Environmental footprint of inhalational paediatric anaesthesia: emission intensity is highest in smaller patients

Introduction

Healthcare systems are recognized as major contributors to global greenhouse gas emissions. ¹ In paediatric patients the use of inhalational anaesthesia, nitrous oxide (N₂O), high fresh gas flows per kilogram of body weight and high-flow induction techniques are frequently applied. A recent modelling study showed that intravenous techniques could reduce carbon dioxide equivalents (CO₂e) emissions compared with inhalational anaesthesia in children. ² The SPRUCE study showed that a combination of educational and technical modifications can provide “environmentally responsible anaesthesia” in a paediatric setting. ³ However, these studies were not based on a patient level and therefore did not provide granular estimates of median CO₂e emissions per case. The aim of this study was to describe in detail CO₂e emissions during paediatric inhalational anaesthesia, both as absolute values and after normalization per kg of body weight and per minute of anaesthesia. This information is necessary to provide a benchmark of CO₂e emissions in children of different ages.

Methods

This prospective observational cohort study was conducted from 3 November to 20 December 2025, at Meyer Children’s Hospital. Paediatric patients below 18 years of age undergoing general anaesthesia were included. Exclusion criteria were cases adopting total intravenous anaesthesia (TIVA) or combined techniques. The study was approved by the Hospital Ethics Committee (Comitato Etico Regionale per la Sperimentazione Clinica della Regione Toscana). The need for informed consent was waived due to the observational nature of the study. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were applied. ⁴ Extracted variables included demographics, American Society of Anesthesiology (ASA) physical status, routine or urgency scheduling, anaesthesia duration, fresh gas flow metrics, anaesthetic consumption, and ventilatory parameters. Clinical data were collected from each enrolled patient on the day of the scheduled surgery. Fresh gas delivery (average flow and total volume during each case), and sevoflurane consumption were extracted from the online Draeger’s Gas Consumption Analytics system (Draeger AG & Co. KGaA, Lubeck, Germany) that anonymously collects information produced by all the ventilators working in our operating theatre and makes them available for analysis.

The primary objective of this study was to describe the median total CO₂e emissions in the analysed period. Secondary objectives were to describe median CO₂e emissions per kg per minute of anaesthesia (CO₂e intensity). Sevoflurane-related CO₂-equivalent emissions were calculated using a density of 1.52 g/mL and a 100-year global warming potential (GWP100) of 130, corresponding to 0.198 kg CO₂e per mL of liquid sevoflurane.

Continuous variables are reported as medians (interquartile range). Group comparisons were performed using Mann–Whitney U or Kruskal–Wallis tests. Age classes were stratified into quartiles. Associations were assessed using the Spearman rank correlation coefficient r. Statistical significance was set at P <0.05 unadjusted. Statistical analysis was performed using the GraphPad Prism 10.0 software package (GraphPad Software, San Diego, CA, USA). Patients were enrolled consecutively as an exploratory sample for a period of 2 months. Artificial intelligence tools were not utilized to conceive any part of the manuscript.

