European Journal of Cardiovascular Medicine

With the recent advances in anaesthesia, perioperative pain management has been emphasized greatly with the usage of regional anaesthesia pre-emptively as a part of multimodal analgesia. Pre-emptive pain management is essential to inhibit the enhanced pain perception and transmission that may happen if not tackled before the initiation of surgical stimulus.[1]

Pain after thyroid surgery is regarded as of moderate intensity and short duration and it has gained more importance recently because thyroid surgeries are being performed on day care basis.

NSAIDs may be less effective in providing analgesia and may increase the risk of postoperative bleeding. Opioid analgesics may cause postoperative nausea and vomiting and postoperative respiratory depression.[2] Bilateral cervical plexus block is used to provide effective analgesia for thyroid surgeries.[3] This technique involves bilateral injection of local anaesthetic behind the lateral border of sternocleidomastoid producing surface anaesthesia of the neck.

The ventral rami of C1-C4 forms the cervical plexus. The plexus of nerves can be blocked by injecting a local anaesthetic mixture based on anatomical landmark.[4]

It can also be visualised with their relation to surrounding structures with the help of ultrasound, thereby avoiding injury to the important surrounding structure.[5]

This study is done to compare the analgesic efficacy of pre-surgical blind (landmark based) Bilateral superficial plexus blocks v/s USG guided bilateral superficial cervical plexus blocks in patients undergoing thyroid surgeries under general anaesthesia.

AIMS AND OBJECTIVES

To assess and compare the analgesic efficacy of bilateral superficial cervical plexus block for thyroid surgery by USG v/s landmark technique.

Primary Objective

To assess and compare the postoperative analgesic efficacy in USG v/s Landmark technique.

Secondary Objective

1. To assess analgesic requirement intra-operatively.
2. Effect on haemodynamic profile between the 2 groups.
3. To assess other side effects like pain during swallowing, pain at incision site, Postoperative nausea and vomiting.

After ethical committee approval and with informed consent from the participating patients, after explaining the risks involved in the procedure, 46 patients undergoing elective thyroid surgery under general anaesthesia, meeting the selection criteria were included in the study.

Patients were randomly allocated into two groups using computer generated number lists and sealed envelopes. It is a double blind study where the assessor did not know the technique used for bilateral superficial cervical plexus block.

The block was given by an experienced anaesthesiologist not involved in the study.

Group A (n= 23) received US guided bilateral superficial cervical plexus block.

Group B (n= 23) received landmark guided superficial cervical plexus block.

12 ml of 0.25% Inj. Bupivacaine + 2mg of Inj. Dexamethasone on each side with total volume of 25ml was used for the study.

Inclusion Criteria

·        ASA I, II & III.

·        AGE -18 to 65 years.

·        Elective thyroid surgery.

Exclusion Criteria

·        Not giving consent

·        History of allergy to local anaesthetic.

·        Infection at puncture site.

·        Pregnancy.

·        Abnormal coagulation profile.

·        History of previous neck surgery causing altered anatomy.

·        CA thyroid patient requiring lymph node dissection.

·        Substernal extension of thyroid.

Place of Study

Major OT complex, BGS Global institute of medical sciences and hospital.

Sample Size

Software-G Power Version 3.1.9.7[6]

Analysis: A priori: Compute required sample size

Input: Tail(s) = One

Effect size d = 0.75

α err prob = 0.05

Power (1-β err prob) = 0.80

Allocation ratio N2/N1 = 1

Output: Noncentrality parameter δ = 2.543

Critical t = 1.680

Df = 44

Sample size group 1 = 23

Sample size group 2 = 23

Total sample size = 46

Actual power = 0.80

Anaesthesia Technique

All patients were fasted for 8 hours before the surgery. In the operation theatre, standard monitors were attached and baseline vital parameters were recorded. A peripheral intravenous access was secured with an appropriate size intravenous cannula.

Patients were pre-medicated with Inj. Glycopyrrolate 0.2mg IV, Inj. Ondansetron 4mg IV and Inj. Midazolam (0.04mg/kg) IV. General anaesthesia was induced after pre-oxygenation for 3 minutes with 6L/min O2 via a circle system with Inj. Fentanyl 2mcg/kg IV, Inj. Propofol 2 mg/kg IV, Inj. Vecuronium 0.1 mg/kg IV to enable endotracheal intubation. Anaesthesia was maintained with O2:N2O: Isoflurane (33%:66%:1-2%). After securing the airway, the patient was positioned for bilateral superficial cervical plexus block.

