Improving resuscitation algorithms with simulation – Ben Symon | PAC 2024

A case study in redesigning a resuscitation algorithm through simulation, simplification, and systems thinking.

Speaker biography:

Dr Ben Symon is a Paediatric Emergency Physician and Simulation Consultant for Queensland Health and Mater Health.

As part of the Queensland STORK service, he works with an interdisciplinary team to improve paediatric resuscitation utilising education and translational simulation.

In his downtime, Ben co-produces ‘Simulcast’ a podcast that translates simulation literature for front line educators.

Insights:
(Powered by OpenAI)

Resuscitation algorithms are intended to support teams in moments of extreme cognitive load, yet many instead add complexity, ambiguity, and stress. Ben Symon argues that this is not a failure of clinical knowledge, but of design. Drawing on concepts of distributed cognition, he reframes algorithms as cognitive tools that should offload thinking from clinicians — much like drug calculators or reference charts — rather than demand interpretation while teams simultaneously manage patients, tasks, and emotional dynamics. In rare, high-stakes events such as paediatric major haemorrhage, teams cannot rely on experience alone, making the usability of systems critical.

Using translational simulation, Symon presents a detailed case study of redesigning a paediatric major haemorrhage protocol across a large health service. Diagnostic simulation revealed consistent failure points: incorrect blood sample labelling, poor role-task alignment, confusion around warming processes, and breakdowns in communication with Blood Bank. Importantly, these issues persisted despite extensive education and simulation — exposing that teams were rehearsing protocol activation, not execution. Visual overload, dense text, branching pathways, and unclear visual flow meant algorithms became unreadable once teams were overwhelmed.

Through iterative simulation testing, the protocol was radically simplified: branching was removed, visual hierarchy clarified, negative space deliberately preserved, and essential actions prioritised over completeness. Supporting information was separated from the core algorithm to reduce distraction. The result was an algorithm that teams found calmer, clearer, and easier to use under pressure. Symon concludes that improving resuscitation performance requires shifting focus from adding information to removing friction — designing systems for real human limitations, and using simulation not just to train clinicians, but to redesign the systems they rely on.

Key insights

• Resuscitation algorithms are meant to reduce cognitive load — but often do the opposite. Many current algorithms add complexity, require interpretation under stress, and compete with patient care for limited cognitive bandwidth.
• Algorithms should function as tools for distributed cognition. Like drug calculators or reference books, well-designed algorithms should offload thinking from the team, not demand it during peak stress.
• Rare, high-stakes events expose design flaws. Major haemorrhage is infrequent, meaning teams cannot rely on experience alone; systems must be designed for overwhelmed, novice performance—not ideal conditions.
• Simulation revealed consistent failure points. Errors in sample labelling, role allocation, warming processes, and communication with Blood Bank were common—even in experienced teams.
• Design matters as much as content. Font size, visual flow, use of colour, negative space, and removal of branching dramatically affected usability under pressure.
• Educational bias reinforced the problem. Teaching focused heavily on medical theory rather than execution, underestimating the operational complexity (often led by nursing staff).
• Translational simulation enabled real improvement. Using simulation diagnostically, iteratively, and for implementation helped redesign an algorithm that teams found clearer, calmer, and more usable.
  
  

Practical recommendations for acute care paediatricians

• Critically review your resuscitation algorithms. Ask: Does this reduce cognitive effort in a real resus—or add to it?
• Design for worst-case cognitive overload. Assume teams are stressed, inexperienced with the event, and juggling multiple tasks simultaneously.
• Simplify relentlessly. Remove branching where possible, minimise text, and separate “need-to-know-now” from “nice-to-know-later” content.
• Use simulation as a design tool, not just training. Observe how teams actually interact with algorithms during stress, and redesign accordingly.
• Test the whole process, not just activation. Rehearse execution, communication, logistics, and handoffs—not just the first 10 minutes.
• Engage multidisciplinary expertise early. Blood Bank, nursing, educators, and designers all bring critical perspectives that clinicians alone may miss.
• Accept that less information can be safer. Excess detail can obscure what matters most when teams are overloaded.
• Iterate continuously. Algorithm design should be a living process, informed by real-world use and repeated testing.