Results

In the analysed period 814 children underwent general anaesthesia. Of these, 84 were conducted as TIVA and in 26 cases the gas analytics system did not report CO₂e data due to connection issues. Ultimately, 704 patients were included in the analysis. Demographic data, anaesthetic characteristics and surgical categories are reported in Table 1. Total cumulative CO2e was 18,903 kg. Total CO₂e emission per patient reached a median value of 21.6 (14–33.5) kg. A group of 339 patients underwent sevoflurane anaesthesia without N₂O, with a median CO₂e emission of 15.1 kg per case (9.4–21.4), whereas 365 children received sevoflurane combined with N₂O during induction, with a median CO₂e emission of 30.6 kg per case (21.1–43.5) (P < 0.001). The median value of average fresh gas flow rate for each case (FGF) was 6.10 (4.36–7.68) L/min; total gas volume was 481 (328–701) L. Total CO₂e was significantly correlated with total fresh gas volume (r = 0.36, P < 0.001). Total emitted CO₂e per minute of anaesthesia showed a significant correlation with average FGF (r = 0.469, P < 0.001). Longer procedures were weakly associated with higher CO2e production (r: 0.25, P = 0.001). Once CO2e emissions were normalized for body weight and per minute of anaesthesia, the association with duration became negative (r: −0,37, P < 0.0001). After stratification by age group (Table 2), median absolute CO₂e emissions per case appeared modestly lower in adolescents (P = 0.012). However, CO₂e normalized emissions showed a marked age gradient, progressively decreasing across older age classes (P < 0.0001). In patients younger than 1 year CO₂e intensity was approximately 12 times higher than in those aged 12 years or older. Average FGF differed significantly across age classes (P=0.0289), but at post-hoc pairwise testing only patients aged 1 to less than 6 years showed higher FGF compared with those aged 12 years or older (adjusted P = 0.025). Average FGF normalized according to patient weight (L/Kg/min) differed significantly across age classes (P < 0.0001), being higher at post-hoc pairwise testing in patients aged less than 1 year compared with all other age groups (Table 2).

Table 1.

Patient, anaesthetic, and ventilatory characteristics (n = 704).

Variable	Median (IQR) or n (%)
Age (years)	8.5 (4.3–12.4)
Weight (kg)	27.5 (17–47)
ASA physical status	ASA I: 396 (56.2%) ASA II: 194 (27.6%) ASA III: 103 (14.6%) ASA IV: 11 (1.6%)
Urgent procedure	195 (27.7%)
Anaesthesia duration (min)	82 (54–126)
Mechanical ventilation (L/min)	3.6 (2.4–5.0)
Tidal volume (mL)	160 (110–250)
Respiratory rate (breaths/min)	18 (14–24)
Surgical category	n (%)
General surgery	245 (34.7)
Orthopaedics	146 (20.7)
Ear nose and throat	102 (14.4)
Neurosurgery Plastic surgery	72 (10.2) 55 (7.8)
Medical ward	48 (6.8)
Ophthalmology	15 (2.1)
Dentistry	11 (1.6)
Maxillofacial surgery	7 (1.0)
Urology	5 (0.7)

Table 2.

Analysis of population, fresh gas delivery and CO₂e emissions stratified by age classes.

Variable	<1 year	1⩽6 years	6⩽12 years	⩾12 years	P value
n	59	185	262	198
Weight (kg)	6.50 (3.80–8.90)	15 (13–18)	31.1 (23.5–40)	54 (45–62)	<0.0001
Anaesthesia duration (min)	116.00 (69.5–182.5)	72.00 (55–107)	76 (52.2–114)	95 (66–144.7)	<0.0001
Average fresh gas flow (L/min)	5.86 (4.34–7.62)	6.41 (4.42–7.93)	5.64 (4.42–7.18)	5.49 (4.24–6.68)	0.029
Average fresh gas flow/kg (L/kg·min)	1.02 (0.59–1.65)	0.42 (0.27–0.54)	0.18 (0.13–0.26)	0.10 (0.07–0.13)	<0.0001
Total CO2 produced (kg)	21.5 (15.9–31.3)	21.6 (15.8–34.6)	23.0 (14.4–35.0)	18.9 (10.3–32.0)	0.0128
CO2/kg/min (g/(kg·min))	35 (20–56)	18 (11–30)	9.6 (5.2–16.3)	3 (1.8–6.9)	<0.0001
CO2e per L of fresh gas (g/L)	33.5 (26.7–38.7)	49.1 (32.8–75.6)	52.7 (33.8–80.3)	31.0 (23.1–53.1)	<0.0001
Sevoflurane CO2e per case	19.10 (13.7–25.9)	13.5 (9.0–19.3)	12.8 (8.8–16.9)	12.8 (8.4–18.8)	<0.0001
Nitrous oxide CO2e per case	0.00 (0.00–2.39)	6.73 (0.00–15.6)	8.75 (0.0–20.3)	0.00 (0.0–11.9)	<0.0001
CO2e without nitrous oxide (kg)	19.8 (13.98–27.54)	16.9 (12.22–22.92)	13.6 (9.05–19.60)	13.2 (7.77–19.60)	<0.0001
CO2e with nitrous oxide (kg)	49.0 (35.7–60.5)	28.5 (18.7–38.7)	30.2 (21.6–40.8)	34.1 (23.9–48.5)	0.0008

