Beyond access: why digital health requires critical engagement in the age of AI

The expansion of digital health over the past decade has been both rapid and necessary.

Across regions, investments in electronic health records, telemedicine platforms, and AI-enabled diagnostic tools have largely been framed as a question of access how to ensure that systems are available, scalable, and integrated into healthcare delivery.

This framing has served an important purpose. It has enabled governments and institutions to prioritise infrastructure, reduce gaps in service delivery, and extend care beyond traditional settings.

Yet, as digital health systems become more deeply embedded, a quieter question is beginning to emerge: is access, on its own, enough?

Increasingly, the answer is no.

Access ensures availability. It does not ensure understanding. In a healthcare system that is becoming progressively data-driven and algorithmically supported, this distinction is no longer marginal it is foundational.

Digital health systems today are not passive tools. They generate recommendations, structure decisions, and, in many cases, shape clinical judgment.

AI-enabled diagnostic systems highlight anomalies in imaging. Clinical decision support tools prioritise treatment pathways. Predictive models identify patients at risk of deterioration. These systems are designed to augment human capability, and in many cases, they do so effectively.

But their value depends on how they are used.

The effective use of digital health systems requires the ability to interpret outputs, understand limitations, question recommendations, and integrate them into real-world decision-making contexts. Without this, two risks begin to surface.

The first is underutilisation where systems are available but not meaningfully used, often due to lack of trust, training, or contextual relevance. This can be seen in settings where clinical decision support tools are deployed but bypassed in practice, or where telemedicine platforms remain underused despite availability due to workflow misalignment or clinician hesitation.

The second is over-reliance where outputs are accepted without sufficient scrutiny, leading to automation bias and potential clinical error.

For instance, AI-enabled diagnostic systems are at times treated as definitive rather than assistive, or triage algorithms influence prioritisation decisions without adequate contextual judgment, particularly in high-pressure environments. Both outcomes undermine the very objective these technologies are intended to serve.

This introduces a new dimension to digital health policy one that has received comparatively less attention. Much of the current policy focus across jurisdictions has centred on infrastructure, interoperability, and regulatory approval.

These are necessary foundations. However, they operate on an implicit assumption: that once systems are deployed, they will be used effectively.

In practice, this assumption does not always hold.

Healthcare systems are complex environments. Clinical decision-making is shaped not only by data, but by time constraints, institutional protocols, resource availability, and human judgment.

Introducing digital systems into this environment does not automatically translate into improved outcomes. It requires alignment between technology, workflows, and the people who use them.

This is where the concept of critical engagement becomes central.

Critical engagement refers to the capacity of individuals and institutions to interact meaningfully with digital systems not simply to use them, but to understand their role within a broader decision-making process. It includes the ability to assess when a system is helpful, when it may be limited, and when human judgment should take precedence.

Importantly, this is not a question of technical literacy alone. It is institutional. It involves training, governance structures, accountability frameworks, and the integration of digital tools into clinical workflows in ways that reflect real-world conditions.

The gap between access and effective use is not unique to digital health. It reflects a broader pattern observed across public policy. Systems are often deployed with the expectation of uniform adoption, yet real-world implementation is shaped by human judgment, institutional constraints, and local context.

Frontline actors interpret and adapt tools based on time pressures, trust, and workflow realities. At the same time, the adoption of new technologies depends not only on availability, but on how well they align with existing systems, how observable their value is, and whether users are equipped to engage with them meaningfully. Without this alignment, systems do not fail visibly they simply fail to deliver their intended value.

Beyond principles: the next phase of AI governance in healthcare

We are beginning to see early signals of this gap. In some settings, advanced digital tools remain underused due to lack of integration or clinician confidence. In others, there is growing concern around over-reliance on algorithmic outputs, particularly in high-pressure environments where time for independent verification is limited. These are not failures of technology, but of alignment.

As digital health systems continue to scale particularly across emerging and rapidly digitising markets this challenge becomes more pronounced. Many health systems are simultaneously building infrastructure and adopting advanced technologies. The pace of adoption can, at times, outstrip the development of institutional capacity required to engage with these systems effectively.

This suggests that the next phase of digital health cannot be defined solely by expansion. It must also be defined by readiness not just in terms of infrastructure, but in terms of capability.

From a policy perspective, this raises several considerations.

First, workforce development must evolve beyond basic digital literacy toward applied competencies how to interpret, question, and act on algorithmic outputs in clinical contexts.

Second, governance frameworks should extend beyond system approval to include how systems are used in practice, including clearer expectations around human oversight and decision-making responsibility.

Third, evaluation metrics must move beyond adoption rates to assess effective use and real-world impact, capturing how digital systems shape outcomes within complex care environments.

These shifts are not about slowing innovation. They are about ensuring that innovation translates into meaningful value.

Because ultimately, digital health is not an end in itself. It is a means to improve patient care and that improvement depends not only on whether systems are available, but on whether they are understood, trusted, and used appropriately.

In this sense, the future of digital health will be shaped less by the technologies we build, and more by the systems and capabilities we build around them.

For institutions and organisations operating in this space, this presents both a challenge and an opportunity. The challenge lies in moving beyond deployment toward integration. The opportunity lies in shaping how digital health systems are governed, evaluated, and embedded within real-world care environments.

Increasingly, this requires a more deliberate and structured approach one that brings together policy design, system-level analysis, and implementation insight.

This is where entities such as HealthTechAsia are increasingly focused: supporting governments, partners, and organisations in translating digital health ambition into operational reality, ensuring that access is matched with the capacity to engage, and ultimately, to deliver sustained value.

As the field continues to evolve, those able to navigate this transition from access to engagement will not only adopt digital health more effectively, but will also play a defining role in shaping its next phase.