Covering about 1 million patients in the Lothian region of Scotland, the DataLoch is one of the largest single repositories of health data in the country.
Funded by the University of Edinburgh’s Data-Driven Innovation Initiative, DataLoch contains routinely collected de-identified data.
The service enables approved researchers to access securely extracts of routine care records from more than 100 general practices and three acute hospital sites in Lothian, as well as national hospitalisation, prescribing and death records.
The ability to visualise and analyse such connection points can allow researchers to discover new insights into population health – aided by a data-driven approach, including artificial intelligence and machine learning. Before researcher access is given to a data extract, identifying information is removed to help protect patient confidentiality.
One researcher who has been working with this resource is Peter Gallacher, a health informatics researcher by day and trainee GP by night.
A graduate of medicine, Gallacher is one of a legion of new doctors who are combining their love of frontline clinical work with data science.
In his particular case, he has used patient data to study kidney disease and kidney failure, and has a PhD in cardio-renal epidemiology.
It is still one of the less well understood fields of medical inquiry, but Gallacher’s research interests and clinical studies have led him to discover potentially ground-breaking interventions for kidney patients’ treatment pathways.
Gallacher explains how collaborating with the DataLoch service has been invaluable in linking datasets across pharmacy, GP, and hospital settings, thanks to it combining primary and secondary care records and combining data across an entire population of patients, irrespective of their age, sex, ethnicity or socioeconomic status.
“One of the studies that I’ve completed recently, on a national level, was looking at patients with kidney failure who need regular dialysis or who have a kidney transplant.
“This study used hospitalisation records, Scottish Renal Registry records, prescribing records, death records, and linked them. All these data were collected as part of routine clinical care,” says Gallacher.
Being able to access and analyse these data has led Gallacher to be able to devise “novel approaches” to the treatment of people with kidney disease, and to understand better cardiovascular risks within the population of patients with kidney disease.
He says: “In this nationwide study, we analysed these linked data and evaluated trends in heart attacks and strokes in patients with kidney failure in Scotland over the past 20 years, before comparing their rates to the general population. We observed some encouraging trends – rates of heart attack and stroke have halved in patients with kidney failure during this period.”
He adds: “Despite this, patients with kidney failure remained up to eight times more likely to have a heart attack and up to four times more likely to have a stroke than the general population, illustrating the hugely increased cardiovascular risk in this patient group.
“Using the available prescribing data, we were also able to show that anti-platelet drugs – a treatment commonly prescribed to the general population after a heart attack or stroke – were being prescribed more often for patients with kidney failure in recent years.
“Overall, our results highlight the urgent need for a clinical trial of these agents to clarify their safety and effectiveness in patients with kidney failure.”
Although the discovery has potentially system-wide impacts on pathways for kidney disease patients, and how they are managed in various clinical care settings, Gallacher comments that there is still much work to be done on understanding the disease more broadly.
He says while it was known that kidney disease patients were at higher risk of heart attacks or strokes, clinicians had not previously known to what extent. The work he and colleagues have completed has allowed them to “fill in that knowledge gap”.
In terms of next steps, there is a need to conduct clinical trials in patients with kidney disease to see what more can be done to reduce the cardiovascular risks in these high-risk patients.
A report published last year by Kidney Research UK indicated that kidney disease costs the NHS across the UK approximately £6.4 billion each year, amounting to around £566 million in Scotland.
Furthermore, it revealed that there are expected to be 2,123 people in Scotland on kidney dialysis, 212 people receiving kidney transplants, and 52,000 acute kidney injury episodes per year.
Over the course of the next decade these figures have the potential to increase significantly, with potentially as many as 10,000 people receiving dialysis and 863 transplants per year.
Yet at the same time, despite being the tenth biggest cause of death worldwide, it only received 1.4 per cent of relevant public healthcare research funds in the last financial year.
Peter Gallacher, above, is based at the Centre for Cardiovascular Science at the Queen’s Medical Research Institute.
Gallacher says kidney disease patients have also traditionally suffered from what he describes as “therapeutic disenfranchisement”, whereby it is known that they are at higher risk of adverse cardiovascular events, but little has been in place to prevent them, largely because they have been excluded from large-scale clinical trials.
He said: “They are at higher risk of these events, but there’s no evidence to suggest that they would benefit from treatments because they’ve not been included in the trials in the first place. That’s an oversimplification because sometimes they may be at a greater risk of side effects from treatments, but you need to start somewhere.”
Gallacher’s studies also looked into the cardiovascular risk of kidney disease patients following a positive Covid test.
He found that the risks were particularly high in the first month after testing positive, and then reduced fairly rapidly to the expected risk levels.
That insight, again, could frame new ways of treating kidney patients affected by Covid and potentially other coronaviruses.
“Again, the key tenet here is the novel use of routinely collected data, which Scotland has been doing for over 20 years and has been world-leading,” says Gallacher, who shared some of his research at Futurescot’s Health & Care Transformation conference in April in Glasgow.
He ascribes much of Scottish healthcare’s data-crunching prowess down to the fact that every patient in the country has a Community Health Index (CHI) number.
“Every patient that’s registered with a GP, since the 1970s, has got a CHI number,” he says. “This means we can link people’s records, de-identify the data, and perform some quite exciting and innovative analyses. In doing so, we can answer some important questions for traditionally disenfranchised patient groups. That’s what I am really passionate about.”
Data-driven innovation as the next frontier in healthcare
Preventative medicine will be a “cornerstone” of our future approach to medicine, says Gallacher. “Data science is probably still in its infancy but it’s showing a lot of promise,” he says.
“There is huge potential for the use of AI algorithms to enhance our ability to predict risk and to instigate preventative therapies. Data science can allow for the implementation of clinical trials as well.
“And going forward, that is going to become increasingly the norm for data-driven research as part of the basis for clinical trials.”
Gallacher, who learned to code during his Master of Public Health degree, sees the profession itself being disrupted by data science and AI.
“Professor Sir John Tooke wrote 10 years ago about the need for the clinician scientist to be educated and trained in data science and analytics. And that is now becoming such a key part of our career.
“Healthcare and data science go hand in hand in my opinion because the potential of these datasets is limitless.”
Partner Content in association with the University of Edinburgh
