The Impact of Anaesthesia Method on the Use of Anaesthesia Department Resources in Total Knee Arthroplasty

ABSTRACT

Background

Spinal and general anaesthesia seem to lead to similar health‐related outcomes in total knee arthroplasty. However, to produce economically and environmentally sustainable health care, we should also consider other factors. Current guidelines recommend preferring regional anaesthesia and total intravenous anaesthesia to reduce environmental impact. In this study, we analysed time, costs, and disposable materials associated with spinal anaesthesia and total intravenous anaesthesia during total knee arthroplasty.

Methods

A total of 395 patients undergoing total knee arthroplasty were randomised to spinal or general anaesthesia. Outcomes were total operating room time, non‐operative operating room time, post‐anaesthesia care unit time, use of drugs affecting haemodynamics, and the need for arterial cannulation and urinary catheterisation. In addition, we evaluated the financial costs and waste produced by single‐use equipment and medicines in both anaesthesia methods.

Results

Anaesthesia method did not affect operating room time (median 152 min [interquartile range 136–168 min] vs. 150 min [137–165 min], Hodges–Lehmann estimate 2 min [95% CI –3 to 6 min], p = 0.49). However, after spinal anaesthesia, patients appeared to spend more time in the postanaesthesia care unit than after general anaesthesia (median 104 min [71–148 min] vs. 73 min [49–98 min], Hodges–Lehmann estimate 32 min [95% CI 23–42 min], p < 0.001). Medication for hypotension was more common in the general anaesthesia group, whereas medication for hypertension was more common in the spinal anaesthesia group. General anaesthesia generated more disposable material (1252 vs. 611 g), and both medicine and equipment expenses were higher (total costs €83 vs. €36).

Conclusion

With this anaesthesia regimen, knee arthroplasty patients occupied the operating room for a similar duration regardless of anaesthesia method, but patients required longer recovery in the postanaesthesia care unit after spinal anaesthesia. Regarding disposable material and medicine use, spinal anaesthesia appeared to be more economical and material‐sparing than total intravenous anaesthesia. Furthermore, spinal anaesthesia appeared to cause less hypotension requiring haemodynamic interventions than general anaesthesia.

Editorial Comment

This health‐economic analysis estimates for how different aspects of general or spinal anaesthesia care contribute to process flow and item costs. Findings suggest that there may be measurably more costs associated with one of the anaesthesia type choices compared with the other, though this will depend on choices for materials and more.

Trial Registration

EudraCT (2016‐002035‐15)

1. Introduction

Clinicians aim primarily for the best results for their patients [1]. Thus, when studying and assessing different treatment methods, health care professionals tend to be most interested in health‐related outcomes such as morbidity, survival, and quality of life. However, responsible decision‐making requires that we take existing resources into account. Hospitals have limited funds to provide health care services for people in need, and we often must compromise between quality and costs. Shortfalls in the health workforce require time‐efficient options. It is ethical to treat patients according to our best knowledge and the most recent techniques, but public health care is also responsible for offering equal care and allocating resources responsibly.

Because climate change has been identified as a fundamental threat to human health [2], the environmental effects of health care systems cannot be ignored. Health care systems produce, on average, over 5% of national carbon footprints worldwide, but there is a 10‐fold difference in health care carbon footprints per capita between some countries [3]. It seems paradoxical to contribute to a global health‐threatening crisis by treating individuals' medical conditions. When examining health care's environmental impacts in detail, some factors play a significantly bigger role than others.

In hospitals, operating rooms (ORs) consume a significant amount of energy and produce notable emissions [4]. According to life cycle assessments, the majority of emissions in ORs arise from anaesthetics, single‐use items, and electricity use [4, 5]. As an anaesthetic, propofol seems to produce multiple times lower carbon dioxide equivalent emissions than volatile anaesthetics [6, 7]. Multiple guidelines recommend favouring total intravenous anaesthesia and regional anaesthesia to provide more environmentally sustainable care [8, 9, 10]. A commonly used volatile anaesthetic, desflurane, has become a well‐known contributor to global warming [11, 12], which has already led to restrictive actions in the European Union [13]. In addition, ORs generate over 20% of a hospital's plastic and total waste [14]. Due to contamination and infectious risk, disposal of this waste is more expensive, and recycling is problematic [14]. Plastics accumulating in landfills and nature are a major environmental burden [15]. Choosing reusable equipment instead of single‐use items generally lowers both emissions and financial costs [5, 16, 17].