For USG guided bilateral superficial cervical plexus blocks, the machine used was Sonosite. Regional anaesthesia tray with sterile towels, gloves and gauzes was kept ready. A 2.5 inch 23 G needle attached to the extension tube was used for the block.

The block was performed with a patient lying supine and the head turned to the contralateral side. The transducer was placed at the lateral aspect of the patient's neck after skin sterilisation at the middle of the posterior border of sternocleidomastoid, such that the tapering edge is at the centre of the screen. Then the needle was introduced from the posterior aspect through the skin & platysma &12.5ml of the solution (0.25% bupivacaine with 2mg dexamethasone) was deposited just behind this landmark to give single point subcutaneous superficial Cervical Plexus Block. Block was repeated on the other side with the same technique.

In the landmark technique of block, the injection was administered at the midpoint of the posterior border of sternocleidomastoid at the level of cricoid cartilage. The needle was inserted to half the depth of the muscle and 12.5 ml of local anaesthetic was injected cephalad and caudad at the posterior border of sternocleidomastoid to block the supraclavicular, occipital and auricular branches. The onset of action of block is around 10-12 minutes, which is approximately the time needed for surgical site preparation and incision.

The intraoperative systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate (HR) were recorded at the time of the incision. Inj. Paracetamol 1G IV was given to all the patients soon after induction of anaesthesia. Anesthesia was maintained using a combination of O₂, N₂O, and isoflurane, with an isoflurane concentration of 1–1.2%. Duration of surgery noted. Fentanyl requirement intra-operatively noted.

At the end of the surgical procedure, after gentle suctioning, the reversal of neuromuscular blockade was done. The endotracheal tube cuff was deflated and removed only after the patients met the criteria of extubation. Patients were transferred to PACU.

Observation

Post-operative analgesia was evaluated by

1.      VAS

2.      Total Postoperative analgesia requirement

3.      Time to first rescue analgesia

VAS noted at the time of transfer and then every 4 hours for the first 24 hours.

Pain during swallowing, PONV, rescue antiemetics was also recorded.

Post-operative pain was evaluated with VAS (0= no pain, 10= worst pain).

If VAS >4 then rescue analgesics was administered as Inj. Diclofenac 75mg IV.

Nausea and vomiting was evaluated by PONV score as-

1= no nausea,

2= mild nausea,

3= severe nausea,

4= retching/or vomiting.

Severe PONV was defined as grades 3 and 4. Mild PONV was defined as Grade 1 and 2. When PONV occurred rescue antiemetics were administered Inj. Metoclopramide 10 mg IV.

The study population consists of 46 patients who were divided into two groups of 23 patients each in group A and group B.

Demographic Data

There were 22 females and 1 male in group A, 21 females and 2 males in group B. The groups were statistically similar with respect to sex distribution.

Table 1: Distribution of Study Participants based on gender

The mean age of patients in group A was 36.3 ± 2.59 years in 30-40 years age group, was 45.0 ± 3.16 years in 41-50 years age group, was 55.3 ± 2.13 in 51-60 years age group 63 ± 1.41 in more than 60 years age group. In group B was 36.3 ± 2.69 years in the 30-40 years age group, 43.8 ± 2.04 years in the 41-50 years age group, 55.4± 2.87 in the 51-60 years age group, and 63.5 ± 2.5 years in the more than 60 years age group. The groups were statistically similar with respect to age.

Table 2: Distribution of Study participants based on age

The table presents the distribution of study participants based on preoperative diagnoses, with identical proportions observed in both Group A and Group B. Each group shows consistent percentages for various thyroid conditions, including colloid goitre with cervical lymphadenopathy (8.7%), diffuse goitre swelling (4.3%), follicular carcinoma of the thyroid (8.7%), and others. Notably, multinodular goitre is the most prevalent condition, affecting 26.1% of participants in both groups, followed by solitary thyroid nodule (17.4%) and papillary carcinoma of the thyroid (13.0%). Other conditions like Graves' disease and solitary carcinoma of the thyroid each account for 4.3% of the cases.

