Desflurane for maintenance of anaesthesia

Mortality or survival

Mortality was not reported in any of the papers on neurological procedures included in the evidence review.

Perioperative complications

Bhardwaj et al. (2018) found no statistically significant differences between desflurane and propofol in the incidence of vasospasm, infarct, tracheostomy, decompressive craniectomy or new onset neurological deficit at 24 hours. In Joys et al. (2019), there were no statistically significant differences between desflurane and isoflurane in the incidence or severity of postoperative delirium on day 1 or day 3.

Resource use

No studies reporting resource use in terms of monetary costs were identified. Studies by Bhagat et al. (2021) and Bhardwaj et al. (2018) found no statistically significant differences between desflurane and propofol in length of hospital stay. Similarly, Dube et al. (2015) found no statistically significant differences between desflurane and sevoflurane in length of hospital or intensive care unit stays.

Short-term recovery

At discharge, Bhagat et al. (2021), Bhardwaj et al. (2018) and Dube et al. (2015) found no statistically significant differences in the degree of disability or dependence on others for help with daily activities. Bhardwaj et al. (2018) also found that similar proportions of people in the desflurane and propofol groups had a good outcome, with no or only slight disability or dependence.

Sharma et al. (2020) found no statistically significant difference between desflurane and propofol in the proportion of people with cognitive impairment (defined as a score of less than 26 on the Montreal Cognitive Assessment Scale, a 30-point scale, with lower scores indicating a higher degree of cognitive impairment) at discharge or 2 weeks after surgery (81.6% compared with 65.4% respectively, p>0.05). By contrast, the mean cognitive impairment score was statistically significantly worse in the desflurane group than the propofol group at the same timepoint (19.09 compared with 22.81 respectively, p=0.013). It is unclear if the difference is clinically significant.

Longer-term recovery

Bhagat et al. (2021) found that, 3 months after discharge, there were no statistically significant differences between desflurane and propofol in 3 different measures of disability and dependence.

Limitations of the evidence

Four of the 5 randomised controlled trials that assessed desflurane for neurological procedures were generally well-designed and reported, but Sharma et al. (2020) failed to recruit sufficient participants and probably lacked statistical power. All studies were undertaken in India, which may limit their generalisability to the UK because of differences in, for example, ethnicity and genetics, socio-economic factors, healthcare systems and clinical practice. The generalisability of the results may also be limited for people undergoing some types of neurosurgery; for example, surgeries that last much longer than those in the studies. One of the studies compared desflurane with isoflurane, which specialist reviewers advised is not widely used in the UK. This means this study may have limited applicability to wider UK practice.

Two of the studies were considered to be at low risk of bias (Bhardwaj et al. 2018 and Joys et al. 2019), but there were some concerns over the other 3 (Bhagat et al. 2021, Dube et al. 2015 and Sharma et al. 2020). All 5 studies were small, with results analysed for between 49 and 91 participants only, divided across 2 groups. Therefore, some analyses may lack statistical power, particularly secondary outcomes in all the studies and all outcomes in Sharma et al. (2020), which means that this evidence is uncertain and we cannot exclude the possibility that clinically important differences may be seen in larger, sufficiently powered studies. Nevertheless, point estimates did not consistently favour 1 general anaesthetic over another.

Participants in all 5 studies were aged between 18 years and 60 or 65 years and were assessed as being relatively healthy, fully responsive with only minor brain injury, or at low risk of mortality. The results of the studies may not be applicable to children or older adults, or people with poor health status, severe brain injury or at higher risk of mortality.

In Bhagat et al. (2021), around 10% of people in each group were lost to follow up. Although the proportions were balanced across the groups, no reasons are reported so it is unclear if outcomes such as mortality were similar in the groups.

Blinding was generally adequate in the studies. However, only the neurosurgeons who measured intracranial pressure were blinded in Dube et al. (2015), and assessors for the outcomes relevant to the PICO were not blinded. This may be a source of bias in this study, but the relevant outcomes are reasonably objective.

In Sharma et al. (2020), the sample size was estimated based on the mean difference in cerebral metabolic rate with propofol compared with desflurane; however, cerebral metabolic rate was not reported in the study, suggesting the study was not powered correctly. Also, a large proportion of people were excluded from the study after randomisation, which was not addressed sufficiently in the sample size calculation.