Total knee arthroplasty (TKA) is a common surgical procedure for the treatment of knee joint arthritis. Both spinal (SA) and general anaesthesia (GA) are suitable anaesthesia methods for TKA, and it appears that anaesthesia method does not affect pain, adverse events, or surgical results after TKA [18, 19, 20]. The aim of this study was to compare time use, additional interventions, and disposable equipment between GA and SA in TKA, and to provide additional information for responsible decision‐making when anaesthesia methods are equivalent from a health‐related perspective.

2. Methods

This study is a prespecified secondary analysis of a randomised controlled trial; the study protocol, including the primary and secondary outcomes and sample size calculations, has been published previously [21]. The original trial was designed to investigate whether anaesthesia method or tourniquet use affected opioid consumption during the first 24 h after TKA.

The study design was accepted by Helsinki University Hospital's Ethics Committee (8 June 2016, ref.: HUS1703/2016) and the Finnish Medicines Agency Fimea (20 May 2016, ref.: KL72/2016). The study was registered with EudraCT (12 May 2016, ref.: 2016‐002035‐15).

The study was performed at HUS Helsinki University Hospital (Helsinki Metropolitan Area, Finland), a large academic medical centre serving a population of over two million people, with 1500 TKA surgeries performed annually. Adult patients waiting for TKA as treatment for Kellgren–Lawrence Grade 3–4 knee arthritis were recruited after providing informed consent. Exclusion criteria were age > 75 years, body mass index ≥ 40 kg m ⁻², ASA class ≥ 4, severe anatomical abnormality in the operating knee, contraindications to either anaesthesia method, or bilateral surgery. Randomisation to either SA with sedation or GA was performed on the day of surgery. According to the study protocol, we aimed to maintain normotension in both groups. Hypotension was treated with intravenous boluses of phenylephrine or etilefrine and/or continuous noradrenaline infusion, and hypertension with intravenous labetalol boluses and/or clonidine infusion. Neither patients nor staff were blinded. We collected variables such as procedural time labels, drugs used, and interventions from the electronic medical record and study forms.

The predefined primary outcome of this secondary analysis was OR time, defined as the period during which the patient occupied the OR before transfer to the post‐anaesthesia care unit (PACU). This included patient arrival in the OR, checklist completion, anaesthesia preparation and induction, surgery, application of wound dressings and elastic bandages, awakening and extubation (GA group) and transfer of the patient to a hospital bed before leaving the OR. Secondary outcomes were non‐operative OR time (OR time before surgical incision and after wound closure), PACU time (from arrival in the PACU until the patient met PACU discharge criteria), additional interventions, associated costs and waste produced.

The sample size for the trial was calculated using a two‐tailed test with an α level of 0.05 and 80% power. The sample size was inflated by 16% to accommodate possible non‐parametric analyses. For sample size analysis, we considered a 5% change in OR time to be clinically significant. Based on a sample from our local electronic medical records, the estimated average OR time was 161.5 min (standard deviation 14.2 min). A minimum of 60 patients per group was required for between‐group comparison.

The normality of continuous data was assessed using the Kolmogorov–Smirnov test. Comparisons between the two groups were conducted using the Mann–Whitney U test, and Hodges–Lehmann estimators with 95% confidence intervals (CIs) were calculated. The continuous data are presented as medians and interquartile ranges (IQRs). Categorical data were compared using the chi‐squared test or Fisher's exact test as appropriate and are presented as frequencies and percentages. In addition, as preplanned, adjusted tests were conducted for two‐group comparisons because of the original four‐group design, which included anaesthesia method with and without tourniquet use. Continuous data were examined using a linear model and results were expressed as adjusted mean differences and 95% CIs, while binary outcomes were compared using binary logistic regression and results were expressed as odds ratios and 95% CIs. All tests were two‐tailed, and p‐values less than 0.05 were considered statistically significant.