Table 3: Distribution of study participants based on pre-operative diagnosis

The table details the distribution of study participants based on their surgical procedures, with both Group A and Group B showing identical proportions. The most common procedure is total thyroidectomy, performed on 40.9% of participants in each group. This is followed by total thyroidectomy combined with Modified Radical Neck Dissection (MRND), which accounts for 22.7% of participants in both groups. Hemithyroidectomy is the next most common, with 9.1% of participants undergoing either left or right hemithyroidectomy in each group. Additionally, hemi/total thyroidectomy, right hemithyroidectomy with neck dissection, and left hemithyroidectomy are each represented by 4.5% of participants. The data reflects a consistent approach to surgical procedures across both groups.

Table 4: Distribution of study participants based on surgical procedure

The table displays the distribution of study participants across two categories, I and II, with identical proportions in both Group A and Group B. In each group, 52.2% of participants are categorised under I, while 47.8% fall into category II. This uniform distribution indicates that both groups have an equivalent representation across the two categories.

Table 5: Distribution of study participants based on ASA grading

Heart Rate

Induction: The mean Heart rate was found to be higher in group B when compared to Group A and the difference was found to be statistically significant (p≤0.05).

Incision: The mean Heart rate was found to be higher in group A when compared to Group B and the difference was found NOT to be statistically significant.

End of Resection: The mean Heart rate was found to be higher in group A when compared to Group B and the difference was found NOT to be statistically significant.

Extubation: The mean Heart rate was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

Table 6: Distribution based on Baseline heart rate in two groups

Systolic Blood Pressure

Induction: The mean SBP was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

Incision: The mean SBP was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

End of Resection: The mean SBP was found to be higher in group B when compared to Group A and the difference was found to be statistically significant (p≤0.05).

Extubation: The mean SBP was found to be higher in group B when compared to Group A and the difference was found to be statistically significant (p≤0.05).

Table 7: Distribution based on baseline SBP in two group

Diastolic Blood Pressure

Induction: The mean DBP was found to be higher in group A when compared to Group B and the difference was found NOT to be statistically significant.

Incision: The mean DBP was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

End of Resection: The mean DBP was found to be higher in group B when compared to Group A and the difference was found to be statistically significant (p≤0.05).

Extubation: The mean DBP was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

Table 8: Distribution based on baseline DBP in two groups

Mean Blood Pressure

Induction: The mean MBP was found to be SAME in group A when compared to Group B.

Incision: The mean MBP was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

End of Resection: The mean MBP was found to be higher in group B when compared to Group A and the difference was found to be statistically significant (p≤0.05).

Extubation: The mean MBP was found to be higher in group B when compared to Group A and the difference was found NOT to be statistically significant.

Table 9: Distribution based on baseline MBP in two groups

VAS Score

The distribution of Visual Analog Scale (VAS) scores over time for Groups A and B reveals similar pain trends with no significant differences between the groups. At 0 hours, Group A reports a mean VAS score of 0.08 (SD 0.28), while Group B reports a higher mean of 0.34 (SD 0.71). Pain levels increase in both groups over time, with Group A reaching a mean score of 2.69 (SD 1.06) at 20 hours, compared to Group B's mean score of 2.78 (SD 1.27) at the same time point. Statistical comparisons show no significant differences in pain scores between the two groups at any time point, with p-values ranging from 0.1 to 1.0. This indicates that pain intensity, as measured by VAS scores, is comparable between Groups A and B throughout the observed period.

Table 10: Distribution based on VAS scores

Figure 1: Distribution based on VAS scores

Time to First Rescue Analgesic

The distribution of time to first rescue analgesic between Group A and Group B indicates notable differences in pain management needs. In Group A, a majority of participants (69.6%) did not require any rescue analgesic, whereas in Group B, only 34.8% of participants did not need additional pain relief. For those who did use rescue analgesics, Group B had higher proportions at earlier time points, with 17.4% requiring it within 4 hours, compared to 4.3% in Group A. Additionally, Group B showed usage at 8 hours (8.7%) and 16 hours (13.0%), while Group A had fewer participants at these intervals. Group A had a greater proportion of participants waiting until 20 hours (13.0%) for rescue analgesic compared to Group B (8.7%). These results suggest that participants in Group B generally required rescue analgesics sooner than those in Group A.

