Lead analyst Lorna Barclay’s skills are helping NHS organisations plan for the future

Following a PhD in statistics and several years as a data scientist, three years ago Lorna Barclay entered the health sector. Here, she talks about her work at Dr Foster and explains why predictive analytics will benefit the NHS.

Lorna Barclay, lead analyst

More than the numbers

I have worked at Dr Foster for two years now and became a lead analyst in June. I think our offer is unique – Dr Foster has a deep understanding of healthcare, and specifically the NHS, combined with technical tools and capabilities. As an analyst, it’s easy to be focused on the numbers, but we work closely with customers and I’ve learned a lot about how the NHS works. When we build a model, we consider the data that is going into it and how the NHS will find it beneficial.

Using my skills to improve patient outcomes

In my work as a data scientist for a customer science company, I gained experience using machine learning techniques to enhance products. I looked, for example, at how the impact of in-store promotions and advertising in supermarkets could be measured and predicted, and I developed a dashboard around this. In another of my previous roles I used machine learning and predictive modelling to support pharmaceutical clients in understanding their target population. I apply the same techniques to the work I do at Dr Foster.

I decided to move into the health sector because I wanted to be involved in improving quality of care for patients. I think predictive analytics can support healthcare and have a real impact on patients’ lives. If my work means someone avoids having to be admitted to hospital, or is prevented from getting a disease, that’s rewarding.

Looking to the future

I’m one of three lead analysts – I lead on statistics, data science projects, and predictive analytics. Recently, there’s been a movement towards predictive analytics and machine learning in the NHS. I think this stems from the realisation that when it comes to improving patient care, it is helpful to know what is likely to happen in the future, particularly in terms of demand. It’s an area Dr Foster is getting much more involved in, and that’s where my skills and experience come in. We have developed a risk model predicting the risk of an emergency admission at a patient level using machine learning techniques. The same technique can also be applied to predict, for example, the risk of falls or high resource use.

Dr Foster has recently launched the Analysis Managed Service, which provides support from the whole analytics team that is tailored to customer needs. I will be working with organisations on building patient-level risk models, implementing our emergency admissions model, and predicting future demand and forecasting costs, so that they can plan better for the future.