Skip to ContentSkip to Navigation
founded in 1614  -  top 100 university
About us Faculty of Economics and Business Research at the Faculty of Economics and Business FEB Research News
Header image Faculty of Economics and Business

Why digital technology in healthcare often has different outcomes

Date:17 November 2026
Technology has no fixed results
Technology has no fixed results

Many healthcare organisations are currently experimenting with AI that automatically turns conversations between healthcare professionals and patients into clinical notes. The promise is attractive: less administrative burden, more time for patients, and more efficient care. But are these optimistic expectations true?

Professor Albert Boonstra looked at these optimistic expectations and found they are often not realised. Healthcare professionals may spend considerable time reviewing and correcting AI-generated notes, while confidently phrased errors can easily be overlooked. Important information may be missing or misrepresented, and old and new ways of working may continue to exist side by side. As a result, workloads may actually increase, while the quality of care may also come under pressure.

This raises a broader question: why does digital technology in healthcare so often have different outcomes than expected? The obvious explanation is that something has gone wrong with implementation. Management may assume, for example, that insufficient attention has been paid to training, communication, or usability. But another explanation is possible. Perhaps the problem lies not so much in the implementation itself, but in the way we think technology changes organisations.

The limitations of technological determinism

Many digital innovations in healthcare implicitly rest on a simple causal model that is widely embraced. Electronic health records make information more accessible and therefore lead to better care. Telehealth replaces in-person consultations and thereby increases efficiency. Artificial intelligence takes over tasks from healthcare professionals and therefore reduces the need for staff.

Although these examples differ, the reasoning follows the same pattern: introducing technology causes a desired change. Better information leads to better care, telehealth to greater efficiency, and AI to lower staffing requirements. These are examples of technological determinism: the assumption that technology is the main cause of change and produces more or less predictable effects once introduced.

Boonstra explains: "My research on more than thirty years of digitalisation in healthcare presents a more complex picture. Electronic health records have made medical information more accessible, but have also increased the administrative burden. Telehealth enables care at a distance, but by no means always replaces existing contacts and may increase the overall volume of care. AI supports diagnosis and documentation, but at the same time creates new work related to validation, monitoring, and professional coordination.

These examples reveal a recurring pattern. Digital technology can solve existing problems, but at the same time changes the context in which it is used, thereby creating new tasks, dependencies, and problems."

Digital transformation as a sociotechnical process

A sociotechnical perspective therefore starts from a different assumption: technology does not have fixed effects. Outcomes emerge from the interaction between technological possibilities and the social, organisational, and institutional context in which technology is used.

A digital application becomes embedded in existing work processes, professional routines, organisational structures, and regulations. But this context does not remain unchanged. Healthcare professionals adapt their ways of working, organisations revise procedures, vendors develop new functionalities, and regulators formulate new requirements. Technology and organisation thus evolve through a continuous process of mutual adaptation.

This perspective explains why the same technology can produce different outcomes in different organisations. When technological possibilities, work processes, and professional expertise are well aligned, quality and productivity may improve. When this alignment is lacking, the same applications may instead create additional work, coordination problems, and higher costs.

But mutual adaptation has another consequence. When technology makes a task faster, easier, or cheaper, this often changes how frequently and extensively that task is performed. An individual activity may therefore become more efficient without reducing the total amount of work or overall costs. Sometimes the opposite occurs: because the activity becomes easier, it is performed more often or more extensively. This mechanism is known as the Jevons paradox.

Telehealth, for example, may make an individual consultation more efficient while simultaneously lowering the threshold for accessing care and thereby increasing the total number of contacts. AI may speed up documentation for each patient, but may also lead to more extensive documentation, additional checks, and new forms of monitoring. Efficiency gains at the task level may therefore be offset by changes at the level of the organisation or the healthcare system.

The relevant question is therefore not only whether a technology makes a particular task more efficient, but above all how that efficiency gain changes behaviour, working practices, and the use of healthcare.

Technology amplifies what is already in motion

The interaction between technology and context has another important consequence. New technologies rarely set organisations in motion from scratch. They interact with developments that are already under way and may strengthen, accelerate, or broaden them.

Electronic health records reinforced the move towards standardisation, accountability, and data exchange. Telehealth aligned with the shift of care towards the home and made it possible to accelerate this development. AI is reinforcing the move towards data-driven diagnosis, decision-making, and quality monitoring.

From this perspective, digital technology does not function as an autonomous engine of change, but rather as a catalyst for existing developments. The effects that emerge therefore depend not only on the properties of the technology, but also on the direction in which an organisation or healthcare system is already developing.

This also has practical significance. Anyone seeking to understand the possible consequences of a new technology should therefore ask not only what the technology can do, but also what it acts upon. What developments are already under way? Which working practices and interests are being reinforced? And which existing problems might thereby be amplified?

Design the change, not only the technology

A sociotechnical perspective leads to a different approach to action. Digital transformation should not begin with the question of how a technology can be implemented as quickly and successfully as possible. The first questions should be: what problem are we trying to solve, what changes in work and organisation are needed to address it, and how should technology, work processes, and organisation develop together? This requires more than user participation or training. It requires joint design, experimentation, evaluation, and continuous adaptation. Not only must the organisation adapt to the technology; the technology must also be adaptable to practice.

Digital transformation should therefore not be managed as the implementation of a technical system with a clear endpoint, but as an ongoing process of organisational change and sociotechnical design. This also answers the question raised at the beginning of this article. Digital technology in healthcare often has different outcomes than expected, not because the technology itself falls short, but because it changes the way people work, collaborate, and organise care.

Those who implement only the technology leave these changes largely to chance. Approaching digital transformation as a process of sociotechnical design does not prevent new problems from emerging. It does, however, create more room to recognise unexpected effects at an early stage, bring competing interest to the open, and adjust technology, work processes, and organisation in an integrated way.

For more information, please contact Professor Albert Boonstra (albert.boonstra@rug.nl).

Share this Facebook LinkedIn