The most powerful tech companies in the world are envisioning the future of the healthcare sector. You will visit the doctor's office, sit in the waiting room, and explain your problems to someone in a white coat; it all sounds too familiar. However, change begins in the treatment mechanism. Rather than relying on expertise and knowledge, the doctor consults an algorithm system trained to detect and diagnose the symptoms of millions of patients, and identify the suitable outputs for treatment. Instead of relying on a radiologist to read X-ray images, a special computer can be used to detect abnormalities and locate tumors or wounds, which is ultimately the desired goal.

These AI systems, developed by international companies like Google and IBM, are incapable of reading books and reports, attending lectures, and doing rounds. Instead, they need millions of real-life examples to understand the different variables between one patient and another. In general, AI quality depends on the data on which it was trained. However, it is difficult to apply this rule to medical data, especially due to its confidential nature. Even the most developed countries treat them as confidential medical information protected by data-protection acts, such as the Privacy Act on the protection of medical information in the United States.

In light of these challenges, companies interested in implementing AI systems in the healthcare sector are eager to find different ways to collect as much data as possible. For example, Google has partnered with Stanford and Chicago university hospitals to gather 46 million data points on patient visits. As for Verily, also owned by Google's Alphabet, it hired 10,000 employees to conduct long-term health studies. In turn, IBM has acquired medical companies in the past few years to obtain their data and gather records of more than 300 million persons.

This methodology, more beneficial to companies with significant resources, is the only way to obtain large sets of health data in the United States due to the visible discrepancies and disparities in the American health system. As usual, healthcare providers keep personal patient files that cannot be shared with any other accredited medical entity, except at the patient's request. Consequently, there is no single space that contains all the health data. Although this mechanism provides high levels of security, it is less effective in terms of research and analysis.

Canada's Ontario seems to have found a solution thanks to its single-payer healthcare system. The latter collects all health information generated by Ontario and maintains it in several large digital databases controlled by the government. (In all cases, the government is allowed to track the bills you pay). Similar organizations can be found in other Canadian regions, such as the province of Quebec. Toronto, which has become a hub for research in the field of artificial intelligence, is making an extra effort to provide these data to companies.

So far, only public entities or researchers who have signed partnerships with the government to study specific diseases have permission to examine and analyze these data. Ontario has entered into a contract with MaRS Discovery District, a cross between a tech incubator and WeWork, and assigned it to build a platform for approved companies and researchers that allows them to access the data, an initiative dubbed Project Spark. The project, launched by MaRS and Canada’s University Health Network, started exploring how to share this data after both entities expressed their desire to the government to offer broader health data access to researchers and companies seeking to develop healthcare tools.

Project Spark aims at building an Application Programming Interface (API) to possibly provide developers with an effective mechanism that requests information from governmental databases. This platform can be broadly used to develop an app that allows doctors to access the full medical history of any first-time patient. Furthermore, Ontarians can access their health records anytime and anywhere with similar software. They can also catalog health issues based on the date they occurred.

There are currently 100 companies waiting to access the data that include the health records of over 14 million Ontarians (MaRS did not yet reveal the names of these companies). Access to high-quality data of millions of patients is perceived as free money for tech companies. The project currently focuses on developing software for patients and medical staff. 

“We are always receiving data that is not necessarily passed around or shared with other entities. This is where the platform steps in by trying to take advantage of being the owner of this consistent and rich source of information while subjecting users to the appropriate governance model," says Joe Greenwood, leader of the Spark project for MaRS, while discussing the platform.

Thanks to this ecosystem, a record is kept for every time an Ontarian contacts a healthcare provider for a routine checkup, an appointment with a specialist, or a medical scan.

This is not considered a major challenge during the first stage of the project, which has not started yet and has no specific timeline. Only doctors, patients, nurses, and specialists can determine if the data is sensitive or not. When it comes to research and people outside the hospital or doctor’s clinic, MaRS is still figuring out how to handle their sensitive data.

If you are interested in studying the health status of Ontarians, all you need to do is access valuable medical records to answer all of your questions. The main questions here are: Why is Google hiring some of the brightest minds to research health-related AI systems? Why would the company allocate over $100 million for a medical study involving 10,000 persons?  According to the World Economic Forum, the global healthcare market is worth $6.5 trillion, yet it remains incapable of developing the required software.

For this reason, MaRS is working with researchers, such as those attempting to build AI systems, to assist doctors in collaboration with Ontario’s Institute for Clinical Evaluative Sciences to receive larger volumes of anonymized data on Ontarians who meet certain criteria. If you want to research every Ontarian who suffered from Alzheimer’s over the last 40 years, the data is only an authorization away. 

On another note, this data, available for sharing, would have a positive impact on enhancing convenience and would make things easier for Ontarians. Although health data is currently maintained in large databases, a request is to be submitted to access health records and transfer them to the doctor, as it is the case in the United States. This separate project will be an opportunity to develop software that complies with health information privacy laws and allows doctors to instantly obtain the patient’s full medical history.

Any person wishing to reach patient data in Ontario must contact the International Credential Evaluation Service (ICES). Liisa Jaakkimainen, a family doctor who worked at ICES for 20 years, says that one of the risks of making this data accessible to a large number of people is the lack of control over how it is analyzed and used.

"Doctors are worried that data will be used for marketing research purposes. I believe this would be extremely upsetting for doctors if data were used to make a profit. This is truly disturbing and unacceptable. Data will be provided to improve the healthcare sector, but where is the line between enhancing healthcare and generating a profit within a certain company? This issue is sometimes clouded by ambiguity,” Liisa explained.