'No blanket approval' to use personal health data in research, focus groups say
Big Data has been called medicine's future, with predictions that artificial intelligence helping researchers sift through every blood pressure, every ECG and MRI looking for cures. But there are fears that data mining is ripe for misuse. A survey published today in CMAJ Open asked the public to weigh in.
ICES is an independent, not-for-profit research institute in Ontario. Since 1992, the institute has been sifting through health-care data derived from the nearly 14 million people who live in the province. Every visit to the doctor or clinic, every lab test or X-ray, and every hospital admission produces data that is collected by the province. The data helps the province manage, administer and pay for health care services.
ICES has the authority to analyze the data. Researchers crunch the data to study the health of Ontarians and the performance of the health care system itself. For instance, a seminal study linking cell phone use while driving and motor vehicle collisions was done using data provided to ICES.
Researchers at the institute wanted to find out what the public thinks of this kind of research. They conducted eight focus groups with a total of 65 people from three cities in Ontario: Toronto, Sudbury and Thunder Bay. In each focus group, participants were given background information about ICES and how it gathers health care data generated by the province. Participants were then asked to talk about specific examples of research studies based on ICES data.
Among those surveyed, they found general support for the kind of research ICES does provided some conditions are met.
Support was strongest when participants could see an obvious benefit to the research. For example, they gave the strongest approval when health data was used to study the long-term safety and efficacy of prescription medications. They were more likely to express disapproval of this sort of research when they thought the results of a study could be misused or could disadvantage certain groups, for instance, seniors.
The researchers also found that participants knew and understood little about how this kind of health-care data is used to do studies. This was despite regular media coverage of the research. Some of those surveyed were under the misconception that researchers needed to ask for consent to use health data even when identifying information had been removed.
One of the major themes to come out of the survey is that participants needed to be reassured that their personal health data was being kept private. There was a high level of awareness about the hacking of the Canada Revenue Agency database. Those surveyed said they were glad to find out the Information and Privacy Commissioner of Ontario has oversight over the way health data is used. As well, participants said they appreciated that ICES removes or codes identifying information before the data is made available to researchers.
However, some of the participants felt these measures were not sufficient. Some felt there are unavoidable increases in the risk of data breaches the more people and organizations have access to the information. Some focus groups expressed concerns about pharmaceutical companies and other members of the private sector being allowed access to this kind of health data. Several thought private sector involvement was acceptable if it led to lower drug prices.
Failure not an option
There are several important lessons to gain from this survey. The public is generally on side with the notion that this rich treasure trove of health data is an asset that can and should be used and developed. But it's not blanket approval. There is a growing movement in health care to involve patients and members of the public in redesigning the system to better meet the needs of patients. Along similar lines, the big lesson is that researchers and organizations like ICES should involve members of the public to ensure that the research is both trusted and trustworthy.
- ANALYSIS | How one researcher harvested data from 50 million people — and Facebook was designed to help
Another lesson is that the research and the sharing of data needs to be transparent. News stories of Cambridge Analytica's reported misuse of Facebook data breeds lack of trust. The study's authors concluded that acting transparently by itself is insufficient to restore faith. The next step is to involve patients and the public by giving them a degree of decision-making authority over this kind of research.
The public has already seen some of the benefits we've seen so far from this kind of research. The aforementioned study linking cellphone use to motor vehicle collisions led to legislation restricting smart phone use while driving. A study showing excess heart attacks linked to the use of the drug Vioxx led to its withdrawal from the market. A study showing the magnitude of the opioid epidemic led to greater efforts prevent overdoses and deaths.
But growing cynicism brought on by fake news and breaches of privacy mean that it won't take much to turn the public against this sort of publicly-funded research. The loss to the public would be staggering. Failure is not an option.