The use of anonymised electronic data from routine clinical care has become an increasingly important tool for identifying the causes of disease and developing more effective treatments. Sara Khalid is among the data analysts whose work in clinical informatics is underpinning medical breakthroughs.
Sara is doing a post-doctorate in biomedical data science in the Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences (NDORMS), with funding from the Oxford BRC’s Musculoskeletal Theme.
Her expertise lies in signal processing and machine learning, and she works in a team, led by Prof Daniel Prieto-Alhambra, that is using routinely collected health service data to understand the epidemiology of musculoskeletal conditions.
“We use observational data: anonymised and de-identified data collected from everyday use in GP surgeries and hospitals. That allows us to conduct research on healthcare utilisation in the real world. What I do is I use my background in signal processing and machine learning to analyse the data, to derive insights that can ultimately inform policy-making,” Sara explains.
Among her recent projects is a study in which Sara analysed data from millions of patients in the UK and Spain to understand how anti-fracture medication prescribed by GPs is used. She found that as well as the expected categories of patients – older women at risk of osteoporosis, a smaller group of men with comorbidities – there were unexpected clusters of younger, healthy women who were prescribed these drugs, potentially raising questions about appropriate prescribing.
In another project, she examined datasets from 30 countries to develop a more accurate way of predicting and classifying the various sub-conditions of vasculitis, a rare condition involving the inflammation of blood vessels.
“These are examples of how I use, in my day-to-day work, machine learning models which can allow for such data-driven analysis. We don’t have an a priori hypothesis, but we allow the data to talk and derive insights from it,” Sara explains.
“The kind of research that’s going on in our department is very real-time in that we see clinical translation very quickly. I’d like to think it’s affecting our patients’ lives in the immediate term, rather than in theory.”
“Whatever I’m doing, there needs to be a connection between my skills and a real-time impact on people’s lives”Sarah Khalid
Sara joined NDORMS in 2016. She previously worked in the University’s Institute of Biomedical Engineering, under Professor Lionel Tarassenko, the Oxford BRC’s Theme Lead for Technology and Digital Health, and had done her PhD in using signal processing and machine learning to develop tools to detect early whether a patient was going to have a cardiac arrest. “It was very exciting work,” she says.
Sara was born in Pakistan and grew up in Karachi. “I came to study at Oxford as a Rhodes Scholar. I first did an MSc in the engineering science department in biomedical engineering. I then stayed on and did my DPhil with Lionel.”
The birth of her daughter coincided with the last year of her DPhil, so Sara decided, once she had submitted her thesis, to return to Pakistan for a year to spend time with her daughter. During that year, she was offered the role at NDORMS.
Sara had the choice, after her PhD, to leave academia and go into industry. “I chose to come back into academia and do research. A key reason for that was work/life balance, including childcare support.”
She adds: “Academic work is challenging. You do have to give it your all if you want to climb the ladder. As a woman working in academia and science, I’ve found it’s been good on the whole. Compared to the alternatives, I’d say life is good, the opportunities are there, and increasingly there’s a lot of recognition and understanding of the issues that can affect a woman in academia and having a family. There’s a lot of support nowadays.”
She says one of her biggest motivations in her current role is being able to help patients: “I wasn’t always sure about what I wanted to do, but the one thing that’s clear to me is that, whatever I’m doing, there needs to be a connection between my skills and a real-time impact on improving people’s lives, and healthcare research is a very direct and obvious way to doing that.
“It gives me a lot of satisfaction to see things I spend my time on during the day are actually directly or indirectly helping people survive or live better lives.”
Next profile: Chantal Hargreaves →