Low-Cost Simulator Proves It Can Train Doctors to Place Subclavian Catheters Safely

Every year, clinicians around the world insert millions of central venous catheters, the thin flexible tubes threaded into the large veins of the chest and neck that deliver medications, nutrition, and fluids to critically ill patients. The subclavian approach, which enters the vein just below the collarbone, is prized for patient comfort and lower infection risk, but it is also technically demanding, with a real risk of striking the lung or an artery on the way in. Teaching that procedure safely has long been a dilemma for medical educators, because the first time a trainee performs it should never be on a vulnerable patient. A new study from the Central University of Venezuela now offers encouraging evidence that an inexpensive, non-living training model can reliably distinguish skilled surgeons from beginners, meeting a key scientific standard for simulation-based education.

The research, published in Global Surgical Education, the journal of the Association for Surgical Education, set out to gather what psychometricians call validity evidence for a low-cost inanimate simulator designed for subclavian central venous catheterization training. Rather than simply asking whether trainees liked the model, the authors, Luz Galvis-Arenas and Omaira Rodríguez of the Department of Surgery in Caracas, structured their investigation around Messick’s framework, the modern conceptual foundation of educational measurement. Under this framework, validity is not a property of a test or a simulator itself but of the interpretations and actions built on its scores. A simulator earns its place in a curriculum only if its scores actually mean what educators claim they mean, such as the claim that a high checklist score reflects genuine procedural competence.

To test that claim, the team ran a cross-sectional validation study with twenty participants divided into two starkly different groups. Ten were sixth-year medical students with no prior experience in central venous catheterization, forming the novice cohort, and ten were experienced general surgeons who constituted the expert group. The logic behind this design, known as the known-groups approach, is straightforward and powerful: if the simulator and its scoring system are truly measuring procedural skill, then experts should systematically outperform novices. A model on which beginners score as well as veterans would be blind to the very skill it is supposed to capture, no matter how realistic it looks or feels.

The study examined three strands of evidence. First, content validity was assessed by surveying experts about whether the model adequately represents the essential elements of subclavian catheter placement, quantified through the Content Validity Index, a metric that reflects the proportion of raters agreeing that each element is relevant. Second, technical performance was scored with a thirteen-item procedural checklist, completed independently by two evaluators who watched each attempt, allowing the researchers to calculate interrater reliability using the kappa statistic. Third, construct validity was examined through the comparison of experts and novices across a battery of outcomes, including checklist scores, procedure duration, first-attempt success, the proportion of participants completing the procedure without assistance, and the occurrence of simulated complications.

The content validity results were as strong as this type of evidence can be. Every item on the expert survey achieved an item-level Content Validity Index of 1.00, and the scale-level average matched that perfect figure, indicating unanimous agreement among the consulted experts that the model covers the essential steps of the procedure. Interrater reliability was equally remarkable: the two independent evaluators assigned identical checklist scores to every performance, yielding a kappa of 1.00. Perfect agreement between raters is an ideal outcome for any scoring instrument, because it means the checklist items are described with enough clarity that subjective interpretation plays essentially no role in the grade a trainee receives.

One statistical nuance deserves attention. The internal consistency of the checklist, measured with the KR-20 coefficient, came in at a moderate 0.63, and the authors noted ceiling effects in several fundamental steps. In plain terms, most participants, including many novices, completed the basic items correctly, so those items could not strongly differentiate performance levels. This is a common and partly unavoidable feature of procedural checklists, since fundamental steps such as hand preparation and equipment assembly are quickly mastered by most trainees. A moderate internal consistency coefficient is not a fatal flaw, but it signals that the checklist may be most informative about overall procedural conduct rather than fine-grained differences in technical finesse, and it suggests room for refinement of the instrument in future versions.

The construct validity results were decisive. Experts achieved a mean checklist score of 100.0 percent, compared with 85.4 plus or minus 10.5 for the novices, a difference that reached statistical significance with a p value of 0.002. Experts were also dramatically faster, completing the procedure in an average of 5.28 plus or minus 0.79 minutes against 11.61 plus or minus 4.67 minutes for the students, again with p equal to 0.002. First-attempt success rates told the same story: 70 percent of experts succeeded on their initial attempt, while only 10 percent of novices did so, a difference with a p value of 0.020. Autonomy diverged even more sharply, with every expert completing the placement independently compared with just 10 percent of the students, a result significant at p equal to 0.0003.

Complication data added a clinical dimension to the statistical picture. No complications occurred among any of the expert attempts, whereas the novice group generated simulated complications in 30 percent of their attempts. Although these were simulated events on an inanimate model, the pattern mirrors real-world epidemiology, where mechanical complications of central venous catheterization, including arterial puncture and pneumothorax, are concentrated among less experienced operators. The fact that the simulator reproduces this gradient of adverse events suggests it captures not only the choreography of the procedure but also the consequences of technical error, which is precisely what makes it valuable as a rehearsal space before students face actual patients.

The implications extend well beyond one medical school in Caracas. Central venous catheterization is among the most commonly performed invasive procedures in hospitals worldwide, and meta-analyses of simulation-based education have linked structured simulator training to fewer catheter-related complications in clinical practice. Studies using simulation-based mastery learning have shown measurable reductions in intensive care unit catheter complications, and multicentre audits have documented the ongoing toll of mechanical injuries even in the ultrasound-guided era. Against that backdrop, the appeal of a low-cost, non-living model is obvious. Commercial high-fidelity mannequins can be prohibitively expensive for institutions in low- and middle-income countries, where the burden of preventable procedural harm is often greatest. Simple, affordable simulators such as gelatin phantoms and assembled low-cost models have been explored by several groups, and this study adds rigorous validity evidence to that movement, showing that cost-cutting need not come at the price of measurement quality.

The Venezuelan study also carries a methodological lesson for the broader field of surgical education. A systematic review of validity evidence in surgical simulation has documented how inconsistently simulators are evaluated, with many devices adopted on plausibility alone. By grounding the evaluation in Messick’s framework, quantifying content validity, verifying interrater reliability, and applying a known-groups design, the authors modeled the kind of disciplined assessment that should precede any simulator’s integration into a training curriculum. The study was approved by the Bioethics Committee of the Instituto Autónomo Hospital Universitario de Caracas, and all participants gave written informed consent. The authors declared no competing interests and received no specific external funding. As health systems worldwide continue to shift invasive skills training away from real patients and toward the simulation laboratory, evidence like this helps ensure that the tools awaiting trainees there are not just cheap and convenient, but demonstrably capable of measuring what matters: the difference between a procedure done right and one that goes dangerously wrong.

Subject of Research: Validity evidence for a low-cost inanimate simulation model used in subclavian central venous catheterization training

Article Title: Validation of an inanimate simulation model for subclavian central venous catheterization training

Article References: Galvis-Arenas, L., & Rodríguez, O. (2026). Validation of an inanimate simulation model for subclavian central venous catheterization training. Global Surgical Education – Journal of the Association for Surgical Education, 5(1), Article 158. https://doi.org/10.1007/s44186-026-00556-w

Image Credits: AI Generated

DOI: 10.1007/s44186-026-00556-w

Keywords: central venous catheterization, simulation-based education, subclavian approach, validity evidence, procedural skills training, medical education, surgical simulation, Messick framework, construct validity, low-cost simulator, patient safety, checklist assessment