The second-stage theatre argument
Decommissioning desflurane and repairing leaky nitrous oxide manifolds are the easy lines in the operating theatre carbon ledger. Nobody defends a leaky pipe, and very few elective lists clinically require a volatile gas with 2,640 times the warming potential of carbon dioxide. The real friction starts on the Monday morning after desflurane is gone. Half the department argues that modern low-flow or target-controlled sevoflurane brings volatile anaesthesia close enough to propofol TIVA; the other half worries that switching to TIVA trades atmospheric carbon for single-use plastic syringes and propofol water toxicity.
Between 2024 and 2026, three large European datasets replaced that debate with measured procedure-level numbers. Together, they audit more than 82,000 real general anaesthetics across both carbon dioxide equivalents (CO₂e) and 10-year freshwater ecotoxicity.
1. Carbon: automated low-flow trims sevoflurane, but by 18 percent, not 80 percent
In the 2025 British Journal of Anaesthesia multicentre observational study (DOI 10.1016/j.bja.2025.01.043, PMID 40187906), investigators analysed 35,242 general anaesthesia procedures across three French university hospitals, each championing a distinct routine technique: 10,717 propofol TIVA cases, 7,873 target-controlled inhalation anaesthesia (TCIA) sevoflurane cases using automated end-tidal control, and 15,461 manually optimised low-flow sevoflurane cases. Crucially, the life-cycle inventory counted drug manufacturing, single-use plastic syringes, infusion lines, breathing circuits, soda lime, and medical device electricity.
Even with every plastic syringe and infusion line charged against the intravenous arm, propofol TIVA averaged 0.4 kg CO₂e per hour. Automated TCIA sevoflurane averaged 3.1 kg CO₂e per hour (7.75 times higher than TIVA), and manually adjusted low-flow sevoflurane averaged 3.8 kg CO₂e per hour (9.5 times higher than TIVA). Automated end-tidal control is worth enabling on every workstation that still uses volatile agents, as it trimmed hourly sevoflurane emissions by about 18 percent over manual low-flow. What it did not do was close the gap to propofol.
At whole-hospital scale, a 2024 Anaesthesia evaluation of 47,157 adult patients across two acute hospitals (DOI 10.1111/anae.16221, PMID 38205529) confirmed how large that gap becomes when nitrous oxide and volatile vapours are retired together: 25,137 patients anaesthetised at the TIVA-only hospital averaged 2.42 kg CO₂e per procedure, compared with 48.85 kg CO₂e per procedure across 22,020 patients at the mixed volatile-TIVA hospital, a 20-fold per-case reduction.
2. Water and PFAS: the trifluoroacetic acid (TFA) blind spot
The most common objection to expanding propofol TIVA is aquatic ecotoxicity: propofol is toxic to aquatic organisms and has a persistence-bioaccumulation-toxicity (PBT) hazard score that rightly forces theatres to dispose of leftover syringes in high-temperature clinical waste incineration bins rather than pouring them into sinks. Volatile sevoflurane, by contrast, is exhaled by the patient and vented off the hospital roof, where it disappears from the pharmacy waste audit.
In 2026, a follow-up British Journal of Anaesthesia modelling study across 33,385 procedures (DOI 10.1016/j.bja.2026.03.042, PMID 42097956; accompanied by July 2026 commentaries DOI 10.1016/j.bja.2026.07.060 and DOI 10.1016/j.bja.2026.07.069) tracked where that vented sevoflurane actually goes. In the atmosphere, sevoflurane undergoes photolytic degradation into trifluoroacetic acid (TFA), an ultrashort-chain per- and polyfluoroalkyl substance (PFAS) that washes back to earth in rainwater. Because TFA has no known environmental sink or natural biodegradation pathway, every hour of sevoflurane anaesthesia adds permanently to freshwater TFA concentrations.
Simulating 1,000 anaesthetic hours per year over a 10-year horizon (using DataEnforce inventory parameters across 10,717 propofol TIVA, 7,207 TCIA sevoflurane, and 15,461 manual sevoflurane procedures), the authors found two completely different trajectories:
- Sevoflurane-derived TFA accumulated linearly and irreversibly, reaching a 10-year water contamination potential of 29 million m³ for automated TCIA sevoflurane and 35 million m³ for manual low-flow sevoflurane.
- Propofol degraded locally in aquatic models and reached an equilibrium plateau after roughly 5 years at 2.6 to 2.9 million m³, a 10-fold lower 10-year water footprint than sevoflurane.
3. The department ledger: comparing the three regimens
For clinical leads updating their operating theatre action plan or grading local formulary changes under our evidence methodology, the combined 2024 to 2026 data can be summarised in a single audit table:
| Regimen & cohort | Hourly carbon | 10-yr water footprint (1,000 h/yr) | Primary environmental driver & audit rule |
|---|---|---|---|
| Propofol TIVA n = 10,717 (BJA 2025/2026) n = 25,137 (Anaesthesia 2024) |
0.4 kg CO₂e/h 2.42 kg / case |
2.6 to 2.9M m³ Plateaus yr 5 |
Plastic syringes, tubing & electricity dominate CO₂e. Mandatory guardrail: never pour residual propofol into sinks; dispose in clinical incineration bins and size vials (20 mL vs 50 mL) to case length. |
| Automated TCIA sevoflurane n = 7,873 (BJA 2025) n = 7,207 (BJA 2026) |
3.1 kg CO₂e/h 7.8x TIVA |
29.0M m³ Linear TFA rise |
Exhaled sevoflurane gas drives >85% of CO₂e and degrades into atmospheric TFA (persistent PFAS). Use automated end-tidal control whenever volatile anaesthesia is clinically indicated. |
| Manual low-flow sevoflurane n = 15,461 (BJA 2025/2026) |
3.8 kg CO₂e/h 9.5x TIVA |
35.0M m³ Linear TFA rise |
Higher fresh-gas flow spikes during induction and maintenance add 0.7 kg CO₂e/h and 6M m³ more 10-year TFA water impact over automated TCIA. |