Speed of spinal vs general anaesthesia for category-1 caesarean section: a simulation and clinical observation-based study†
Presented in part at the Society for Obstetric Anesthesia and Perinatology 42nd Annual Meeting, Texas, May 2010 and the Obstetric Anaesthetists' Association Annual Meeting, Newcastle, May 2010.
You can respond to this article at http://www.anaesthesiacorrespondence.com
Correction added on 29 May 2013, after first online publication: adrenaline changed to ephedrine in the Methods section.
Correspondence to: A. Kathirgamanathan
Controversy exists as to whether effective spinal anaesthesia can be achieved as quickly as general anaesthesia for a category-1 caesarean section. Sixteen consultants and three fellows in obstetric anaesthesia were timed performing spinal and general anaesthesia for category-1 caesarean section on a simulator. The simulation time commenced upon entry of the anaesthetist into the operating theatre and finished for the spinal anaesthetic at the end of intrathecal injection and for the general anaesthetic when the anaesthetist was happy for surgery to start. In the second clinical part of the study, the time from intrathecal administration to ‘adequate surgical anaesthesia’ (defined as adequate for start of a category-1 caesarean section) was estimated in 100 elective (category-4) caesarean sections. The median (IQR [range]) times (min:s) for spinal procedure, onset of spinal block and general anaesthesia were 2:56 (2:32 - 3:32 [1:22 - 3:50]), 5:56 (4:23 - 7:39 [2:9 - 13:32]) and 1:56 (1:39 - 2:9 [1:13 - 3:12]), respectively. The limiting factor in urgent spinal anaesthesia is the unpredictable time needed for adequate surgical block to develop.
Rapid delivery of the fetus by category-1 caesarean section is necessary when there is immediate threat to the life of mother or fetus. Under these circumstances, the National Institute for Health and Clinical Excellence recommends delivery as quickly as possible after the decision is made . Thus, the time taken to achieve adequate surgical anaesthesia should be kept as short as possible .
In the absence of a pre-existing epidural in labour, anaesthesia may be provided by spinal or general anaesthesia. Some anaesthetists believe that spinal anaesthesia can be administered as quickly as general anaesthesia and often persist in performing spinal for fear of the risks associated with general. Prospective randomised controlled trials comparing the two techniques under emergency situations may be ethically unacceptable. Simulation allows anaesthetists to practise safe emergency anaesthesia. In a pilot simulation study, insertion of spinal anaesthesia was found to be as quick as general, but the time to achieve surgical anaesthesia was longer .
We undertook a simulation and clinical observation-based study to compare the speed of spinal vs general anaesthesia for category-1 caesarean section. We also performed a questionnaire survey to investigate the techniques commonly used by anaesthetists in our hospital to expedite readiness for surgery.
After approval by the University of British Columbia Clinical Research Ethics Board, written informed consent was obtained from 16 consultants and three fellows in obstetric anaesthesia. The ethics board advised that informed consent for the clinical section of the study, which consisted of timing spinal anaesthesia from intrathecal injection to adequate surgical anaesthesia, was unnecessary because it was an observation of normal practice, and women were managed as per the existing protocols at the hospital.
A number of assumptions were made regarding the simulation environment, parturient and preparation before start of the study procedure. The scenario was that the anaesthetist was fast-bleeped to attend a ‘stat’ category-1 caesarean section and advised on arrival of the type of anaesthetic to administer. The patient was of ASA physical status 1 and was in the left lateral position for spinal insertion and supine with lateral pelvic tilt using a wedge for general anaesthetic induction. The lumbar spine spaces were easily palpable such that spinal insertion occurred on the first attempt. There was a Cormack & Lehane grade-1 laryngoscopic view, allowing intubation on the first attempt. Pre-operative assessment had been completed, the anaesthetic machine checked, and an experienced anaesthetic assistant was available to position the patient and apply cricoid pressure. There was a working intravenous cannula in situ, and emergency drugs (thiopental, suxamethonium, phenylephrine and ephedrine) had been drawn up as is normal in our institution.
Spinal simulation was performed on a Styrofoam® model of the back with a saline bag attached posteriorly to simulate cerebrospinal fluid. Simulation time included opening of the spinal tray (Med-Rx®; Benlan Inc, Oakville, ON, Canada), application of gloves, sterile preparation, drawing up of drugs, local anaesthetic infiltration, spinal needle insertion and injection. The volume and vials of simulated intrathecal drugs drawn up were identical to the drugs used in clinical practice at our institution. The time taken for the simulated spinal procedure was defined as from the entry of the anaesthetist into the operating theatre until the end of the intrathecal injection.