Table 11: Distribution based on time to first rescue analgesic

Figure 2: Distribution based on time to first rescue analgesic

Total Analgesic Consumption in the First 24 hours Post-Operatively

The distribution of total analgesic consumption in the first 24 hours postoperatively reveals distinct patterns between Group A and Group B. In Group A, 73.9% of participants did not require any analgesic, whereas 39.1% of Group B reported no analgesic use. Among those who did receive analgesics, Group B had a higher proportion using Inj. Diclofenac 75mg alone (47.8%) compared to 13.0% in Group A. Additionally, 13.0% of Group B participants used Inj. Diclofenac 75mg combined with another 75mg dose, while 8.7% of Group A participants used this combination. Only 4.3% of Group A used Inj. Diclofenac 75mg 8th hourly, with no participants from Group B using this regimen. These findings indicate that Group B had a higher overall consumption of analgesics (Inj. Diclofenac 75mg IV) compared to Group A.

Table 12: Distribution based on total analgesic consumption in the first 24 hours postoperatively

Post-Operative Nausea and Vomiting

The distribution of post-operative nausea and vomiting (PONV) across different grades in Group A and Group B shows some differences in incidence rates. In Group A, 39.1% of participants reported no nausea, while 30.4% experienced mild nausea, 17.4% had severe nausea, and 13.0% experienced retching or vomiting. In contrast, Group B had a lower proportion of participants reporting no nausea (30.4%) but a higher incidence of mild nausea (43.5%). Severe nausea was reported by 21.7% of Group B participants, and retching or vomiting occurred in only 4.3% of this group. Overall, both groups show variation in PONV severity, but Group B had a higher percentage of mild nausea and fewer cases of retching or vomiting compared to Group A.

Table 13: Distribution based on Post-operative nausea and vomiting

Pain While Swallowing

The distribution of pain while swallowing between Group A and Group B reveals a notable difference in the prevalence of this symptom. In Group A, 82.6% of participants reported no pain while swallowing, whereas 17.4% experienced pain. Conversely, Group B had a lower percentage of participants without pain (73.9%) and a higher percentage experiencing pain (26.1%). This indicates that Group B had a greater proportion of participants reporting pain while swallowing compared to Group A.

Table 14: Distribution based on pain while swallowing

Patients complain of moderate intensity pain following thyroid surgery.[7] Factors contributing to the perioperative pain include surgical incision and hyperextension position during surgery.[8] General anaesthesia alone does not suffice to control pain. There is an increase in stress hormones and other complications due to the pain, hence adequate analgesia is a must.[8] Patients may also experience discomfort in swallowing, burning sensation in throat, nausea and vomiting, hoarseness of voice.[9]

Bilateral superficial cervical plexus block can reduce intra- and postoperative pain. It is associated with the typical complications commonly seen in nerve blocks, including intravascular injection of local anesthetic (LA) into a vein or artery, hematoma, infection, and LA toxicity.[10] It is a popular regional anaesthesia used in thyroidectomy because of its feasibility and efficacy.[9]

Several studies have demonstrated the effectiveness of bilateral superficial cervical plexus blocks. Conventional landmark or ultrasound guided technique has been used for this block.[4]

Ultrasound-guided nerve blocks are highly effective due to the precise placement of local anaesthetic near the nerve, resulting in quicker and more robust anaesthesia. This technique enables real-time visualisation of needle movement and the direct dispersion of the local anaesthetic, enhancing both the safety and efficacy of the procedure compared to traditional landmark-based methods.[11]

In this study we have used 12 ml of 0.25% Inj. Bupivacaine + 2mg of Inj. Dexamethasone on each side with a total volume of 25ml.