To evaluate consumption of single‐use equipment and waste production associated with anaesthesia, we collected, weighed, and tabulated items used in the anaesthesia protocols and any additional perioperative procedures when necessary. We collected and calculated the prices of the single‐use equipment and medicines used. The weight and price of single‐use items for additional perioperative procedures are expressed as the mean weight and cost per patient (i.e., the total weight and cost for the actual number of procedures divided by the total number of patients in each anaesthesia method group). Equipment prices were obtained from Valtti‐service (HUS Logistics, HUS Group, Helsinki, Finland) and drug prices from OSTi‐service (HUS Pharmacy, HUS Group, Helsinki, Finland). Organisational contract prices for individual products and medicines are confidential. We excluded oxygen and electricity consumption because data were not available, as well as non‐anaesthesia‐related equipment at the surgical site, such as sterile surgical drapes, maintenance for reusable instruments and implant packaging.

Data were analysed using the latest version of IBM SPSS Statistics (IBM Corp., Armonk, NY). Microsoft 365 Copilot (Redmond, WA) was used to edit the spelling and grammar of the final manuscript.

3. Results

A total of 395 patients undergoing TKA were included in the study between October 2016 and December 2018. Of these, 200 patients were randomised to the SA group and 195 to the GA group (CONSORT flowchart; Figure 1. For more detailed information about excluded patients, see [18]). The results are shown in Table 1. Neither total OR time (SA median 152 min [IQR 136–168 min] vs. GA 150 min [IQR 137–165 min], Hodges–Lehmann estimate 2 min [95% CI −3 to 6 min], p = 0.49) nor non‐operative OR time (SA median 65 min [IQR 56–73 min] vs. GA 63 min [IQR 55–73 min], Hodges–Lehmann estimate 1 min [95% CI −2 to 4 min], p = 0.42) showed a statistically significant difference between groups. However, patients in the SA group spent statistically significantly longer in the PACU than patients in the GA group (SA median 104 min [IQR 71–148 min] vs. GA 73 min [49–98 min], Hodges–Lehmann estimate 32 min [95% CI 23–42 min], p < 0.001). The PACU time data were missing from three patients.

FIGURE 1.

CONSORT flowchart. Modified from Palanne et al. [18].

TABLE 1.

Elapsed time and additional interventions.

	Spinal anaesthesia (n = 200)	General anaesthesia (n = 195)	p
OR time (min)	152 (136–168)	150 (137–165)	0.49
Non‐operative OR time (min)	65 (56–73)	63 (55–73)	0.42
PACU time (min)	104 (71–148) a	73 (49–98) a	< 0.001
Bolus for hypotension	47 (24%)	159 (82%)	< 0.001
Infusion for hypotension	1 (0.5%)	18 (9%)	< 0.001
Bolus for hypertension	18 (9%)	1 (0.5%)	< 0.001
Infusion for hypertension	8 (4%)	2 (1%)	0.11
Arterial cannulation	0 (0%)	1 (0.5%)	0.49
Urinary catheterisation	21 (11%) b	28 (14%)	0.27

Note: Values are presented as median (IQR) or number of patients (%).

Abbreviations: OR = operating room, PACU = post‐anaesthesia care unit.

^a Missing data for one patient in SA group (n = 199) and two patients in GA group (n = 193).

^b Missing data for three patients in SA group (n = 197).

Patients in the GA group required statistically significantly more interventions for hypotension than patients in the SA group. In the GA group, 159 patients (82%) received vasoactive drug boluses and 18 patients (9%) received a continuous noradrenaline infusion, whereas the corresponding numbers in the SA group were 47 (24%) and 1 (0.5%) (p < 0.001 for both comparisons). Conversely, more patients in the SA group (18, 9%) received labetalol bolus doses for hypertension than patients in the GA group (1, 0.5%; p < 0.001), but there was no significant difference in the number of patients receiving clonidine infusions (SA 8 [4%] vs. GA 2 [1%], p = 0.11). The frequency of urinary catheterisation was 21 (11%) in the SA group and 28 (14%) in the GA group (p = 0.27). Arterial cannulation was performed in only 1 of 395 patients; therefore, it was excluded from further evaluation of waste production and costs. Adjusted analyses showed comparable results (Supplementary Table 1).

Generated waste and costs related to anaesthesia methods and haemodynamic interventions are shown in Table 2. For SA anaesthesia protocol, single‐use equipment later sorted as mixed waste weighed 585 g per patient. The corresponding weight of single‐use mixed‐waste equipment for GA protocol was 1212 g. When the incidence of haemodynamic interventions and urinary catheterisation in our data was taken into consideration, the average weights increased to 611 g for SA and 1252 g for GA protocol per patient. Equipment used in each anaesthesia method is specified in Supplementary Table 2. Average total costs of medicines, single‐use equipment and haemodynamic interventions were higher for GA (€82.99 per patient) than for SA (€35.72 per patient).

