Using big data to fight chronic diseases

In Singapore, a fast-ageing population is spurring the rising burden of chronic diseases. This will place great strain on the already-stretched healthcare system in the coming years.

In a commentary in this newspaper last month, Dr Lee Kheng Hock highlighted this problem, which he said could be addressed by focusing on community healthcare through collaborative networks of general practitioners (GPs) and by training family physicians. In a follow-up commentary, Dr Jeremy Lim took the debate further by noting that most chronic diseases are managed in the primary-care system (GP clinics and polyclinics), which is driven by a “fee-for-service” model that emphasises volume and profitability of services.

This is not a sustainable framework to handle chronic diseases, Dr Lim rightly argues. Instead, a better financial model for managing patients with chronic diseases is capitation, a system where the doctors are paid a guaranteed sum per patient for a specified total population. This provides incentives for all stakeholders towards keeping the patient as healthy as possible at minimal cost, and forms the basis for value-based healthcare. As Dr Lim noted, capitation models were eschewed previously because quality of care was difficult, if not impossible, to ascertain in a data-poor milieu. But the situation today is radically different, with electronic records readily available, making objective assessments of clinical quality and payments feasible.

Technology is the key enabler, particularly big-data analytics, in the transformation to such a value-based healthcare system. So, how does it work?

WHAT BIG DATA CAN DO

Big data refers to large, complex sets of data that are difficult to access or manage traditionally. Big data in healthcare consists of electronic health records, doctors’ notes, pharmacy prescriptions, insurance claims, sensor data (such as blood pressure, glucose level), genetic information, and more.

Unlocking the hidden value in this information has the potential to improve care and attain cost effectiveness. Data analytics allows us to synthesise and discover patterns and correlations within the data that would not have been revealed otherwise.

Doctors will be able to make better and more accurate diagnoses, and provide the patient with more personalised treatment plans based on their profiles. Genetic information, though not widely available yet, can be integrated with traditional medical data to help doctors decide on tailored treatment with a higher degree of success.

Real-time analysis of large volumes of data from vital monitors and sensors can reveal early signs of a patient’s deteriorating condition, long before a nurse or doctor detects it.

Big data analyses also accelerate research and development efforts in drug development and medical therapy, and improve the analysis of clinical trial results.

Prevention and lifestyle management

Chronic disease is an area where big data and predictive analytics can be of real benefit. To stem the tide of chronic diseases, a new model of care focused on prevention and lifestyle management is required.

An example of this is diabetes, which is one of the most prevalent chronic diseases in Singapore. An even larger pool of people are prediabetic, which is the precursor to diabetes.

Analytics of available health data can better identify prediabetics who are at highest risk of developing type 2 diabetes. Focusing preventive measures on this group of people will be highly cost-effective. These individuals will benefit most from proven interventions such as the Diabetes Prevention Programme, which showed that intensive lifestyle therapy involving weight loss, dietary changes and physical activity significantly reduce the risk of developing diabetes.

For patients already living with diabetes, technology and analytics will be able to influence their health outcomes. The biggest component of direct medical costs for diabetes is hospital inpatient care and frequent hospital readmissions.

Reducing readmission rates will significantly improve the quality of care and reduce costs. Predictive analytics can stratify patients into different readmission risk categories based on their past medical history, clinical markers and even demographic factors, and greater resources can be allocated to ensure that high-risk patients receive close follow-up care, reducing avoidable readmissions.

A patient’s engagement in the care of their health strongly contributes to improved health outcomes. People with diabetes who are able to positively modify their lifestyle have improved glucose control, reduced risk of diabetic complications, and require fewer medications.

To engage patients in making better choices, timely feedback is vital. Using big-data analytic techniques, dynamic and predictive measurement tools can be built. These tools integrate clinical parameters, biochemical markers, personal nutrition and physical activity information to provide the patient with a more current indicator of their disease status.

Being able to witness on a continuous basis the results of lifestyle changes will be a strong motivator for positive habit formation.

The global landscape for healthcare is evolving rapidly. A “fee-for-service” system has worked well in the past, but a fundamental new path is needed to optimise value for patients.

Technology will spur this change, as medical services become more akin to consumer goods. In this new landscape, big data will play an increasingly important role. Big data and predictive analytics offer tremendous opportunities.

ABOUT THE AUTHOR:

Dr Yau Teng Yan is chief medical officer at Holmusk, a big data and digital health company.