General anaesthesia was simulated with an airway training manikin (Laerdal Medical, Stavanger, Norway) placed on an operating theatre table. To simulate induction agents, saline 0.9% was drawn up in the same sizes of syringe as those used in clinical practice at our institution. Simulation time included pre-oxygenation, injection of the drugs into the intravenous cannula, application of cricoid pressure and intubation of the trachea. The time for general anaesthetic induction was from the entry of the anaesthetist into the theatre until the anaesthetist said that surgery could commence. Each anaesthetist administered simulated anaesthesia in the order spinal-general-spinal-general.
The time to obtain surgical anaesthesia could not be simulated, so it was measured clinically in 100 women having elective (category-4) caesarean section with the spinal inserted in the lateral position. The intrathecal anaesthetic consisted of 11.25 mg hyperbaric bupivacaine, 10 μg fentanyl and 100 μg morphine. The ‘ready for surgery’ time was defined as the period from the end of intrathecal injection until when the anaesthetist would allow the surgeon to commence surgery if it was a category-1 caesarean section. A ‘total spinal time’ was also calculated by adding the simulated spinal induction time to the observed block onset time; this was done for the shortest of these two times, the median of the two and the longest of the two to give an estimate of the optimistic, expected and pessimistic values. Data were entered onto Microsoft™ Excel 2003.
Two weeks after completion of the simulation study, the participating anaesthetists completed an open-ended questionnaire pertaining to the modifications they might make for spinal and general anaesthesia for a category-1 caesarean section. A further question related to the minimum acceptable block height required for caesarean section under spinal anaesthesia and the modality used for testing the block in the situation of either an elective case, a category-1 case with no suspected difficulty in intubation or a category-1 case with suspected difficulty in intubation.
All 19 anaesthetists completed the study and questionnaire. The time from entry of the anaesthetist into the operating theatre until completion of anaesthetic induction is shown in Table 1. For each anaesthetist, the best time for each type of anaesthetic was used. The second attempt was faster for 16 (84%) anaesthetists performing spinal and 15 (79%) performing general. For one (5%) anaesthetist, the times for performing general were identical. The median (IQR [range]) difference in times (min:s) between the faster and slower attempts for spinal procedure and general anaesthesia was 0:44 (0:24 - 1:7 [0:2 - 1:12]) and 0:18 (0:7 - 0:42 [0:0 - 1:19]), respectively.
Table 1. Times (min:s) for simulated and clinical spinal anaesthesia, and simulated general anaesthesia. Values are median (IQR [range]) or median (range)
|2:56 (2:32 - 3:32 [1:22 - 3:50])||5:56 (4:23 - 7:39 [2:9 - 13:32])||8:52 (3:31 - 17:22)||1:56 (1:39 - 2:9 [1:13 - 3:12])|
Table 1 also shows the time from intrathecal injection to ‘ready for surgery’ in 100 parturients. One woman required conversion to general anaesthesia because of intra-operative pain. She was judged by the anaesthetist to have had an adequate block height for a category-1 caesarean section at 7 min 48 s.
Replies to the questionnaire are shown in Tables 2 and 3. The total number of responses exceeds 19 as some anaesthetists gave more than one suggestion. 17 (90%) used temperature as the first-line modality to test block height. One (5.3%) anaesthetist used pinprick as the first modality, and another did not state the modality used. Second-line modalities used were pinch, light touch and pinprick used by six (32%), four (21%) and four (21%) anaesthetists, respectively.
