Comparing the carbon footprint of telemedicine vs in-person visits across a patient population
Before any digital substitution occurs, every outpatient encounter carries a travel footprint: a car trip from home to clinic and back. Lifecycle carbon accounting of healthcare delivery consistently finds that this Scope 3 patient-travel component, not the clinical encounter itself, dominates the emissions of a typical visit. Establishing this baseline is the necessary first step before any telemedicine substitution can be credited with avoided emissions.
A full lifecycle accounting of an outpatient visit includes the clinical encounter (lighting, HVAC, sterilization, equipment), the supply chain behind consumables and pharmaceuticals, and the patient's own travel to and from the point of care. When researchers itemize these components, patient and staff travel routinely emerges as the single largest line item — frequently 40-70% of a visit's total footprint for anything beyond the shortest local trips.
This is why the earliest peer-reviewed telemedicine carbon studies, such as Holmner et al.'s 2014 analysis of Swedish telehealth consultations, focused specifically on avoided travel distance as the primary carbon lever, rather than on differences in the clinical encounter itself. The clinical content of a visit is nearly carbon-neutral by comparison; the mode of arrival is what matters.
A single 15-mile-each-way clinic visit by car emits roughly 12 kg of CO2 from the round trip alone — comparable to the emissions of driving a passenger car for a full day of ordinary errands.
Outpatient visit volume has grown steadily for decades, and health systems have consolidated specialty services into fewer, larger regional centers — improving clinical efficiency but lengthening the average trip a patient must take. Rural and exurban patients disproportionately bear this cost, often driving 30-60+ miles one way to reach a specialist, while urban patients may face shorter distances but heavier traffic and correspondingly higher idling emissions.
Because this baseline is built almost entirely from car travel, it is also the most tractable target for carbon reduction: unlike the physics of running an MRI or heating a hospital wing, a car trip can, for a meaningful share of visit types, simply be avoided.
Telehealth utilization expanded roughly 38-fold during the early COVID-19 pandemic relative to pre-2020 baselines, and while volumes have settled well below that peak, adoption has remained structurally higher than before. Each visit that shifts from a car trip to a video or phone connection avoids essentially its entire travel-related footprint, replacing it with a comparatively tiny digital one. The carbon benefit scales directly with how many visits are clinically appropriate to substitute.
Not every encounter can move online, but a substantial share can: routine medication management, chronic disease follow-up (hypertension, diabetes check-ins), mental health therapy, post-operative check-ins, and dermatology triage via photos are all well-established telehealth use cases with strong clinical evidence behind them.
Visits that still generally require physical presence include anything needing hands-on examination, vital-sign instrumentation not available at home, imaging or lab draws, procedures, and a patient's first complex diagnostic workup where physical findings matter. The carbon reduction opportunity is therefore bounded by clinical appropriateness, not simply by technology adoption — pushing every visit online would be both clinically unsound and unnecessary to capture most of the achievable savings.
When a visit is substituted, the patient's round-trip car emissions (distance × 2 × 0.40 kg CO2/mile) are avoided in full, replaced by a small digital footprint. At a 15-mile average one-way distance, that is roughly 12 kg of avoided tailpipe CO2 per substituted visit — before accounting for the modest energy cost of the video connection itself, covered in the next stage.
Because avoided emissions accrue per visit and scale linearly with both adoption rate and average travel distance, health systems serving geographically dispersed or rural populations see disproportionately large carbon benefits from even modest increases in telehealth adoption.
Every 10-percentage-point increase in telehealth adoption across a 50,000-visit annual population avoids on the order of 500,000+ miles of car travel at typical suburban travel distances.
Telemedicine is not a zero-carbon activity — video calls draw on data centers, network infrastructure, and end-user devices that all consume electricity and embody manufacturing emissions. A rigorous carbon comparison has to account for this digital footprint honestly rather than treating telehealth as automatically free. The evidence nonetheless shows this footprint is an order of magnitude (or more) smaller than the travel it replaces for any visit beyond a very short local trip.
A video visit's footprint has three components: the electricity used to encode, transmit, and route video data through the network and data center; the electricity the patient's and clinician's devices draw during the call; and an amortized slice of the embodied carbon from manufacturing those devices, spread across their multi-year lifespan.
