AI in Healthcare - A Risk to Patient Safety
AI is being used more and more in healthcare, but the lack of guidelines and uncertainty around handling sensitive data makes the technology risky for patients. The Swedish Medical Association is calling for clearer government guidelines on how and when AI should be used in healthcare.
AI technology is already being tested and used in several medical fields, for example to detect breast cancer in mammography images and to make initial assessments of patients in emergency care. Despite the potential, many doctors feel there are risks to using AI without clear guidelines. A survey from the Swedish Medical Association shows a quarter of Swedish doctors use AI in their daily work, often for writing medical records. At the same time, 55 percent of respondents say they lack a policy for when and how AI should be used.
Uncertainty Around Sensitive Data and Legal Consequences
One of the biggest challenges with AI in healthcare is how the technology handles sensitive patient data. Sofia Rydgren Stale, chair of the Swedish Medical Association, points out that healthcare is a particularly vulnerable environment, where incorrect conclusions can have serious consequences. Despite this, employers don't appear to have prioritized the issue of AI use. Rydgren Stale does see opportunities for AI to make healthcare more efficient, but without clear rules and guidelines, Sweden risks falling behind in AI development. According to the survey, 47 percent of doctors feel uncertain about the legal consequences of increased AI use.
The Need for Government Guidelines
The Swedish Medical Products Agency has developed general guidelines for AI in healthcare, recommending that healthcare providers conduct a risk and impact assessment before introducing AI technology. The government has also set up an AI Commission, which will report its findings in 2025. The commission's purpose is to strengthen AI development in Sweden, but it addresses healthcare only to a limited extent.
To ensure safe, secure use of AI in healthcare, Rydgren Stale believes the state needs to take greater responsibility. She emphasizes that more specific guidelines are needed for how and when AI may be used in healthcare.
- There's a political expectation that AI can save on staffing and make healthcare more efficient. But many tasks in healthcare and social care are hard to automate, says Rydgren Stale.