Using FMEA to strengthen radiation therapy quality assurance

In contemporary radiation oncology, quality assurance can no longer rely solely on end-of-process checks. Patient pathways now span CT simulation, multi-modality image registration, inverse-planned dose calculations, image-guided delivery and adaptive replanning, all of which introduce points where small errors can propagate into clinically meaningful harm. Failure mode and effects analysis offers a disciplined way to anticipate those points before they affect a patient, making it a natural fit for the kinds of complex radiotherapy programmes now operating across Australia and New Zealand.

For delegates heading to the COMP Annual Scientific Meeting, the topic also reflects a wider shift in how bodies such as the Australasian College of Physical Scientists and Engineers in Medicine expect departments to demonstrate proactive risk management. The pages linked from the main conference website describe how the programme was designed to bridge clinical physics, dosimetry and patient safety, and this article unpacks how FMEA can be applied to a busy linear accelerator workflow.

Why systematic risk assessment matters in modern radiotherapy

Reactive incident reporting has been the backbone of radiotherapy safety for decades, and it remains essential. However, near-miss data alone tends to capture only the failures that have already happened, which means departments can overlook latent weaknesses in newer techniques such as stereotactic body radiotherapy, surface-guided delivery or MR-guided adaptive workflows. A prospective method like failure mode and effects analysis forces the team to think about what could go wrong at every step, rather than waiting for the inevitable close call.

The case for FMEA has been reinforced by regulators and learned societies. The Australian Radiation Protection and Nuclear Safety Agency, together with state regulators, has long emphasised a risk-based approach to authorisation of new treatment techniques, while the Australian Clinical Dosimetry Service has published audit findings that point to vulnerabilities in commissioning and in vivo dosimetry. Local audits conducted in centres from Melbourne to Perth and from Brisbane to Hobart repeatedly show that small documentation gaps, rather than dramatic equipment faults, are the most common precursors to incidents, and FMEA is well suited to uncovering them.

Building the right multidisciplinary FMEA team

A useful FMEA cannot be performed by a single physicist at a desk. The classic structure of a failure mode review brings together the people who actually perform each step: radiation therapists, dosimetrists, medical physicists, radiation oncologists, engineering staff and, increasingly, patient safety officers. In a department such as the Peter MacCallum Cancer Centre in Melbourne, this kind of cross-disciplinary workshop has become a familiar annual exercise, often coordinated by a senior physicist who chairs the meeting but deliberately steps back from dominating the discussion.

A good team will include someone with facilitation skills who can keep the conversation focused, someone who knows the information systems intimately (especially when record-and-verify platforms such as Mosaiq or Aria are involved), and someone who can challenge assumptions from a clinical perspective. In smaller regional centres, where a single physicist may be responsible for multiple linacs, FMEA sometimes doubles as a useful mentoring exercise, pairing senior and junior staff so that tacit knowledge about machine quirks is captured in the risk register rather than left in someone's head.

Mapping the patient pathway from simulation to follow-up

The first analytical task is to construct a detailed process map. Most Australian departments break the pathway into functional blocks: referral and consent, CT simulation, contouring and planning, plan checking, treatment delivery, and on-treatment and post-treatment review. Each block is then expanded into a numbered list of substeps, with arrows showing dependencies and handoffs. The Trans-Tasman Radiation Oncology Group has published workflow templates that many centres adapt for their own FMEA documentation.

Once the map exists, the team systematically lists everything that could fail at each step. A typical list for an external beam workflow might look like this:

• Patient identification error at CT simulation, with potential to propagate through every downstream step.
• Incorrect immobilisation device selected, leading to systematic shifts during delivery.
• Image fusion mismatch between diagnostic MRI and planning CT for head and neck cases.
• Misinterpretation of a structure-of-interest contour by the planning dosimetrist.
• Plan transferred to the wrong treatment unit in a multi-linac department.

The map and its failure modes become a shared reference document. Importantly, the same map can be reused whenever a new technique is introduced, so the team's investment pays dividends well beyond the first analysis.

Scoring risk: severity, occurrence and detectability

Each failure mode is then scored on three dimensions: how severe the consequence would be if it reached the patient, how likely it is to occur given current controls, and how likely it is to be detected before reaching the patient. The classic Risk Priority Number multiplies the three scores, although many modern teams prefer to display them as a colour-coded matrix so that the most worrying combinations stand out visually.

In practice, the conversation around scoring is often more valuable than the numbers themselves. A high-severity, low-occurrence item such as wrong-patient treatment still ranks highly because the consequence is unacceptable, prompting the team to invest in fail-safe patient identification protocols. A high-occurrence, lower-severity item such as late completion of on-treatment image review may justify a workflow redesign that automatically notifies the radiation oncologist when a review is overdue. Australian departments contributing to eviQ protocols often align their local scoring with the language used in those national guidance documents, making it easier to benchmark against peers and to satisfy accreditation reviews.

Turning analysis into action and ongoing oversight

An FMEA that ends with a long spreadsheet has not finished its job. The final phase is to translate prioritised risks into concrete improvements, assign owners, and set review dates. Practical actions that frequently emerge include the following:

• Adding forcing functions or hard-stops in the record-and-verify system to prevent unapproved plan transfers.
• Introducing or refining time-outs at the treatment console, mirroring the surgical safety checklist culture.
• Updating competency packages so that every new staff member is trained on the specific failure modes identified.
• Scheduling quarterly audits of the highest-ranked items to confirm that control measures remain effective.
• Linking the FMEA register to the departmental incident learning system so that emerging trends can be cross-referenced.

Sustainability depends on rhythm. Some Australian centres link their FMEA review cycle to their annual quality programme calendar, presenting outcomes to the local radiotherapy executive and feeding key findings into national forums such as the ACPSEM scientific meetings or the annual TROG symposium. Others integrate the FMEA register into their quality management software so that the document lives alongside incident reports, audits and competency records rather than gathering dust in a shared drive. Departments looking to coordinate with peer institutions or arrange collaborative audits can always reach the event team for introductions to relevant working groups.

For those wanting to explore how the COMP community is approaching these topics in person, the Exploring Lake Minnewanka story offers a reminder that professional meetings also depend on collegial downtime, with walking trails, lake cruises and informal dinners forming part of the week. Anyone interested in contributing to next year's FMEA-focused sessions, or who simply wants to share what their home department has learned, is warmly encouraged to submit an abstract, propose a workshop topic, or sign up as a reviewer through the organising committee before the call closes.