TABLE 2.

Anaesthesia‐related mixed waste and costs of single‐use equipment and medicines.

		Spinal anaesthesia	General anaesthesia
Mixed waste (g)	Single‐use equipment and medicines	585	1212
	Treatment of hypotension a	1.7	11.4
	Treatment of hypertension a	2.9	0.7
	Urinary catheterisation b	21	28
	Total per patient	611	1252
Costs (€)	Single‐use equipment	22.45	65.74
	Medicines	11.37	14.11
	Treatment of hypotension a	0.68	2.53
	Treatment of hypertension a	0.82	0.07
	Urinary catheterisation b	0.40	0.54
	Total per patient	35.72	82.99

^a Bolus and infusion treatments, calculated as an average per patient.

^b Calculated as an average per patient.

4. Discussion

In this study, the anaesthesia method did not affect total OR time or non‐operative time, but SA was associated with a longer PACU stay. GA was associated with additional interventions for hypotension, more single‐use waste, and higher equipment and medicine costs. In GA, both generated mixed waste and expenses were approximately double those observed with SA.

Our data showed that patients in the SA group had an approximately half‐hour longer PACU stay than patients in the GA group. Although we are not aware of any previously validated or widely used minimal important difference for PACU time, we consider this difference clinically relevant, as it may play a role when planning ambulatory TKA, OR and PACU opening times, and personnel resource requirements. Nevertheless, estimating the cost of this 30‐min difference is complex. The average cost of a PACU stay varies considerably by healthcare system and surgical procedure, and includes, for example, heating and air conditioning, electricity, equipment investments, and staff costs. In our hospital system, PACU staff have preplanned working hours and care for several patients simultaneously. Therefore, a 30‐min reduction in one patient's PACU stay would not directly reduce costs unless, for example, the patient was the last PACU patient of the day and staff overtime was avoided. Furthermore, because length of hospital stay did not differ between groups in the original trial [18], transfer from the PACU to the ward would not stop the accumulation of other hospital costs [22]. Thus, 30 min in the PACU could theoretically narrow the cost gap between SA and GA, but comparing these figures directly is an oversimplification.

Previous literature has associated regional anaesthesia with shorter PACU stays [23]. The longer PACU stay in our study may partly have resulted from the study protocol, in which the standard bupivacaine dose was 15 mg. This dose may be slightly higher than that used by many anaesthetists outside study settings and could prolong motor block of the limb. At our hospital, the ability to raise the unoperated leg has been one of the PACU discharge criteria for recovery from regional anaesthesia, although discharge criteria vary. The anaesthetic and dose used in this protocol were chosen to match the study by Harsten et al. [24] to provide comparable results. PACU time may be shortened, for example, by using a smaller dose of standard bupivacaine or a shorter‐acting spinal anaesthetic, but this may affect the need for additional anaesthesia, conversion from SA to GA, surgical conditions, or patient anxiety or satisfaction. On the other hand, in OR setups where induction rooms and teams are available, performing SA induction during the previous surgery may reduce non‐procedural OR time and improve efficiency, allowing more operations to be performed [25].

Our findings are in line with multiple guidelines that suggest preferring regional anaesthesia over GA to reduce environmental impact [8, 9, 10]. In contrast, in McGain's study in Australia, SA had a greater carbon footprint than GA [26]. However, the results were mostly due to differences in OR time, oxygen consumption, washing of reusable items, and electricity‐associated emissions, which cannot be generalised worldwide. During McGain's study in 2021, the carbon intensity of electricity generation was several times greater in Australia than in Finland (606 vs. 135 gCO₂ kWh⁻¹) [27], which also increases the carbon footprint of reusable equipment in Australia [16]. Even if the 600‐g difference in waste produced per anaesthesia may appear clinically modest, it would result in tonnes of avoidable waste when considering the annual number of procedures performed in large arthroplasty centres or nationwide.

Our finding of lower costs in the SA group is consistent with previous studies suggesting that SA is associated with cost savings [23, 28]. However, the costs of anaesthesia equipment and medicines account for less than 0.5%–2% of the total costs of the TKA procedure. As with the difference in waste production, any potential savings become clinically relevant only at a much larger scale.