Continuous variables are reported as medians (interquartile range).

Global P values were calculated using the Kruskal–Wallis test.

Discussion

The key findings of our study are: first, in our centre the environmental burden of paediatric volatile anaesthesia was substantial even for single cases, ranging from 10 to 40 kg of CO₂e due to the use of rather high levels of fresh gas volumes and the application in half of cases of N₂O at induction. Average FGF is a condensed measure of total fresh gas per each minute of anaesthesia; accordingly, it was associated with CO2e emissions per minute. Second, while anaesthesia duration showed a weak association with absolute per-case emissions, younger children undergoing shorter procedures incurred a higher environmental burden, as shown by the inverse relationship between duration and CO2e emissions normalized for body weight and per minute of anaesthesia. We estimate that this happened because with high induction flows (i.e. 10–15 L/min)), irrespective of patient age and size, the emission intensity was higher during the shorter procedure in the smallest children because the induction-related fresh gas volume represented a larger proportion of total fresh gas exposure. Third, as shown by average FGF values stratified by age class, no consistent reduction in FGF was observed in younger paediatric age groups throughout the procedure, despite their smaller body size and lower ventilatory requirements. However, FGF normalized according to patient weight was significantly higher in patients aged less than 1 year compared with all other age groups. Finally, in absolute terms, anaesthesia in children produced broadly similar amounts of CO₂e across age classes. However, patients younger than 12 years were not only the most frequent, but also the least efficient, as in these age groups CO₂e emissions normalized per kilogram of body weight and per minute of anaesthesia were strikingly higher.

While the benefits of low-flow anaesthesia and N₂O restriction are intuitive, their real-world impact relies heavily on implementation strategies. Further reductions in fresh gas flows may be achievable with modern anaesthesia workstations, not available in our centre, which support the safe use of lower FGF rates through electronically controlled vapor delivery. Furthermore, there is considerable potential in systematically minimizing fresh gas flow and avoiding routine N₂O use at induction and in short procedures, where the impact on emission intensity is greatest. This is possible in all children, with anaesthesia circle breathing systems, and some attention to mask fit at induction and to reducing airway devices leaking during maintenance phase anaesthesia. Educational and behavioural interventions should be specifically targeted at the highest-carbon-intensity subgroup, namely patents below 12 years, who also represent the most clinically challenging to manage in daily paediatric anaesthesia practice. Based on these findings, FGF management practices have been reviewed and addressed in our centre specifically to emphasize the need to adapt fresh gas delivery to patients’ body size and age during both induction and maintenance.

This study has several limitations. First, its observational design precludes causal inference. Second, this analysis, conducted at a single tertiary paediatric centre, may have limited generalizability compared with institutions with different case mixes, equipment, or anaesthetic practices. Third, even if sevoflurane consumption was derived from a last generation gas analytics system, previous reports have highlighted potential inaccuracies associated with this methodology. ⁵

Conclusion

In conclusion, we showed substantial CO₂e emissions associated with inhalational anaesthesia in our centre. Although total emissions were largely driven by fresh gas delivery, N₂O use, and procedure duration, emission intensity was highest in younger patients, even during shorter procedures, reflecting disproportionately high average FGF relative to body weight. These findings highlight a substantial and readily actionable margin for improvement by promoting N₂O-free anaesthesia and tailored FGF management.

Data availability statement

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.