Introducing Niamh Wilson: Associate Director at Kensa Health

Tell us about your background and how you ended up working in healthcare analytics.

I’m a registered nurse and spent 10 years in critical care, where I began my interest in technology. From there, my move to a project nurse role at Imperial in 2001 involved implementing electronic prescribing and stock management.

Through this experience, I learned that technology can improve things, but only if you consider the whole patient pathway. That broader perspective is crucial when making changes in complex health systems.

Excited by the potential of technology and data, I returned to university to study computing and systems. Since then, I’ve spent 20 years in digital health, focusing on electronic patient records and, more recently, how operational data and statistics support evidence-based decision-making.

What kinds of healthcare or operational challenges have you worked on most closely?

I’ve worked with clients on challenges like ED performance, length of stay, and bed occupancy. My approach is always to bring data into the conversation practically, segmenting operational data to help clinicians understand what’s truly going wrong. Sometimes the data confirms assumptions; other times, it challenges them. Either way, it helps identify interventions with the greatest impact.

What makes healthcare analytics different from standard reporting or dashboards?

Healthcare is still catching up with other industries in using statistical techniques to predict performance. While reporting and dashboards show what’s happened, analytics should help us understand variation, spot problems early, and support better decisions. Other sectors use well-tested methods to reduce variation and make performance predictable. Though we acknowledge that patients aren’t parcels, there’s much to learn from these industries about system behaviours and how to respond to demand.

For me, that’s where healthcare analytics really adds value: it helps us identify system behaviours that aren’t responding well to patient needs and may be contributing to poorer outcomes.

Can you talk about a piece of work or insight that genuinely changed decision-making or outcomes?

A project on winter pressures challenged common assumptions. Many believed busyness was due to increased demand or more complex patients. Our analysis showed admission rates for complex patients were stable; instead, bed occupancy

increased because the length of stay rose after Christmas. Discharges slowed, partly due to annual leave and interface issues between acute and social care. The health system used this insight to plan a focused discharge effort over the holidays, reducing bed occupancy.

For me, this was a great example of how analytics revealed the true operational cause of pressure, moving past assumptions.

What excites you most about where NHS analytics is heading over the next few years?

I think AI is one of the most exciting developments, but it won’t automatically make things better. We need to monitor its impact and ensure it genuinely improves care. What excites me most is the potential to simplify the complexity of healthcare data and support better decision-making across every level of the healthcare system.

What do you think operational leaders most need from analytics today?

I believe operational leaders need faster feedback from analytics. Formal evaluations can take months, but applying a statistical lens to operational data in real time provides timely, reliable insights.

A fast feedback mechanism helps leaders understand whether changes are making things better or worse, giving confidence to scale what works and stop what doesn’t quickly.

What’s one misconception people have about NHS analytics?

Many think imperfect data is useless for analysis. In reality, by looking at trends over time and combining them with frontline insights, we can still gain a valuable understanding. Time series charts and statistical process control clarify trends, and consistent data quality issues rarely prevent useful interpretation.

What signals or patterns do you look for first when something starts going wrong operationally?

At Kensa Health, we use seasonal and cyclical trends alongside statistical process control charts. The first step is to determine whether there’s a true special-cause signal or if the system is still within normal variation.

For example, an ED may escalate, but data might show demand is within the normal range, just sustained at the high end for several hours, leading to queues. Once congestion starts, delays build, and staff struggle to recover. Data helps teams have grounded conversations about early warning signs and appropriate responses.

What makes system-wide analytics difficult in practice?

Healthcare data is often fragmented across organisations and teams, making it hard to get a complete picture. Different organisations may have varying priorities and definitions, which complicates shared decision-making. Good system-wide analytics brings together different parts of the system, helping everyone look beyond their own organisation to understand how the whole pathway works for patients.

What’s the most valuable thing operational teams can gain from real-time analytics?

Real-time analytics lets operational teams be proactive rather than reactive. Combining operational data with predictive analytics shows what’s likely to happen next, not just what’s happened. For example, unscheduled care demand is often predictable within a normal range. If a system admits extra frail patients or loses capacity, analytics helps anticipate the operational impact and risk of escalation earlier. This proactive approach lets teams act sooner, manage risk, and address pressures before they escalate.

Can you describe a moment where the data told a different story from assumptions?

Often, teams are excited about new models of care, only to find the data shows little improvement. That’s when analytics is most helpful, it reveals what’s really happening and where models need refinement. For example, a frailty model may assume earlier assessment leads to earlier discharge, but data might show patients follow the same pathway or arrive out of hours when the model isn’t available.

These insights help clinical teams move past assumptions and refine models for better outcomes.

To discuss how these insights apply to your system, connect with Niamh on LinkedIn.

If you’re ready to manage demand, capacity, and variation with greater confidence, contact the Kensa Health team today.

‍