Cardiovascular and interventional radiology
September 9, 2026 · via Europe PMC
Simulation and Virtual Reality as Training Modalities in Interventional Radiology: A Systematic Review.
Deshwal A, Temperley HC, Deshwal A, +8 more, Lee MJ.
AI summary — for doctors, generated from the abstract only
A systematic review of 18 studies found simulation and virtual reality training probably improve IR technical performance, especially radiation and contrast use, with low-certainty evidence of transfer to clinical practice.
What was studied
PROSPERO-registered systematic review of 18 studies (495 learners; 7 randomized trials) evaluating simulation or VR training in IR with objective outcomes, synthesized narratively with GRADE assessment.
Key findings
- All studies reported improvement on at least one metric
- Pre-post reductions: fluoroscopy time 50-72%, procedure time 27-53%, contrast volume 28-76%, radiation dose 31-71%; specific statistic not fully captured in the abstract
- Low-certainty evidence for transfer to clinical performance
Clinical relevance
This supports incorporating simulation into IR training, particularly for radiation and contrast stewardship, while recognizing that patient-level benefits are not yet established.
Who should care
Limitations
Small studies, substantial heterogeneity precluding meta-analysis, and probable publication bias.
Simulation training likely makes IR trainees more efficient with radiation and contrast, though proof of patient benefit is limited.
This is an AI-generated summary of the published abstract, not a clinical recommendation. It may not reflect the full text or subsequent updates — verify against the original source before acting on it.
Abstract
Purpose
To synthesise and appraise the certainty of evidence on simulation- and virtual-reality-based training in interventional radiology.
Materials and methods
MEDLINE, EMBASE, Web of Science, and Scopus were searched to April 2026 following PRISMA 2020, under an a priori protocol registered with PROSPERO (CRD420261358144). Eligible studies evaluated simulation or virtual-reality training with objective outcomes. Risk of bias was assessed using design-specific tools (RoB 2, ROBINS-I, the NIH before-after tool) and certainty with GRADE. Heterogeneity precluded meta-analysis; findings were synthesised narratively (SWiM), without pooling.
Results
Eighteen studies (495 learners; seven randomised trials) were included; all reported improvement on at least one metric. Pre-post-reductions ranged from 50 to 72% (fluoroscopy time), 27 to 53% (procedure time), 28 to 76% (contrast volume), and 31 to 71% (radiation dose) (p < 0.001 to 0.05). In controlled trials, simulation improved technique (for example fewer catheters advanced without a guiding wire, 25% versus 82%, p = 0.01) and reduced contrast volume (28.0 versus 40.0 mL, p < 0.01), whereas procedure time was rarely significant. Transfer studies reported real-patient gains, including superior first-attempt cannulation (16 versus 12 of 16, p < 0.05) and a 30% radiation-dose reduction during real CT-guided procedures (p < 0.001). No randomised trial was at low risk of bias; GRADE certainty was low for radiation- and contrast-related outcomes and very low for procedure time and clinical transfer.
Conclusion
Simulation probably improves interventional-radiology technical performance, most clearly for radiation- and contrast-related outcomes, with low-certainty evidence of clinical transfer. Small samples, heterogeneity and probable publication bias limit the evidence; adequately powered multicentre trials reporting patient-level outcomes are needed.
Level of evidence
Level 2a (Oxford Centre for Evidence-Based Medicine).
DOI
10.1007/s00270-026-04592-3
PMID
42717110
Research metadata (titles, abstracts, authors, publication dates) is sourced from Europe PMC, which aggregates content from PubMed/MEDLINE, PubMed Central, and other providers under the U.S. National Library of Medicine (NLM/NCBI). NLM does not claim copyright on abstracts; the original journal publishers or authors may hold rights to the full articles. AI-generated summaries are derived only from the abstract, never the full text, and are always labeled as AI-generated.
Want more like this?
Browse more from the feed. A free account unlocks the full feed, saving and communities.
Back to the research feed