Paul Harper

Professor Paul Harper

In summer 1996, I found myself amongst the seeming chaos of an HIV and AIDS clinic in downtown Mumbai. It was 7:00am. Patients were already queued on benches along corridor after corridor. It was an eye-opening introduction to the three months I spent in India working on my master’s dissertation.

What really struck me was the inadequate resources, given demand. Clinicians (at least those who would touch patients thought to have HIV) were unfamiliar with the virus, and the Indian Government denied it was a problem at all. My task was to apply Operational Research (OR) and mathematical modelling techniques to measure the scale of infection, and the resources needed to tackle it.

I look back fondly on that summer. No teacher or lecturer had ever told me my mathematics could be useful for healthcare, let alone that it might save lives. So my fascination for healthcare modelling began, which would eventually lead me back to University for PhD research.

NHS services

That passion persists, and I’m excited to be leading a talented team of collaborative researchers funded by various healthcare partners. The NHS currently has to deal with increasing demand and complexity within constrained budgets. Designing and delivering services that maximise resources is vital.

...there is a great deal of personal satisfaction in knowing our research is not only helping to improve the NHS, but is really saving lives.

Professor Paul Harper, Deputy Head of School, Professor of Operational Research

Healthcare systems typically operate in an uncertain environment. Imagine managing services in a large hospital on a day to day basis and, more strategically, planning ahead. Hundreds of patients with varying resource needs will pass along different care pathways every day. Help is needed to forecast demand, schedule clinics, reduce waiting times and so on. 

But a hospital is just one part of an interconnected system. What happens here is influenced by, and influences, policy decisions within primary care. In my opinion, whilst the NHS works hard to improve services, it does so in silos: changes in one part can have an unintended impact elsewhere. Planning plays a vital part in balancing the system. It can help overcome geography, so services reach patients within a critical time limit, and ensure the cost-effective delivery of policies such as screening for disease.

Modelling improvements

With OR, we build mathematical based models of current processes and use them to explore “what if?” scenarios of different ways of working, and move towards optimal services. This is much safer than experimenting with changes ‘for real’. It’s a technique that really does save lives. In one major hospital, our research completely redesigned care for stroke patients, resulting in a reduction in mortality rates by 60%. In another hospital A&E department, our work helped save the Health Board £1.6m per year through optimised capacity planning.

What is really exciting about our collaboration with the Aneurin Bevan University Health Board is a step change in capability following the creation of an embedded modelling unit. Having a team of four Operational Researchers in the hospital, working in partnership between the University and Board, is truly innovative and creates a virtuous circle. From a research perspective, the team’s methods create outputs and impact case studies, and so act as a catalyst for further grant applications and funding, including PhD studentships. In turn, research activities enhance our student experience, so knowledge improves.

We are unaware of a unit of this nature and scale in the UK, and potentially worldwide. Our work is putting the University, the Board and NHS Wales on the map as a hub for innovation in the design and delivery of healthcare services. Reflecting on my early experiences in India, there is a great deal of personal satisfaction in knowing our research is not only helping to improve the NHS, but is really saving lives.

See Professor Paul Harper discuss the 'Maths Saves Lives' project.