The Ambulance Is Becoming an Interface

Estonia's plan to move toward mostly two-person ambulance crews sounds like a staffing reform. It is actually a redesign of the whole emergency-care system.

When one person disappears from the vehicle, documentation, communication, lifting, monitoring, medication preparation, hospital coordination, and scene management still need to happen. The work moves into software, dispatch protocols, specialist response units, hospital networks, and the hands of other responders.

The ambulance stops being a self-contained unit. It becomes the moving interface of a regional healthcare network.

Most people imagine an ambulance as a small emergency room on wheels. A call arrives, a crew drives to the patient, the crew provides treatment, and the patient is taken to a hospital.

The quality of the system appears to depend on three visible things: how quickly the ambulance arrives, how many people are inside it, and how quickly it reaches a hospital.

Modern emergency care is increasingly organized as a layered system. A basic crew may arrive first. A physician, specialist unit, fire crew, or helicopter may join later. The ambulance may transmit an ECG before moving. Dispatch may recommend a destination based on the patient's condition, specialist coverage, travel time, and hospital availability.

The important unit is no longer the ambulance. It is the network.

Estonia's emergency medical strategy proposes four stronger operational regions and a network designed to maintain or improve access across the country. The controversial part is the planned shift toward mostly two-person crews.

The weak version of the reform removes one person and expects the remaining two to do the same work. The stronger version distributes more crews geographically, improves dispatch, adds clinical support, uses telemedicine, and sends additional capability when a case requires it.

Evidence on crew size is mixed. Some simulation research has found little difference between two-, three-, and four-person crews for several resuscitation measures. Other research has found three-person teams faster at time-critical actions. A recent scoping review concluded that two-person advanced resuscitation can work, but often depends on adaptations such as rapid backup, mechanical compression equipment, prefilled medication, and protocols designed specifically around smaller teams.

Two people can be enough. But two people is not a complete operating model.

Taiwan offers another side of the same transformation. Its regular ambulances are commonly operated by fire departments with small firefighter-EMT crews, but the more interesting development is the regionalization of emergency care.

Taiwan introduced nationwide emergency-care regionalization and a three-tier categorization of hospital emergency capability. Hospitals are not treated as interchangeable destinations. They are nodes with different levels of capability for trauma, cardiac emergencies, stroke, and critical care.

This changes the question inside the ambulance. The old question is which hospital is closest. The newer question is which reachable hospital can provide the treatment this patient is likely to need.

For a major heart attack, a nearby hospital without a catheterization laboratory may be geographically closer but clinically farther away. The patient can arrive quickly, wait for assessment, wait for another hospital to accept the case, be loaded into another ambulance, and then travel again.

A slightly longer direct journey to a specialist center may produce a shorter time to definitive treatment.

A hospital may own the correct equipment but still be unable to help at that moment. The specialist may not be on site. The operating room may be occupied. The intensive-care unit may be full. The scanner may be unavailable. The emergency department may be overloaded.

A destination screen therefore cannot simply show hospitals sorted by driving time. It must distinguish theoretical capability from live capacity, while also showing the uncertainty behind the recommendation.

A useful recommendation explains why a hospital was selected, which required capability it has, how much additional travel time is involved, whether the hospital has accepted the patient, and what alternative remains if the assessment changes.

This is not a route-planning app with a medical skin. It is safety-critical decision support.

Finland is probably the most relevant comparison for Estonia. Finnish ambulances commonly use two-person crews, while field commanders, first-response units, advanced clinicians, and physician-led helicopter teams add capability when required.

Denmark follows a similar network logic. Dispatch centers assess calls, ambulances work alongside paramedic and physician units, and prehospital information can be transferred directly to the receiving hospital.

These systems do not prove that smaller crews are automatically safe. They show what must exist around them: visible backup, reliable medical consultation, shared patient information, destination flexibility, and a receiving hospital that understands what has already happened before the doors open.

The crew size is visible. The coordination architecture is not.

Healthcare digitization often promises to reduce workload. It can also create a second workload layered on top of the first.

A major UK study of electronic ambulance records found clinicians still writing observations on gloves, notebooks, or scraps before entering them into a device. The systems often worked mainly as stores for patient data rather than as useful tools for decisions, information transfer, or clinical handover.

A clinician can be treating a patient in a moving vehicle while the official digital interface is so poorly matched to the work that the glove becomes the real interface.

When an organization removes human capacity, it cannot compensate by adding forms. The interface must remove coordination work from the crew.

The future ambulance interface should behave less like an electronic clipboard and more like a shared operational picture.