Too much healthcare data? AI can help

As Singapore's population ages, there is greater focus on healthcare innovation and greater investments in digital health.

For the elderly and disabled for example, developments in assistive technology, analytics and robotics are making a difference to the way they complete tasks and activities.

However, increasing digitalisation of health records is producing more data, relating to patient health status and delivery of care. This has resulted in a new data dilemma: Having too much data and not knowing how to unlock its full value.

According to International Data Corporation (IDC), global healthcare data is predicted to grow from 153 exabytes in 2013 to 2314 exabytes in 2020.

One exabyte is equivalent to one billion gigabytes.

From a clinical perspective, current health IT requires clinicians to use common database tools, such as the electronic health record (EHR) to retrieve quantifiable, measurable data.

This provides clinicians with patient information like lab results, blood sugar levels and cholesterol, which they have to manually inspect, making it hard to identify critical information. So healthcare providers need to ensure they have the right tools to gather, organise and analyse this information.

WHY DATA MATTERS

In today's environment, clinical evidence is developed based on scientific research, which is then used to develop clinical decision support (CDS) tools to be implemented in care settings.

Such tools are meant to help medical practitioners make the best decisions at the point of care, addressing inconsistent care practices, which pose the greatest threat to the quality, cost efficiency and outcomes of care delivery.

However, while scientific research is important for CDS, this research data is not always available and it takes a long time to find its way into practice.

Research has found that it takes 17 years for only 14 per cent of new scientific discoveries to find their way to daily practice. This leads to a high degree of knowledge variability that directly impacts the quality and safety of patient care.

THE PROMISE OF AI

AI has the potential to reduce the time for new discoveries to find a way into practice.

It can analyse existing patient data and clinical data to draw new hypotheses not covered in larger scientific research.

However, real-world evidence data alone is not enough.

By combining patient-generated data with academic evidence, healthcare professionals are able to create personalised treatment options.

AI could be an even bigger catalyst to provide them with real-time capability to develop real-world evidence, which can positively impact patient outcomes.

Moreover, with deep analytics, data collected through structured content can be easily analysed to feedback into care practices.

THE THREAT OF CYBER RISKS While digitalisation can reap benefits for the country's healthcare system, there is an additional dilemma.

The huge quantities of patient data may become a major target of cyber attacks.

For instance, SingHealth's data breach in July resulted in the health data records of 1.5 million people being stolen.

However, in the words of Minister for Communications and Information S. Iswaran, such incidents should not "cow or deter us from pursuing this path, because it is a path that's going to create the opportunities for this and future generations of Singaporeans".

Resolving this dilemma necessitates a balance between pursuits of the future that such sophisticated technology promises, and ensuring there are safeguards in place to protect the identity of patients.

The writer is managing director for Asia-Pacific at Elsevier, an information and analytics company. This is an edited version of an article that appeared in The Business Times yesterday.