The biggest data science trends in healthcare

The healthcare sector has been leveraging data science to accelerate operations and aid patient recovery from various ailments, including COVID-19. As the industry continues to navigate through the post-pandemic landscape, this is showing no signs of slowing down.

“There are so many areas where data science can help enable more effective care, from capacity demand management, to predicting the length of stay, to alignment on discharge and lower care demands for patients exiting acute care,” said Rob O’Neill, head of analytics at University Hospitals of Morecambe Bay NHS Foundation Trust (UHMBT).

“Since the pandemic, there’s been an accelerated use of data. COVID-19 has increased the need for health leaders to be able to make decisions in real-time and predict what resources they will require for upcoming demand. For example, being able to understand our current patient population’s risk for readmittance is crucial to effectively executing unscheduled demand forecasting and potentially managing an influx of crisis-related patients, and limit numbers of patients having to come back into hospital environments during a pandemic.”

In this article, we take a look at the biggest data science trends that are occurring in the healthcare space.

Evolving analytics platforms

NHS analytics leader O’Neill went on to explain how analytics platforms have been evolving to help meet demand for services.

He said: “At UHMBT, we’ve been combining data science and predictive analytics with Qlik, Snowflake and DataRobot. The companies have come together to create a modern analytics platform that enables the real-time analytics needed to predict patient readmittance, which, in turn, has helped plan for COVID-19 surge capacity. It’s a combination of a cloud data platform, enterprise AI and machine learning for predictive modeling and actionable business intelligence.

“With Qlik reading the live data from Snowflake, and by executing the predictive model from DataRobot right within the Qlik application, healthcare decision-makers have an up-to-date window into their current state of care, while also getting a data-driven sense of the strategies they need to execute to best serve their entire patient population.

“This is enabling clinicians at UHMBT to drill down specific modalities and departments to get a clear understanding of where, how, and why readmissions are occurring. This is helping us more confidently understand current demand and predict what resources we will need to deliver complete care to our entire patient population.”

Hot topics and emerging trends in data science

We gauged the perspectives of experts in data science, asking them about the biggest emerging trends in data science. Read here

Increased use of AI

AI and machine learning deployments have proved key in accelerating communication and data management within healthcare. According to Nick Mandella, managing consultant at Harnham, this looks set to increase further, to make care for patients even more efficient.

“There is huge potential for AI and data science to have a sizeable, positive impact in healthcare in the near future,” said Mandella.

“One example is using machine learning to optimise patient procedures, ensuring they’re in and out as efficiently as possible consequently freeing up bed space and allow for more operations to be undertaken – a problematic issue for the NHS currently.

“AI is also being used to help diagnosing illnesses. Using computer vision and deep learning to understand images from scans, those illnesses that can be harder to detect, such as certain types of cancers, can be found and treated a lot earlier, meaning a much higher rate of survival for patients.

“In many cases and tests that have been done, AI has been more successful than using a doctor with several years’ experience.”

DevOps adoption for cost reduction

In addition, Mandella believes that DevOps has played a vital role, particularly in the area of pharmaceuticals.

He said that DevOps has helped businesses in the space to reduce costs, achieve compliance quicker, and maintain productivity: “The healthcare industry is heavily regulated to ensure that the drugs created do not cause harm, and this includes monitoring its software and hardware components as much as anything else.

“Using Computer System Validation (CSV) is the most common way of companies being regulated by the FDA, but there’s no denying that this system is time-consuming and expensive. Using DevOps for this process allows businesses to autonomously reduce the risk of bugs, avoid bottlenecking all without damaging productivity and reliability.

“Not only do all these elements within DevOps mean the regulation process becomes far more streamlined, but regulations are more likely to be adhered to, and products are able to be taken to market much faster, improving ROI and revenue.”

How to drive impact and change via DevOps

Stephen Magennis, managing director for Expleo Technology (UK technology), discusses how impact and change can be driven via DevOps. Read here

Digital twins

Digital twins is another aspect of data science that’s been making moves within healthcare, helping to drive the sector’s post-pandemic recovery. The technology is allowing organisations such as the NHS to make decisions quicker, with the aid of modelling and simulation.

“This technology, which is already in practice across everything from cancer patient pathways to GP surgery waiting times, is one of the most effective ways to quickly play forward different scenarios, and produce the evidence needed to streamline processes and improve patient outcomes,” said Frances Sneddon, CTO at Simul8.

“NHS Trusts were using this approach for ICU capacity planning over the pandemic, feeding real time data from local and national government figures, internal resource planning systems and straight from the hospital floor into a digital twin so that they would know exactly how many beds, ventilators and staff would be required in advance of the peak. It’s been so effective that now these hospitals are building new models to look deeper into patient pathways and find ways to combat the much-talked about challenge of expanding NHS waiting lists.

“Breaking down everyday processes into numerical information – things like waiting times, staff numbers, resource availability, floorplans, typical timings for surges in demand – is where data science is contributing towards healthcare process excellence. This data provides the blocks on which to build out a digital replica – or twin – and it’s in this virtual playground that you can experiment and optimise by running simulations.”

Shifting towards preventive treatment

The healthcare sector has demonstrated an aim to shift from reactive to preventive measures for patients, and is leveraging data science capabilities in order to achieve this.

“Healthcare’s wealth of historic patient data coupled with the steady rise of home monitoring equipment means health services have vast stores of structured data,” explained Andre Van Gils, senior general manager, global sales & marketing at Omron Healthcare.

“Yet, to date, the overwhelming majority of treatment is reactive – managing an established condition or responding to medical emergencies.

“The mission to move that needle from reactive to preventive is why those growing stores of data are so important. With the right data science tools and patient home equipment, we’re learning as an industry how to spot trends and early indicators of conditions, allowing for a more preventive approach to care.

“The NHS is currently trialling a range of schemes using AI and automation in this respect. And we ourselves at OMRON have just kicked off research project with Kyoto University, exploring how AI could predict cardiovascular diseases at early stages. It is our hope that the results will directly transfer to pre-existing care platforms within the NHS, platforms we have already built and rolled out across the country.”

See also: Blockchain technology applications in healthcare