Of these, data-center and network energy is the most variable — it depends heavily on local grid carbon intensity, video resolution, and call duration. Device manufacturing is a fixed, one-time cost the visit merely borrows a fraction of. None of these components scale with the distance between patient and clinic, which is precisely why telehealth's footprint stays flat regardless of how far apart the two parties are — unlike a car trip, where footprint rises linearly with distance.
Even using the higher end of published videoconferencing estimates (roughly 1 kg CO2e per hour of high-definition video on a carbon-intensive grid), a typical 15-20 minute telehealth visit totals well under 0.3 kg CO2e — compared to roughly 12 kg CO2e for a 15-mile round-trip car visit. The gap widens further for rural patients traveling 30-60+ miles, and narrows only for the shortest local trips or for patients using low-carbon transit alternatives.
Grid decarbonization also compounds the advantage over time: as electricity grids add more renewable generation, the digital footprint of telehealth falls automatically, while the physics of tailpipe combustion for a gasoline car trip does not change.
Comparing like-for-like: a 15-mile round trip by car emits roughly 40-200x more CO2 than the video connection that could have replaced it.
Comparing emissions across visit types clarifies an important point: it is travel distance, not clinical content, that mainly drives the carbon gap between in-person and telemedicine care. A short local follow-up and a long-distance specialist consult may involve similar clinical time, but their footprints diverge enormously once travel is factored in — while their telehealth equivalents remain roughly constant regardless of visit type.
Routine follow-ups tend to have the smallest in-person footprint because they are typically scheduled at a nearby primary care site. Mental health therapy, despite often being clinically similar in intensity to a routine visit, carries an outsized annual footprint because it recurs weekly or biweekly — the same round trip repeated dozens of times a year compounds quickly.
Specialist consults carry the largest single-visit footprint because referral networks frequently route patients to regional centers well outside their local area. In every case, the telehealth alternative stays essentially flat: video calls do not get more carbon-intensive because the specialist is farther away.
Physical examination, in-person diagnostic imaging, lab draws, and procedures cannot be replaced by video regardless of the carbon incentive to do so — and safety must always take priority over emissions accounting. The realistic model for most health systems is a hybrid one: routine and recurring low-acuity visits shift to telehealth by default, while diagnostic, procedural, and first-time complex visits remain in person.
Modeled this way, even conservative substitution rates concentrated on recurring visit types (like mental health follow-up) produce large annual carbon reductions, because those are exactly the visits that repeat most often across a patient population.
A single patient attending 12 monthly in-person mental health visits at 15 miles one-way generates roughly 145 kg of CO2e per year from travel alone — nearly all of it avoidable through telehealth substitution.
Aggregated across an entire patient population, avoided car trips compound into carbon savings measured in thousands of tons annually for a mid-size health system — while simultaneously reducing the travel, time, and cost burden that falls hardest on rural, mobility-limited, and low-income patients. Carbon reduction and health equity turn out to be closely linked co-benefits of the same underlying shift.
Rural patients, patients with disabilities affecting mobility, caregivers without backup childcare, and low-wage workers who lose pay for missed shifts all carry a disproportionate share of the travel burden that telehealth removes. For these groups, avoided car trips translate directly into avoided lost wages, avoided caregiving logistics, and improved ability to keep recurring appointments — which is a major reason no-show rates tend to fall meaningfully when telehealth is offered as an option.
This equity dimension does carry an important caveat: telehealth introduces its own access barrier in the form of reliable broadband and a suitable device, meaning digital-divide gaps can offset some of the equity gains unless addressed deliberately through device lending programs, phone-based visit options, and broadband investment.
Telehealth's carbon and equity benefits are strongest for exactly the patients who travel farthest and can least afford to — rural, mobility-limited, and low-income populations.
Health systems modeling their own carbon savings should watch for rebound effects: telehealth can lower the barrier to seeking care, generating additional visits that would not otherwise have occurred (a genuine clinical positive, but one that partially offsets per-visit carbon savings at the system level). Accurate accounting nets true avoided trips against any induced-demand increase in total visit volume.
The most defensible model treats telehealth as one lever within a broader hybrid-care carbon strategy — alongside site consolidation planning, active-transport incentives for short local trips, and clinical protocols that default to telehealth wherever safe and appropriate — rather than a single silver-bullet solution.