Shah PM et al in their study ‘Utility of bilateral superficial cervical plexus block in thyroidectomy patients for post- operative analgesia’ used 0.25% bupivacaine 15 ml solution and found that the VAS score at frequent and regular time intervals in the postoperative period were significantly lower as compared to the control group.[8]

Ozgun et al in their study ‘Effect of bilateral superficial cervical plexus block on postoperative analgesic consumption in patients undergoing thyroid surgery’ administered a total volume of 20ml 0.5% Inj. levobupivacaine, 10 ml on each side and found that it significantly reduced postoperative analgesic consumption, rescue analgesic consumption requirement and pain severity.[7]

During this study, the heart rate and blood pressure parameters were compared between two groups at various stages. At induction, Group B had a significantly higher mean heart rate than Group A (p≤0.05), while Group B also had higher mean systolic blood pressure (SBP), but this

difference was not statistically significant. Throughout the incision and end of resection phases, Group A showed a higher mean heart rate, though the difference was not significant. Group B consistently had higher mean SBP at the end of resection and extubation, with statistical significance at both stages (p≤0.05). For diastolic blood pressure (DBP), Group B had higher values at the end of resection, which was statistically significant (p≤0.05), but not at other stages. Mean arterial blood pressure (MBP) was similar between the groups at induction, but Group B showed higher values at incision and end of resection, with statistical significance at the latter stage (p≤0.05).

However, P Veena et al in their study ‘Assessment of analgesic efficacy of bilateral superficial cervical plexus block for thyroid surgeries under general anaesthesia: A routine data based observational study’ found that statistical analysis revealed beneficial effect in intraoperative heart rate, systolic BP, diastolic BP which invariably points that the block is efficient enough to reduce intraoperative stress response during thyroid surgeries, which is in accordance with our study.[9]

Rasha M et al in their study ‘Analgesic efficacy of ultrasound guided versus landmark-based bilateral superficial cervical plexus block for thyroid surgery’ found that the block showed a significant decrease in SBP, HR (both at the time of incision and throughout the study) and intraoperative fentanyl given in patient who received the block (both US and LM guided) when compared to their control group (patients who received no block). However no significant differences were noted between the US and LM groups.[1]

In the current study, The Visual Analog Scale (VAS) scores for pain were similar between Groups A and B throughout the study period, with no significant differences at any time point, as p-values ranged from 0.1 to 1.0. At 0 hours, Group A had a mean VAS score of 0.08 (SD 0.28) compared to 0.34 (SD 0.71) for Group B. By 20 hours, pain levels increased to 2.69 (SD 1.06) for Group A and 2.78 (SD 1.27) for Group B. In terms of rescue analgesic use, Group A had a higher proportion of participants (69.6%) who did not need additional pain relief compared to Group B (34.8%). Group B participants required rescue analgesics sooner, with 17.4% needing it within 4 hours versus 4.3% in Group A. The total analgesic consumption also differed, with 73.9% of Group A participants not needing analgesics, while 39.1% of Group B did not use any. Group B used Inj. Diclofenac 75mg more frequently and in higher doses than Group A, reflecting greater overall analgesic consumption in Group B.

Rasha M et al. in their study ‘Analgesic efficacy of ultrasound guided versus landmark-based bilateral superficial cervical plexus block for thyroid surgery’ found that bilateral superficial cervical plexus blocks (BSCPB), administered under general anaesthesia for thyroid surgeries, notably decreased the need for postoperative analgesics compared to systemic narcotics alone, regardless of whether the technique was ultrasound-guided or landmark-based. VAS pain scores were significantly lower in the first 24 hours after surgery, and the time to first rescue analgesic was longer with BSCPB, suggesting more effective pain control. Nonetheless, there were no significant differences in analgesic efficacy or safety between the ultrasound-guided and landmark-based approaches.[1]

In our study, Group A had 82.6% of participants reporting no pain while swallowing, compared to 17.4% who experienced pain. In Group B, 73.9% reported no pain, while 26.1% experienced pain. Thus, Group B had a higher incidence of pain while swallowing than Group A. Meanwhile, Elmaddawy et al said in their study ‘Ultrasound- guided bilateral superficial cervical plexus block for thyroid surgery: the effect dexmedetomidine addition to bupivacaine-epinephrine’ that hoarseness of voice was detected in 24% and 19% of their patients in two groups. They found that the incidence of hoarseness was variable 4% to 28% when using US guided block and that it may be explained by the difference in LA concentrations or the injection close to carotid artery, infiltration of vagus nerve which may cause temporary paralysis of the ipsilateral recurrent laryngeal nerve.[11]

In our study, The incidence of postoperative nausea and vomiting (PONV) varied between Groups A and B. In Group A, 39.1% of participants reported no nausea, while 30.4% experienced mild nausea, 17.4% had severe nausea, and 13.0% experienced retching or vomiting. In Group B, 30.4% reported no nausea, 43.5% had mild nausea, 21.7% experienced severe nausea, and only 4.3% had retching or vomiting. Overall, Group B had a higher rate of mild nausea and fewer instances of retching or vomiting compared to Group A.

