Using An Artificial Intelligence Telephone Call To Reduce The Carbon Footprint In The Cataract Pathway

In 2020, the NHS’s carbon footprint was 24.9 million tonnes carbon dioxide equivalents (CO2e). To achieve net zero by 2045 while meeting rising healthcare demand, service delivery must evolve. Cataract surgery, the most common surgical procedure, saw over 450,000 surgeries in 2019/20, with demand continually increasing. Artificial intelligence-driven tools like Dora (Ufonia Ltd, Oxford, UK) which automate clinical assessments over the telephone replacing both clinician-led in-person and telephone assessments have the potential to lower carbon emissions whilst also enhancing clinical capacity. 

The change in carbon emissions was assessed across four NHS sites where Dora has been implemented as the standard of care to conduct consultations at four time points in the cataract care pathway. Prior to the implementation of Dora the pre-surgery assessments and post-surgery follow-ups were face-to-face (F2F) for three sites and telephone based at one. Patients had a pre-surgery reminder over the telephone and patient reported outcome collection was modelled compared to paper based surveys.

A hybrid carbon footprint analysis compared emissions pre and post implementation of Dora. Using the greenhouse gas (GHG) emissions associated with patient travel, staff commuting, energy use, some appointment consumables and Dora’s impact, a process-based analysis was undertaken. Data sources included patient surveys, NHS Estates Returns and Greener NHS data. For Dora’s impact, smartphone energy use, text reminders, router, wifi network, server and laptop hardware was considered. Due to lack of available data for some of the consumables and pharmaceuticals used, an environmentally extended input output analysis was undertaken where financial expenditure was converted to CO₂e using conversion factors from the UK Government SIC database. 

At baseline, the carbon footprint per pre-assessment and post-op face-to-face (F2F) review was 12.78 kg and 12.24 kg CO₂e, respectively. The primary contributor was patient travel (61%–64%), followed by consumables and pharmaceuticals (18%), overhead energy use (11%–12%), and staff commuting (8%–9%). A human-led surgery reminder call emitted 0.018 kg CO₂e.

Dora’s emissions depended on call length, averaging 0.1261 kg CO₂e per call, with server usage accounting for 57%–63% of emissions. Across a full patient pathway, four Dora calls totaled 0.5044 kg CO₂e, reducing to 0.2519 kg CO₂e with higher patient numbers and 90% server utilisation. Across four sites, Dora saved 129.31 kg CO₂e per 100 cataract surgeries.

Integrating AI-driven automation into the cataract care pathway has demonstrated a measurable reduction in carbon emissions. Dora’s implementation across four NHS sites significantly lowered greenhouse gas emissions, with potential annual savings of 16,037.9 kgCO2e for 14,500 surgeries—equivalent to eliminating the emissions from 27 return flights between London and Munich. Expanding Dora’s adoption across more providers could further enhance efficiency, maximise carbon savings, and support the NHS’s journey toward net-zero emissions.