the conventional training video session had a very detailed description of the safety steps and requirements, underscoring the importance of the simulation session. Despite the superior procedural skills overall, we found no significant improvement in procedure time, fluoroscopy time or radiation dose. By protocol, there was a time limit of two minutes for each step, which could explain the lack of superiority in simulation group. Nonetheless, this lack of difference may also be attributed to an increased risk awareness, which may have triggered the simulation group participants to exercise greater caution.

The Zone of Proximal Development (ZPD) and scaffolding play pivotal roles in the acquisition of skills like coronary angiography, especially when utilizing simulator models to train final year medical students. The ZPD concept emphasizes the range between a learner’s current abilities and their potential with guidance, underscoring the importance of tailored instruction to optimize learning outcomes [23]. Scaffolding, as proposed by Wood, Bruner, and Ross, involves providing temporary support to learners within their ZPD to facilitate skill development [24]. In the context of coronary angiography training, educators can offer structured guidance and feedback which is enhanced by the experiential learning on the simulator. This supports students to gain proficiency but more importantly to act safely. Applying these theories to simulator-based learning enables students to engage in authentic practice while receiving timely assistance, fostering both competence and confidence in mastering complex medical procedures. While this study lacked sufficient power for secondary endpoint analysis concerning theoretical knowledge, we did find an improvement in the theoretical test results. Simulation might impact the type of memory retention process. Medical students trained with simulation sessions compared to traditional lectures have the same results in immediate testing but perform better in delayed tests (5 weeks). Conventional lectures stimulate explicit memory, which is a recollection of previously attained information. In contrast, it is hypothesised that simulation potentiates the development of implicit memory, from subconscious thought processes, leading to higher knowledge retention in the follow up [25]. In addition, 3D visualization might aid in the learning process of coronary anatomy and fluoroscopy projections, as recently shown in a study enrolling 34 cardiology residents exposed to 3D-printed coronary models [26]. The medical knowledge of healthcare professionals on congenital heart disease was improved with the use of 3D models for teaching, versus traditional classes, in a recent randomised trial [27].

An advantage of simulation training over conventional classes is the active involvement of students in the learning process, improving motivation. The training session presented a technical challenge in an informal setting, resembling a game. Adjedj J. and co-authors reported that a playful environment, such as a video game, encourages student’s participation and improves retention and satisfaction rate [28].

Our report contributes additional evidence to reinforce the importance of simulation in medical training. Despite the major obstacles to widespread adoption, including high costs, limited access to simulation centres and lack of a standardized curriculum, its integration has the potential to significantly enhance the knowledge and skills of future physicians [29]. There are several limitations in medical training that can be addressed with simulation. For instance, recent reductions in clinical shifts duration and mandatory rest after night work result in less instruction and fewer procedural opportunities for trainees [30]. In addition, radiation-free training in interventional cardiology poses an opportunity to address gender inequalities, as reported by a fellow that was able to maintain skills and proficiency in performing coronary intervention during maternal leave, due to regular training with simulators [31]. Moreover, simulation can serve as an assessment tool, establishing a standardized minimum proficiency level that should be achieved by all trainees.

Therefore, the cardiovascular community is eager to foster simulation training for complex procedures [32], and professional societies propose the inclusion of simulation programmes in the interventional cardiology core curriculum [33, 34].

One hopes that the expression “see one, do one, teach one” will evolve into the principle of “never the first time on a patient”, further advancing into “never a first time without simulation training” [29]. In that situation, simulation-based certification in interventional cardiology would commence before working with real patients, to accelerate the learning curve and ensure a proper shift to a scenario where the margin of error is narrow and not without medicolegal risk.

Limitations

As a pilot study, the sample size was small and the analysis, albeit pre-specified, were exploratory. Although a post-hoc power analysis was performed, its value is inherently limited when compared with that of an a priori powered study. As multiple outcomes were assessed, and no correction for multiple comparisons was performed, the findings should be interpreted with some caution regarding the potential for Type I error. The population was selected according to the research question; however, the overall simulation effect might have been amplified with medical students. In a trial of virtual reality simulation, Bagai A. et al. demonstrated that less proficient operators derived greater benefit from simulator training [10], suggesting that the effect size is potentially smaller the more experienced the individual. However, this reflects the potential benefit of simulation in shortening the learning curve and time to gain proficiency. Recognizing that cardiology residents and interventional cardiology fellows are the primary audience for ICA training, additional research in this population is warranted to more robustly establish the value of simulation-based education.

In an ideal setting, evaluation would involve assessing performance during a real ICA procedure to determine skill transfer to patient care, complemented by measures of decision-making under stress and teamwork ability. In addition to the obvious ethical concerns of such approach, the variability of patients’ anatomies would mandate a larger sample size. With our current protocol, all students were subjected to the exact same evaluation setting. Moreover, efforts were made to provide an immersive experience, with a high-fidelity simulator installed under surgical drapes, in a catheterization laboratory. Besides, it is already proven that skills acquired with simulation training are transferable to real life practice [35]. Recognizing the pilot nature of this study, we acknowledge that both procedural and knowledge assessment instruments employed have not yet been validated, weakening the conclusions. The results may be biased considering that the same simulator was used for training and evaluation, potentially favouring the simulation group; despite this, the conventional training group received very clear instructions in the video session, which used the same simulator and material for demonstration. Hence, the sole distinction between trial groups was the mode of information delivery: passive approach versus an active and engaging simulation session. Knowledge retention was not tested in the present study. As with most early-stage educational interventions, factors such as the Hawthorne effect, trainer-related variability, and peer-learning dynamics may have contributed to the observed outcomes and should be considered when interpreting the results.

Conclusion

In this single-centre randomised controlled trial, simulation-based training with a 3D-printed coronary simulator demonstrated superior efficacy and safety in a simulated coronary diagnostic procedure performed by medical students, as compared to conventional training. This study highlights the significant impact of simulation training using 3D-printed simulators on the acquisition of knowledge and competencies in interventional cardiology. It also suggests a broad range of potential applications in this field, warranting further investigation.

Abbreviations

3D

three dimensional

ICA

Invasive coronary angiography

LCA

Left coronary artery

RCA

Right coronary artery

ZPD

Zone of Proximal Development

Authors’ contributions

MOS, CG, BL, GPM, JSM and AK designed the study, undertook experiments, analyzed results and interpreted the data. JR, JG, MC and LG participated in study design. EOS and AVM contributed to the experiments. CG and AK rated practical assessment. JR reviewed the statistical analysis. MOS drafted the first and subsequent versions of this report with key input from all other authors, who reviewed and approved the final submitted report.

Funding

This study was supported by the non-profit Associação para o Desenvolvimento e Investigação em Cardiologia de Coimbra [The Coimbra Association for Development and Research in Cardiology]. The funders was not involved in the study design, conduct, analysis, or reporting.

Data availability

The trial protocol, statistical analysis plan, de-identified individual participant data, statistical code, and other relevant materials used and/or analysed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The trial protocol was drafted in accordance with Good Clinical Practice guidelines and the principles outlined in the Declaration of Helsinki, registered at clinicaltrials.org (NCT06224101) and approved by local authorities: the Ethics Committee of the Faculty of Medicine of the University of Coimbra (OBS.SF.052-2023). All the participants consented to be included in the trial.

Consent for publication

All the participants consented for the publication.

Competing interests

MOS, EOS and JSM are co-founders of 3D CardioSolutions, the company that developed SimulHeart^®.