The data governance frameworks that will boost interoperability in healthcare

When Heather Fraser, Global Life Sciences and Healthcare Lead at healthcare solutions provider IBM, had to rush her husband to the emergency department of a remote Scottish hospital, she had to provide his medical history manually due to the national borders between England and Scotland. That got her thinking about the need for interoperability in healthcare.
 

“When applications, devices, and systems are able to interact and exchange information in real-time, healthcare organisations can operate much more efficiently, and that’s going to help patients receive that seamless and personalised care wherever they go,” said Fraser at a recent event hosted by GovInsider, Connectivity at the core of healthcare transformation. 
 

Fraser said that this seamless integration can not only improve patient experience, but also patient safety, by helping to ensure the accuracy of health data. This will especially impact some health inequities that are prevalent in underserved populations, which may see more fragmentation when it comes to communicating patient data.
 

A more integrated global healthcare ecosystem could really help the world tackle the next public health crisis much more quickly than the world did previously. It can also help to reduce the global cost of healthcare when resources can be shared easily, said Fraser.
 

GovInsider hears from the IBM team on how healthcare organisations can begin to break down their siloed systems to achieve better overall patient outcomes.
 

The three principles of healthcare data governance

It may not be that healthcare organisations lack the will to work in tandem with each other, but the means to execute this will. A 2021 Future Health Index published by health tech provider Philips found that the difficulties with data management and the lack of interoperability and data standards present the biggest barriers to adopting digital health technologies in healthcare organisations.
 

Rather than isolated healthcare systems, Fraser suggested that healthcare organisations can consider three key principles to govern how patient data can be accessed, exchanged, and stored.
 

The first is having open data transmission standards. This means ensuring that everyone speaks a common language and takes a common approach to sharing healthcare data. International partnerships such as the Global Digital Health Partnership, for example, help bring countries together to establish best practices for data sharing and the effective implementation of digital health.
 

Second, developing an open architecture and open source software. An open-source approach enables greater collaboration and allows innovators and researchers to avoid duplicated innovation efforts, said Fraser. For instance, the team behind Singapore’s newly launched Health Appointment System has made the programme’s code open source, so that other countries may adopt similar solutions easily.
 

Finally, to maintain cybersecurity standards, healthcare organisations should adopt a zero-trust model. Medical organisations have suffered nearly 5,000 data breaches affecting over 342 million records since 2009 in the US alone, reported Comparitech previously. This is not only unsafe, but costly – according to IBM’s annual Cost of a Data Breach report, healthcare has incurred the highest costs in data breaches across all industries for the last 12 years.
 

“Zero trust is a cybersecurity framework that goes beyond perimeter base controls, and assumes that breaches have already occurred. It always verifies processes through access controls, identity management, and contextual data,” explained Fraser.
 

Dealing with the scale of data

Once these pieces are in place, healthcare organisations then need to think about processing huge amounts of both structured and unstructured data, said Ed Macko, General Manager, Global Healthcare & Life Sciences Industries, IBM Technology Global Sales, at GovInsider’s event.
 

Traditionally, healthcare professionals would deal with datasets in the clinical domain, such as patients’ health histories, allergies, and medications. They might also deal with patients’ financial records, such as past transactions, and socioeconomic tiers. “But most people don’t realise that if you look over the life cycle of an individual, that only constitutes about 10 per cent of their data,” said Macko.
 

Moving forward, Macko expects that healthcare organisations will look at more holistic social determinants of health, such as family history, and even “omics” – genomic studies that aim to provide a fuller genetic makeup of any individual. Further, the increased uptake of IoT devices like smart watches and Fitbits will have a part to play in supplying real-time data in remote patient monitoring programmes.
 

These are massive datasets – according to IBM Watson, the amount of medical data has doubled every 73 days since 2020. And healthcare organisations historically spend 80 per cent of their time integrating data from a variety of sources, and then interpreting this data, said Macko.
 

“Using things like natural language processing to understand even unstructured data such as clinical notes is critically important,” said Macko.
 

For instance, IBM helped the second largest integrated delivery network in the US in predicting the early onset of specific diseases, including the SARS-Cov-2 virus responsible for the Covid-19 pandemic. By building data science models, the IBM technical team managed to streamline the processing of massive amounts of data, cutting the prediction process down to six weeks.
 

“These predictive models also helped our clients analyse social determinants that help them estimate the overall cost of care within the ecosystem. When we leverage these platforms, physicians can spend more time on patients’ personal health and wellness,” said Macko.