However, P Veena et al in their study ‘Assessment of analgesic efficacy of bilateral superficial cervical plexus block for thyroid surgeries under general anaesthesia: A routine data based observational study’ that they couldn't find a statistically significant reduction of PONV after the block, meanwhile it was reported El Galeel et al that the incidence of PONV is less after the block. But we, in our study, found that Overall, both groups show variation in PONV severity, but Group B had a higher percentage of mild nausea and fewer cases of retching or vomiting compared to Group A.[9]

Limitations of Our Study

The patient's subjective experience of the addressed pain differed from individual to individual.[9]

In patients undergoing thyroidectomy, ultrasound-guided superficial cervical plexus blocks resulted in lower postoperative visual analog scale (VAS) scores, improved intraoperative hemodynamic parameters, and reduced postoperative analgesic use compared to landmark-guided cervical plexus blocks.[12] However, the differences observed were not statistically significant.

1. Hassan RM, Hashim, RM. Analgesic efficacy of ultrasound guided versus landmark-based bilateral superficial cervical plexus block for thyroid surgery. Egyptian Journal of Anaesthesia 2017;33(4):365-73.
2. Andrieu G, Amrouni H, Robin E, Carnaille B, Wattier JM, Pattou F, et al. Analgesic efficacy of bilateral superficial cervical plexus block administered before thyroid surgery under general anaesthesia. British Journal of Anaesthesia 2007;99(4):561-6.
3. Shih ML, Duh QY, Hsieh CB, Liu YC, Lu CH, Wong CS, et al. Bilateral superficial cervical plexus block combined with general anesthesia administered in thyroid operations. World Journal of Surgery 2010;34:2338-43.
4. Syal K, Chandel A, Goyal A, Sharma A. Comparison of ultrasound-guided intermediate vs subcutaneous cervical plexus block for postoperative analgesia in patients undergoing total thyroidectomy: a randomised double-blind trial. Indian J Anaesth 2020;64(1):37-42.
5. Mayhew D, Sahgal N, Khirwadkar R, Hunter JM, Banerjee A. Analgesic efficacy of bilateral superficial cervical plexus block for thyroid surgery: meta-analysis and systematic review. Br J Anaesth 2018;120(2):241-51.
6. Faul, F, Erdfelder E, Lang AG, Buchner A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods 2007;39(2):175-91.
7. Ozgun M, Hosten T, Solak M. Effect of bilateral superficial cervical plexus block on postoperative analgesic consumption in patients undergoing thyroid surgery. Cureus 2022;14(1).
8. Shah PM, Kheskani DN. Utility of bilateral superficial cervical plexus block in thyroidectomy patients for post-operative analgesia. Int J Res Med Sci 2021;9(12):3643-7.
9. Veena P, Rajasree O, Koshy RC, Krishna J. Assessment of analgesic efficacy of bilateral superficial cervical plexus block for thyroid surgeries under general anaesthesia: a routine data based observational study. Int J Anesthesiol Pain Med 2021;7:40.
10. Karakış A, Tapar H, Özsoy Z, Suren M, Dogru S, Karaman T, et al. Perioperative analgesic efficacy of bilateral superficial cervical plexus block in patients undergoing thyroidectomy: a randomized controlled trial. Revista Brasileira de Anestesiologia 2019;69(5):455-60.
11. Elmaddawy AE, Mazy AE. Ultrasound-guided bilateral superficial cervical plexus block for thyroid surgery: The effect of dexmedetomidine addition to bupivacaine-epinephrine. Saudi Journal of Anaesthesia 2018;12(3):412-8.