Big data will change the way South Africans receive health care and how medical schemes administer it and Metropolitan Health, one of the country’s largest medical scheme administrators, wants to be at the forefront of the change, says CEO Dylan Garnett.

The company, which has 3m individuals on its books, is deploying “cognitive processing” and other advanced technologies in the hope of bolstering the capabilities and effectiveness of health care in the public and private sectors. It’s signed a broad-ranging agreement with IBM to do this.

“If used strategically, technology can assist in transforming the entire health ecosystem and it permeates every dimension of health care,” Garnett tells TechCentral. “We believe that the right technology applied in the right environment and with the right intent is a powerful tool.”

Metropolitan’s collaboration with IBM is meant to provide better and more cost-effective services to members and the medical schemes that serve them. And large volumes of easily accessible data will play a key role, says Garnett.

The company operates one of the largest medical schemes in South Africa, serving approximately 3m individuals across 19 medical schemes and a number of large corporations, including Absa, Standard Bank, First National Bank, Woolworths, BP Southern Africa, Engen and Pick ’n Pay.

Metropolitan Health is also the administrator of government’s employee medical schemes. It forms part of MMI Holdings, a financial services group that was formed after the merger of Metropolitan and Momentum in 2010.

Garnett says Metropolitan Health plans to collect and analyse large volumes of patient data and give that data context. “This will allow us to make smarter decisions and constantly adapt and improve our offerings, allowing us to reach far more South Africans.”

The challenge is taking raw data and turning it into something that will allow medical scheme administrators to have a “useful conversation” with their clients, he says. This could mean selling them new products or simply offering them medical and related advice.

In medical terms, contextual data is key to giving patients the best advice possible, while collating medical data so that it is accessible by the doctor or specialist providing care is imperative.

Patients’ records are typically made up of structured and unstructured data. The latter is typically made up of data sets that are not easily or broadly accessible — doctors’ notes, for example, or a patient’s lifestyle habits. Cognitive processing can play an important role here.

In layman’s terms, cognitive computing is the process of taking complex data sets and information and using clever algorithms to derive knowledge and context from the data.

IBM’s Watson technology is one example of cognitive computing. Watson is able to process data and output meaningful information based on that data. In future, computers will do much more than just compute. They’ll be able to sense, learn and predict the consequences of actions or deliver results based on historical and predictive data.

The goal of these technologies is not only to provide better health care to patients, but also to provide more accurate risk assessments.

Quoting a recent IBM study, Garnett says only 10% of a person’s future health can be predicted through health claims data — doctors’ notes, prescription information from pharmacies and so on. To predict what is going to happen in the future, or in some way influence a patient’s health risk positively, more information is needed.

According to the study, 40% of a patient’s future health risks are determined by their behavioural data. This type of data is typically highly unstructured and includes such things as what someone eats and whether they exercise or not. It could even include someone’s mental status based on what they post to Twitter or Facebook.

A person’s genetic profile is also a big determinant of risk. According to the IBM study, almost a third of some’s future health risk predictability is in their genetic make-up.

This is typically the data that shows someone’s disposition for certain medical conditions.

Garnett explains how some people can following certain eating plans without any negative effect on their health, while the same diet could negatively impact the health of others. Garnett says a quick and cheap way to check one’s genetic profile is to look at their family medical history.

This data, which forms part of the data sets required for cognitive processing, will ultimately lead to better health care capabilities as the data is processed into workable, contextual information.

Garnett says the real benefits of Metropolitan’s partnership with IBM will only become clear after it makes an announcement in a couple of months’ time. It will be able to show “real deliverables and practical applications and toolsets based on cognitive processing and its capabilities”, he says.

Metropolitan Health also intends to solve part of the unstructured data problem facing public sector health, applying cognitive processing to it and converting that data into smart advice for patients.

Cognitive processing will provide better insights and advice for medical scheme clients and patients. Garnett believes we can start to use technology effectively to improve treatment protocols and outcomes using the data that is collected on patients.