articles

Go Back

Driving Operating Theatre Productivity Through Machine Learning and Digital Innovation

Lee Scothern, Managing Director, Four Eyes Insight

The Elective Care Challenge Facing the NHS

The pressure on elective care within the NHS has become one of the defining challenges of modern healthcare delivery. While the system has demonstrated extraordinary resilience, treating record numbers of patients in recent years, the backlog for planned care remains stubbornly high. As of 2025, the waiting list in England alone sits at over 7 million treatment pathways, representing more than 6 million individual patients. For many, the promise of treatment within 18 weeks, once a cornerstone of NHS performance, has become increasingly out of reach, with barely 60% of patients currently seen within that standard.

These figures are often framed as a capacity problem. But from where we stand at Four Eyes Insight, the issue is more nuanced. Capacity certainly matters, but so too does how effectively that capacity is deployed.

Why Operating Theatres Matter So Much

Nowhere is this more evident than in the utilisation of operating theatres.

Operating theatres are among the most resource-intensive environments in healthcare. They require highly skilled multidisciplinary teams, complex coordination, and significant financial investment. Yet despite this, variation in productivity remains widespread.

It is not uncommon to see theatre lists start late due to upstream delays or finish early because of overly cautious scheduling. Cases overrun because of poor estimation, while in other sessions, valuable time goes unused. Cancellations, often driven by bed shortages, staffing gaps, or incomplete pre-operative pathways, continue to disrupt planned activity.

A System That Reacts Rather Than Anticipates

What emerges is a pattern not of failure, but of fragmentation.

Decisions are frequently made in silos, reliant on static data or retrospective reporting. The system reacts, rather than anticipates. And in that gap between planning and reality, productivity is lost.

This is precisely where machine learning and digital solutions have the potential to reshape the landscape.

Moving Beyond Averages With Machine Learning

Machine learning offers a fundamentally different approach to operational decision-making. Rather than relying on averages or assumptions, it enables predictions grounded in real-world complexity.

For example, the duration of a surgical procedure is not simply a fixed number. It varies based on the patient’s condition, the surgical team, the type of intervention, and even time-of-day effects. By analysing historical data at scale, machine learning models can generate far more accurate predictions of case length.

When applied to theatre scheduling, this allows lists to be constructed with greater precision, reducing both overruns and idle time.

From Prediction to True Orchestration

But the real value lies not just in prediction but in orchestration.

Operating theatre productivity is dependent on a web of interrelated factors: pre-operative readiness, bed availability, staffing levels, equipment, and downstream recovery capacity. A delay in any one of these elements can cascade across the system.

Traditional planning approaches struggle to account for this level of complexity. Digital solutions, underpinned by machine learning, can.

By integrating data across pathways, it becomes possible to move towards dynamic scheduling, where theatre lists are continuously optimised based on real-time conditions. If a patient is not fit for surgery, another appropriately prepared patient can be prioritised. If bottlenecks are predicted, such as critical care bed shortages, activity can be adjusted proactively rather than cancelled at the last minute.

This shift from static planning to adaptive systems is critical in an environment where demand consistently exceeds supply.

Forecasting Demand to Align Capacity

Equally important is the ability to forecast demand with greater accuracy.

Elective care demand is not random; it follows patterns shaped by referral behaviour, seasonal variation, and population health trends. Machine learning models can identify these patterns and project future demand, enabling organisations to align capacity more effectively.

This has profound implications not only for theatre scheduling, but also for workforce planning, estate utilisation, and investment decisions.

Technology Alone Is Not Enough

However, technology alone is not the answer.

The most successful transformations we see are those where digital tools are embedded within a broader operational redesign. Surgical hubs, for example, have demonstrated how dedicated elective pathways, supported by data-driven planning, can significantly increase throughput while reducing cancellations.

In some cases, productivity gains of over 20% have been achieved – not by adding more theatres, but by using existing ones more effectively.

The Cultural Shift Behind Data‑Driven Productivity

What is often underestimated is the cultural shift required to support this change.

Moving towards data-driven decision-making requires trust: trust in the data, in the models, and in the systems that surface insights. It requires clinicians and operational leaders to work in closer partnership, using a shared view of performance and opportunity. And it requires organisations to move beyond retrospective reporting towards real-time visibility.

Small Improvements, System-Wide Impact

At Four Eyes Insight, we have seen firsthand how even modest improvements in theatre utilisation can translate into meaningful impact at scale.

A small increase in on-time starts, a reduction in cancellations, or a more accurate alignment of case duration can collectively unlock thousands of additional procedures each year. In the context of a 7‑million‑strong waiting list, these gains are not marginal – they are transformative.

Putting Patients Back at the Centre of Productivity

Yet perhaps the most important perspective is not operational but human.

Behind every delayed procedure is a patient whose life is on hold. Someone living with pain, uncertainty, or deteriorating health while they wait for treatment. The conversation about productivity must therefore move beyond efficiency metrics and towards outcomes.

Every additional case completed, every avoided cancellation, and every reduced delay represents a tangible improvement in someone’s quality of life.

Working Smarter, Not Harder

The NHS has already demonstrated that it can deliver extraordinary levels of activity under pressure. The challenge now is to ensure that this activity is delivered as effectively as possible.

Machine learning and digital innovation provide the tools to do this, but their impact will depend on how they are applied, integrated, and embraced.

The future of operating theatre productivity will not be defined by working harder but by working smarter. By combining clinical expertise with advanced analytics, and by designing systems that are responsive rather than reactive, we have an opportunity to unlock capacity that already exists within the system.

At Four Eyes Insight, we believe that this is not just an opportunity – it is an imperative.

Contact us to learn how predictive analytics and intelligent scheduling can help reduce cancellations, improve utilisation, and treat more patients – without adding theatres.