From drug discovery to diagnostics, population-level patient data has a powerful role to play in overcoming the biggest health challenges of our generation. Dr Petros Kotsidis, Chief Data Officer at FITFILE, explores how properly-managed health data could be used to accelerate progress in the life sciences sector, and how existing barriers to progress can be overcome.
According to research and analysis carried out by the WHO, the top 10 most common diseases and health conditions account for approximately 55% of the 55.4 million annual deaths. Including cardiovascular diseases, respiratory conditions and neonatal conditions, these illnesses pose a huge obstacle to governments’ efforts to ensure healthy lives and improve well-being for all ages (WHO’s 2030 Sustainable Development Goal 3).
One positive consequence of these conditions’ high incidence is that national health systems hold vast quantities of data on their causation, diagnosis, progression and transmission. This is data that could be key to reducing disease morbidity and mortality, with use cases including drug discovery, accelerated diagnostic pathways and epidemic prevention.
However, owing to fundamental flaws in the way that the majority of global health data is stored, accessed and leveraged, its wealth of potential remains largely unrealised. Before we are able to accelerate efforts to tackle the global disease burden, we must first unlock the power of patient data by giving healthcare professionals, scientists and population health experts complete, convenient and connected access.
Data: healthcare’s most precious resource?
To fully understand the urgency of the need to better engage with patient data, it’s necessary to appreciate the nature of the information captured, how this delivers actionable insights and the scope of these insights’ transformative potential for global health outcomes.
Trillions of new raw data points are generated every day: from national health systems, wearable devices, medical technologies and the life sciences sector. Every time a patient visits their doctor, has a blood test or starts a course of medication, they leave behind a data footprint. When millions of these footprints are integrated with other data points – including social, personal and activity – a holistic, multi-dimensional picture of population health can be captured and leveraged.
This detailed picture can be used to target health interventions and to map their impact, track and prevent disease spread and guide the allocation of funding, service provision and research resources. In addition, data from specific diseases can train AI models to diagnose patients earlier, accelerate drug R&D and identify previously unknown comorbidities, symptoms and risk factors.
Data in action
In 2022, the NHS launched the Galleri Trial, a pioneering project to improve the early detection of more than 50 types of cancer. Today, DNA data gathered from around 140,000 trial participants is currently being used to test the efficacy of a new diagnostic test – the Galleri Test – which could dramatically improve health providers’ ability to identify and treat cancer patients in the early stages. This is a project which depends heavily on accessible patient data – it was used to accelerate the recruitment of eligible patients and will be used to draw conclusions on the efficacy of the test itself. The study data will be stored and used in a highly impactful way to conduct additional studies to advance scientific research and public health – projects that may involve bringing together coded information from the NHS-Galleri Trial with information from other studies, electronic health records or biobanks.
It’s not just the treatment of major global diseases such as cancer where data has a pivotal role to play. A recent study assessing the impact of rare diseases (RD) on patients and healthcare systems concluded that ‘Machine Learning strategies applied to healthcare system databases and medical records using sentinel disease and patient characteristics may hold promise for faster and more accurate diagnosis for many RD patients and should be explored to help address the high unmet medical needs of RD patients’. This proves that although a greater data volume is likely to deliver more reliable insights, smaller sets of relevant data can be equally useful when brought together under the right conditions.
Crafting the keys to unlock data’s potential
While these are excellent use cases for connected patient data, we must also evaluate the obstacles to its effective utilisation. Foremost among these is the siloed, disconnected nature of the majority of existing social, demographic and health data. Information is ‘locked’ in internal computer systems, coded in a multitude of ways, disjointed and dispersed. In the NHS, for example, disjointed and dispersed patient data often makes it incredibly difficult for clinicians to ‘join the dots’ regarding a patient’s treatment journey across different NHS services. For research and planning, regulatory requirements such as the GDPR and UK DPA include having a valid legal basis for secondary data use, such as specific individual consent.
As a rule, accessing and uniting multiple types of source data is therefore complex, slow, expensive and inaccurate. This is hampered further by the mistakenly perceived inability to integrate anonymised data, which preserves privacy, negates the need for consent, and maximises data volume and utility for all stakeholders’ and patients’ ultimate benefit.
It follows that the first step towards reaping the benefits of connected data is to deploy technologies that enable a clear, unified and holistic view of patient information. We need to transition to systems designed for interoperability and complete health profiles, so we can put an end to the time-consuming, frustrating work of patching together poor quality information to gain poor-quality insights.
Security as first priority
Finally, it’s worth noting that the most stringent safeguards and access controls must always remain in place to protect patients when unlocking the latent value of their data. This can be best achieved by minimising the movement of data through the use of single point-of-access cloud storage platforms and by employing ongoing ‘fire drill’ security tests on regularly updated systems.
In addition, while pseudonymisation or tokenisation has its uses, wherever possible the anonymisation at the source of all types of patient health data ought to be non-negotiable as Article 89(1) of the GDPR requires, meaning that Personal Identifying Information (PII) cannot be linked to health data.
In summary, population-level patient data has a powerful role to play in overcoming the biggest health challenges of our generation. To realise this potential, we must deploy proven systems to unlock and unites health and activity data and to deliver safer, faster and better profiles of record-level health. From these firm foundations of evidence, informed decisions can be made to accelerate global disease eradication.