Previous studies have shown no significant difference in safety or surgical and patient‐reported outcome measures between GA and SA among selected patients [18, 19, 20, 29, 30]. The original RCT based on this patient cohort did not detect any significant differences in anaesthesia‐related or postoperative complications, such as pneumonia, hypoventilation, delirium, post‐dural puncture headache, nerve damage, or falls [18]. However, GA is associated with a higher incidence of intraoperative hypotension [29], as observed in this study. Without efficient treatment, intraoperative hypotension has been associated with a higher risk of complications, such as organ injury, in observational studies [31]. Managing hypotension requires more single‐use equipment and additional drugs, which increases waste production and expenses, especially during GA.

Despite significant actions at the management level and in product development, individual anaesthetists play a critical role in promoting sustainability. Clinicians might find it difficult and frustrating to consider environmental impacts in every small action and for every piece of plastic used. According to our results, a single anaesthetist, who in our system is often responsible for two ORs with two to three TKAs each per day, could reduce waste production by more than 3 kg day⁻¹ simply by choosing SA over GA when there are no patient‐related contraindications to SA. Moreover, the difference could become even more substantial if sedation is not required, or if it is administered as bolus doses instead of a continuous infusion. Environmental impacts might also be a priority for patients participating in decision‐making regarding their care. Simultaneously, there are multiple carbon emission‐producing factors that are hard to compromise without risking patient safety; for example, efficient OR air conditioning and adequate sterility practices are crucial in preventing complications, which can both be life‐threatening and increase financial and environmental costs substantially [32].

In addition to waste production and carbon emissions, we would like to mention possible water pollution due to pharmaceuticals. Although propofol has lower carbon dioxide equivalent emissions than volatile anaesthetics [6, 7], it is known to be environmentally toxic and, despite metabolism and correct disposal, measurable amounts have been detected in wastewater [33, 34]. Regional anaesthesia provides an opportunity to reduce polypharmacy and propofol use even when sedation is desired, due to fewer induction drugs and a less frequent need for hypotension treatment.

The strengths of this study are the randomised controlled trial design, prespecified secondary analysis plan with power calculation and prospective, systematic data collection. However, the study has several notable limitations. The inclusion criteria resulted in a selected subgroup of patients, and the results cannot be extrapolated to all patients undergoing TKA. In the data collection and analysis, we excluded non‐anaesthesia‐related factors because the aim of this study was to compare differences between anaesthesia methods rather than evaluate the overall cost and environmental impact of surgery. The routine equipment used at the surgical site in both groups, such as maintenance and packaging of instruments and implants, sterile operating gowns and drapes, adds up to a remarkable amount of single‐use waste across the entire operation. We were unable to evaluate electricity and oxygen use, which might differ between anaesthesia methods and have a significant environmental impact [26]. Unfortunately, we were unable to evaluate the carbon dioxide equivalent emissions or the overall pollution severity of the mixed waste. Further multidisciplinary studies with life cycle assessments would provide more precise information.

5. Conclusion

In patients undergoing TKA, anaesthesia method did not affect OR time. SA was associated with a longer PACU stay. GA required more interventions for hypotension and was associated with greater use of single‐use items and higher per‐patient costs. These findings support the use of SA for patients without contraindications in comparable anaesthesia units, particularly when reducing resource use and waste is a priority.

Author Contributions

K.L., E.R. and N.S. were responsible for the study design, data collection and analysis, reporting, interpretation and discussion of the results, manuscript preparation, and the publication process.

Funding

This randomised study was funded by the HUS Helsinki University Hospital Grant and HUS Helsinki University Finnish Government Science Grants (Y102011095, TYH2017239, TYH2019113).

K.L. has received grants for her PhD studies by The Paulo Foundation and Instrumentarium Science Foundation.

Ethics Statement

The study design was accepted by Helsinki University Hospital's Ethics Committee (8 June 2016, ref.: HUS1703/2016) and the Finnish Medicines Agency Fimea (20 May 2016, ref.: KL72/2016).

Consent

All participants provided informed consent.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Table 1. Elapsed time and additional interventions with adjusted tests.

Supplementary Table 2. Single‐use equipment and packaging used in spinal and total intravenous general anaesthesia.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.