Big data is as important to medical science as medical science itself. In fact it is often the harbinger of news according to Dr Mark Baker.
“Consider the link between smoking and cancer,” he says.
“Originally people thought smoking was actually helpful for people with lung conditions. The coughing up of phlegm was seen as a sign that the cigarettes were a positive influence. The smoke expelled by mouth was also seen as a virtue.’
As Dr Baker points out it was only when Sir Richard Doll, then an epidemiologist working in the aftermath of world war 2, began to look at why Britain had one of the highest instances of lung cancers, that his work with statistics began to make a difference. Originally the high instances were attributed to environmental causes such as the tar on the roads, but Sir Doll’s pioneering research in Oxford University into using statistical data proved the link between smoking and lung cancer, and other diseases.
— MediChain (@MediChainOnline) March 14, 2018
It was a long process battling big Tobacco but Sir Doll’s statistical approach heralded a new era in medicine: the intellectual ascendancy of medical statistics. Dr Baker also attended Oxford: “Richard Doll as an intellectual mentor for my career,” he says.
“Similarly statistical analysis was the used in tracing cholera to fresh water supplies. And now we look at the use of big data in examining current issues such obesity, mental health and cancer.”
Dr Baker’s first foray as a medical scientist into big data was while at Oxford early in his career. The UK Cancer Research Trust sponsored his work in reviewing data to determine trends and causal effects. He moved to Cambridge and back again to Oxford working in early big data projects.
A brief stint working with Johnson and Johnson in the trails into their anti-psychotic drug Risperidone led him to examine the sleep patterns of patients using the drug.
“I established the largest sleep laboratory in the world,” says Dr Baker. “We set up the Laboratory with each subject wired up with 40 electrodes which continuously recorded information over a ten hour period.
“In the end we had tens of thousands of physical tapes containing the data that required a storage warehouse the size of an aerodrome. Originally we had skilled laboratory technicians recording and analysing the data but the sheer volume made that impossible. It was this stage that we introduced Artificial Intelligence (AI) to simulate the workings of the technicians, using AI and machine learning to analyse the data.”
According to Dr Baker this process was a great success and the drug was approved for use and went on to make 42 million sterling for the manufacturer and to this day improves the quality of the patients using it.
Dr Baker was now hooked on statistical analysis. He became CTO of Europe’s largest predictive analysis company which used code running on every European retail website to examine retail buying patterns.
Dr Baker insists this company was not snooping on individual’s privacy. “I am very mindful of the dignity of people’s personal information. All the data was collected under strict data privacy regulations which allow for the anonymization of the person behind the data.
“But it was a powerful use case of how data from millions of people, some 50 million, could be extracted, anonymised and used to create predictions.”
Dr Baker moved back into mainstream medicine again, this time with a project for the NHS looking at the outcomes from different treatments applied by doctors in the UK.
The ambition was to bundle up sets of patients with similar medical conditions, look at the different treatments by different regions down to individual doctors, and record the outcomes.
The project, while having the possibility of getting over the old adage that doctors differ and patients die, was shelved by the NHS for, well NHS reasons, and this was the impetus that led Dr Baker to initiative his own project.
Already a fan of AI and increasingly of blockchain, Dr Baker set up Medichain, a project aiming to do two things: firstly to give access to patient data regardless of origin and secondly to build up a global platform for big data to be gathered on an infinite range of medical conditions.
Blockchain is new technology underpinning modern innovative projects from cryptocurrencies, to banking to the means of sharing value without the need for intermediaries. It has value most particularly for this project in its immutability. Data once collected cannot be tampered with. Putting Medichain on the blockchain also opens the door to funding in cryptocurrencies. Medichain aims to raise anything from between $2million and $40million to build its software platform: figures which are not unusual for cryptocurrency companies.
MediChain states that its core principle is that data is owned by the individual, not the system.
“It has taken us a year to get this far,” he explains, “as we are working in highly regulated environment. All patient data must be protected under the HIPAA, the Health Insurance Portability and Accountability Act, we only provide access to where the data is: we don’t record the data directly.”
This has advantages to patients who can direct doctors to retrieve medical data from clinics or hospitals in different geographies or from different years. Core to the patient control is that patients can direct which of their records can be seen by anyone else. This has applicability in America where medical insurance is very expensive and sometimes people can be denied cover with certain conditions. This can in turn lead to potential job loss in some cases according to Dr Baker.
— MediChain (@MediChainOnline) March 16, 2018
“If a patient does not want to share sensitive information, they don’t have to,” insists Dr Baker.
On the flip side, pharmaceutical companies devour data when developing drugs. ‘This can be facilitated by authorised permissions,” he says.
“In some cases patients with a particular condition may consider donating their data with the view that a cure may be advanced. Or in some cultures they may receive a payment for their data.
“I don’t think patients in the UK would look for payment: that would just be weird.”
Dr Baker maintains that big data can hold big business, like pharmaceuticals, to account. “If drugs are being oversold, then big data research can filter out untrue claims.
“It works both ways – big data found out tobacco was linked to cancer and it can also find cures for cancer.”
— MediChain (@MediChainOnline) March 12, 2018