Telemedicine cuts carbon footprint by up to 99%, yet AI and wearables pose new climate challenges

A comprehensive new review has placed the spotlight on how digital technologies are reshaping the climate footprint of healthcare. The study examines the potential of healthcare digitalization to both reduce and increase greenhouse gas emissions depending on its application and management.

The paper, titled “The Climate Impacts of Healthcare Digitalization: A Scoping Review” and published in Digital Health, systematically analyzed 32 studies covering telemedicine, diagnostics, artificial intelligence, health informatics, medical appliances, robotics, and self-tracking technologies. The findings show that while digitalization offers significant opportunities to cut emissions, particularly through telemedicine and virtual education, it also poses risks when technologies are poorly managed or when energy-intensive systems, such as large-scale AI models, are not optimized.

Can telemedicine and virtual care cut healthcare emissions?

Telemedicine substantially reduces carbon emissions compared with traditional in-person visits. The carbon footprint of a virtual consultation was estimated between 0.005 and 3 kgCO₂e, in stark contrast to 0.57 to 178 kgCO₂e for physical visits, largely dependent on travel mode. This translates into a reduction of between 79 and 99 percent.

The review also highlights similar trends in medical education. Virtual continuous medical education was shown to have a carbon footprint reduction of nearly 99.7 percent compared with in-person events, with travel avoidance being the main contributor. Telepsychiatry and hybrid telehealth systems were also found to cut emissions, although the extent varied depending on whether analyses considered electricity use, equipment production, or data transmission.

The review found emerging examples of digital interventions with climate benefits, such as autonomous AI reducing facility needs, machine learning supporting sustainable pharmaceutical 3D printing, and virtual health services lowering the demand for physical infrastructure. However, the authors caution that methodological inconsistencies, such as narrow system boundaries and low transparency in carbon accounting, often weaken the reliability of the evidence.

Where do AI, wearables, and diagnostics fit into the carbon equation?

The review pays close attention to artificial intelligence, identifying it as a double-edged sword for healthcare’s climate footprint. AI can optimize processes, identify overdiagnosis, and streamline care, all of which have the potential to cut emissions. Yet, training large machine learning models like GPT-4 or multimodal medical systems requires immense computing power, with training alone producing thousands of tons of CO₂ emissions. The authors emphasize that smaller, optimized models may achieve similar medical outcomes with drastically lower environmental costs.

Self-tracking technologies, such as wearable devices, also demonstrate mixed impacts. On one hand, they can prevent hospitalizations by enabling early detection and monitoring of chronic conditions, thus lowering the carbon footprint associated with hospital care. On the other, the mass production, energy use, and eventual disposal of millions of wearable devices create new environmental burdens. The rapidly growing wearables market, projected to exceed $180 billion by 2030, highlights the urgency of addressing this tension.

In diagnostics, point-of-care testing emerged as a low-carbon alternative compared with laboratory testing, reducing waste and travel requirements. However, the review raises concerns about overdiagnosis, excessive use of diagnostic tools that lead to unnecessary treatments, travel, and resource use. Studies estimate that up to 40 percent of clinical care may be low-value or harmful, with significant carbon implications. For example, unnecessary vitamin D testing in Australia alone was found to account for tens of tons of CO₂ emissions annually.

What role do leadership and strategy play in steering low-carbon digital health?

Perhaps the most striking gap identified in the review is the absence of leadership and strategic planning in the studies reviewed. While evidence points to clear opportunities for emissions reductions, few studies addressed the role of hospital or healthcare leadership in aligning digitalization with climate goals.

The authors argue that without strategic direction, digitalization risks reinforcing high-carbon practices instead of transforming them. Investments in ICT infrastructure, medical appliances, and robotics need to be guided by policies that explicitly prioritize low-carbon solutions. Similarly, the adoption of AI, telehealth, and electronic records requires transparent accounting of emissions across their full life cycle.

The review introduces a framework that organizes digitalization’s role in healthcare into action, foreground, and background layers. “Action” includes visible patient services like telehealth, diagnostics, and treatments. “Foreground” encompasses supporting equipment like medical appliances and robotics, while “background” covers unseen but crucial systems such as health informatics, AI infrastructure, and logistics. Leadership and planning are identified as the cross-cutting forces needed to ensure that all these layers contribute to emissions reduction rather than escalation.

Innovations, such as smart logistics systems and renewable-powered data centers, should be integrated into healthcare’s digital transformation strategies. By embedding sustainability into procurement, investment, and innovation, healthcare systems can pursue a credible pathway toward climate neutrality.