Cardiovascular and Interventional Radiological Society of Europe
Slider
PublicationsCIRSE InsiderCIRSE updates standards of practice document to improve complications reporting

CIRSE updates standards of practice document to improve complications reporting

November 19, 2025

CIRSE has released a new standards of practice document modifying the CIRSE Complications Classification system. To learn more about this update, CIRSE Insider spoke with the authors: Prof. Christoph Binkert, Prof. Dimitrios Filippiadis, Prof. Klaus Hausegger, CIRSE President Philippe Pereira, and Prof. Anthony Ryan.


Christoph Binkert
Dimitrios Filippiadis
Klaus Hausegger
Philippe Pereira
Anthony Ryan

CIRSE Insider: The need for an update to the 2017 CIRSE system for classification of complications was identified during the International Conference on Complications in Interventional Radiology (ICCIR) in 2023. What was the context behind this action?

Filippiadis: Between 2017 and 2023, the previous classification system went through the necessary “field-testing” in multiple centers and countries. When it came time for the ICCIR, we felt the event provided a suitable international forum to survey and validate the system with many different users. This approach aligned with the natural cycle of review to ensure that the CIRSE classification system remains reliable, reproducible, and clinically meaningful.

Hausegger: It’s also worth noting that our main challenge in updating the system was to strike a balance between preserving the classification’s simplicity and incorporating greater detail where necessary.

CIRSE Insider: How does the current modified version improve on the previous system?

Filippiadis: The 2017 version of the CIRSE classification system was published with the goal to standardize the reporting of complications by avoiding subjective definitions and an excessive number of grades. Now, the current modified CIRSE classification improves the original system by clarifying gray zones, especially regarding incomplete procedures and prolonged hospitalizations.

Binkert: Dimitrios’ last point on the gray zones is key. We realized that short-term outcomes, like spending one or two nights in the hospital, don’t truly capture a patient’s experience. What matters more is the long-term impact. If a complication means you can’t return to your favorite activities or sports, that affects your quality of life far more than a brief hospital stay.

CIRSE Insider: One of the main updates in the modified system is the introduction of new subdivisions such as Grades 1a/1b and 3a/3b. What prompted this change?

Pereira: During the validation process at ICCIR 2023, it became clear the previous grades were too coarse and sometimes not precise enough. They grouped events with substantially different clinical impact, management, and resource use into the same grade. Indeed, even if inter-observer agreement and inter-rater reliability were excellent, only a mean agreement of 68% was found on evaluation of audience voting results during ICCIR. Inconsistencies were observed in audience voting particularly for Grades 1 and 3, which the prompted the change.

Binkert: To elaborate on what Philippe’s describing, we divided Grade 1 to distinguish between complications that were solved within the same session and complications that were solved within the same session, but treatment was not performed as a result. Grade 3a involves a hospital stay longer than two days, whereas Grade 3b involves a hospital stay longer than two weeks. Both have no sequelae, meaning the patient can return to their normal life.

CIRSE Insider: With these refinements, how do you expect the modified system to influence daily clinical practice and complication reporting?

Ryan: I expect a bigger impact on the reporting of complications, rather than a direct impact on clinical practice, although the principle of ‘what gets measured, improves’ may apply. The increased specificity will be relevant in terms of more accurately documenting patients’ outcomes and the impact on healthcare resource utilization. The latter is particularly salient; while image-guided IR is thought to be more cost-effective than more invasive approaches, the negative impact on cost-efficiency of a procedure must also be documented.

Hausegger: This is an important point. These classification systems, no matter which one, may not directly impact clinical practice, but they are essential tools for quality assurance and reporting.

CIRSE Insider: Implementing a revised framework across diverse healthcare settings can take time. What challenges do you foresee in achieving widespread adoption?

Pereira: Interventional radiologists, trainees, clinicians, and administrators may be unaware of the update. To overcome this, CIRSE is promoting targeted education and continuous professional development. I encourage other societies to also disseminate the document.

Hausegger: I anticipate a rapid uptake in the IR community, especially by authors publishing in CVIR, but one potential challenge going forward is that other medical specialties will likely continue using their own classification systems.

CIRSE Insider: Do you see a possibility for a global classification system in IR?

Pereira: A harmonized global system seems to me achievable but will require structured international collaboration, validation across varied settings, and tools to bridge existing classification systems.

Filippiadis: Agreed. A global classification system for reporting complications in IR would improve comparability, education, research quality, and patient safety worldwide. The main challenge here is coordination, not concept.

CIRSE Insider: How do you see the CIRSE classification system evolving in the future?

Filippiadis: I could see future developments focusing on digital integration within electronic health records and registries to enable automated and real-time complication logging.

Binkert: Actually, I believe the system should remain stable for a considerable length of time. If too many changes are made, it becomes difficult to classify a procedure now and then again with refinements in ten years. But you need the same system in the long-term, or else it becomes difficult to determine robustness.

Ryan: To put it metaphorically: the proof of the pudding is in the eating. As with any system, modifications will undoubtedly come to light due to its widespread usage and subsequent feedback from our colleagues; however, maintaining the clarity and usability of this updated system will be a priority.