AI holds a huge promise for transforming healthcare, but does it truly offer the revolution we need, or does it simply reinforce existing routines and flaws? This blog explores how the interplay of technology, context, and people shapes AI’s impact in healthcare, and how technological determinism can hinder its success.

Healthcare systems worldwide are struggling with ageing populations, increasing demands, and huge staff shortages. These pressures highlight the need for innovations, where AI is seen as a promising technology. Yet, expectations often outpace reality. For instance, Health Minister Fleur Agema envisions a revolutionary role for AI, including halving administrative workloads by 2030.

AI is already in use, and some isolated successes are realized in areas like radiology, where AI-powered imaging detects early signs of diseases such as cancer with greater accuracy than radiologists. Natural language processing reduces physicians’ documentation burdens by transcribing medical notes, while AI-driven virtual assistants streamline patient triage, freeing healthcare professionals to focus on complex cases.

Despite these apparent successes, integrating AI into healthcare systems is challenging. Research shows that new technologies often reinforce existing routines rather than drive meaningful change. For example, while AI can analyze electronic health records (EHRs) to predict patient outcomes, the cumbersome nature of these systems and fragmented workflows often lead to workarounds that undermine AI’s effectiveness.

Technological determinism is a significant obstacle in this context. This common perspective rests on the mistaken belief that technology alone can drive change, independent of the social, economic, and political environments in which it operates. Such a mindset can lead to an overconfidence in implementing AI while neglecting the structural and cultural barriers that hinder its effective use.

The Social Shaping of Technology theory emphasizes that technology does not exist in a vacuum. Instead, its impact is shaped by the contexts in which it is implemented. Without careful consideration of these factors, AI can end up reinforcing inefficiencies or even creating new problems rather than solving healthcare’s biggest problems.

Unlocking AI's potential requires a shift from viewing it solely as a tool for efficiency to recognizing its role in enhancing care quality and user experience. Our own research showed that clinical Decision Support Systems can distract clinicians with unnecessary alerts, causing “alarm fatigue” and system ignorance. However, when carefully tailored, such as flagging high-risk patients, such systems can save lives by prioritizing critical information.

This demonstrates the importance of sociotechnical design, this is an approach that aligns AI tools with the needs of both healthcare professionals and patients. For example, wearable devices like continuous glucose monitors, combined with AI algorithms, can offer diabetics real-time insights for managing their condition. When integrated with care plans and provider training, these tools improve both positive health outcomes and patient satisfaction.

So, what is holding us back from reaping the benefits of AI in healthcare? The answer lies in recognizing and addressing the interplay of technology, context, and people. By moving beyond technological determinism and focusing on sociotechnical design, we can ensure that AI solutions are not just quick fixes but lasting, meaningful and systemic changes. Success will depend on interdisciplinary collaboration, adaptive policies, and a commitment to ongoing refinement based on real needs and user feedback.

The transformative potential of AI in healthcare is there, but only when it is thoughtfully designed, implemented and integrated.

Author: Albert Boonstra - albert.boonstra@rug.nl