Table 2. Suggested techniques for expediting anaesthesia. Values are number (proportion) of anaesthetists who made the suggestion
| Prepare drugs before arrival of patient, change gloves to help position if necessary||1 (5%)|
| Experienced assistant to position while anaesthetist prepares drugs||1 (5%)|
| No local anaesthetic to skin||1 (5%)|
| Larger dose of intrathecal bupivacaine (12–15 mg)||5 (26%)|
| No opioid or fentanyl with intrathecal bupivacaine||2 (11%)|
| Use 3-ml syringe to draw up all intrathecal drugs (rather than using 1-ml syringe to draw up morphine)||1 (5%)|
| Head-down tilt, barbotage, Valsalva manoeuvre||3 (16%)|
| Start surgery once rising block has reached T10 dermatome||1 (5%)|
| Experienced assistant to position, pre-oxygenate and apply cricoid pressure||9 (47%)|
| Get help from another anaesthetist||1 (5%)|
| Start pre-oxygenating early; 3–8 deep breaths to pre-oxygenate||5 (26%)|
| Give opioid, for example fentanyl or remifentanil before induction||3 (16%)|
| Give induction agents quickly as soon as abdomen sterilised and surgeon scrubbed||7 (37%)|
| Use overpressure of volatile agent to speed alveolar uptake||1 (5%)|
| Use 50:50 inspired oxygen/nitrous oxide rather than oxygen/air||1 (5%)|
| Commence surgery as soon as tracheal intubation confirmed||1 (5%)|
| Cancel instrument count and X-ray afterwards||1 (5%)|
Table 3. Anaesthetists' recommendation of minimum height of spinal block required for surgery to commence in various clinical circumstances. Values are number (proportion)
|Elective caesarean section|
| Temperature||T4||13 (68%)|
| Temperature||T6||4 (21%)|
| Pinprick||T6||3 (16%)|
| Pinch||T8||4 (21%)|
| Not specified|| ||3 (16%)|
|Category-1 caesarean section without suspected difficulty in intubation|
| Temperature||T6||7 (37%)|
| Temperature||T8||4 (21%)|
| Pinprick||T6||4 (21%)|
| Pinprick||T10||3 (16%)|
| Not specified|| ||3 (16%)|
|Category-1 caesarean section with suspected difficulty in intubation|
| Temperature||T4||5 (26%)|
| Temperature||T6||6 (32%)|
| Temperature||T8||2 (11%)|
| Pinch||T6||3 (16%)|
| Pinch||T8||1 (5%)|
| Not specified|| ||2 (11%)|
Regional anaesthesia is recommended to facilitate delivery of the fetus as it is ‘safer and results in less maternal and neonatal morbidity than general anaesthesia’ (Grade A recommendation) . However, general anaesthesia is used more often in category-1 caesarean sections as surgery can commence upon confirmation of tracheal tube placement. In a UK survey, the median (IQR [range]) rate of use of general anaesthesia for category-1 caesarean section was 51% (29–80% [6–100%]) . Marx et al. found that spinal anaesthesia can be induced acceptably quickly for caesarean section for fetal distress . However, they used a 22-G spinal needle which is significantly bigger than those in current use, and amethocaine which is also not commonly used in modern practice.
Physiological changes of pregnancy can lead to complications of general anaesthesia such as difficult tracheal intubation, pulmonary aspiration of stomach contents and awareness of intra-operative events [6, 7]. The increased use of regional techniques means that anaesthetists are becoming less experienced in administering general anaesthesia for obstetric patients [8, 9]. In the last Confidential Enquiries into Maternal Deaths report into maternal mortality, four deaths were directly attributable to general anaesthesia . The cause of death in these cases was oesophageal intubation, dislodgement of a tracheostomy tube in critical care, aspiration of gastric contents at tracheal extubation and opioid toxicity secondary to patient-controlled analgesia. Problems due to general anaesthesia are likely to become more frequent as the use of general anaesthesia declines and the prevalence of obesity in parturients rises.
A direct comparison of the speed of spinal and general in a prospective randomised trial for category-1 caesarean section would not be ethically acceptable. We therefore designed a combined simulation and observation study under controlled conditions. In our study, the median (IQR [range]) time (min:s) to perform the procedural aspect of the spinal was 2:56 (2:32 - 3:32 [1:22 - 3:50]). This time included preparation for spinal and intrathecal injection, but not the time required for adequate surgical block to develop. The median (IQR [range]) time for general anaesthetic induction was 1:56 (1:39 - 2:9 [1:13 - 3:12]). Both spinal and general induction times in our study are shorter than those found in a simulation study by Macafee et al., who reported the median (range) times for spinal and general anaesthesia to be 5:12 (2:57 - 9:9) and 3:12 (1:47 - 3:24), respectively . Macafee et al. used high-fidelity simulation where the drug and equipment availability was comparable to a labour ward theatre. In addition, their anaesthetists had to read a scripted scenario before performing anaesthesia. In our simulation, all the measurements were taken in our regular operating theatre, the drugs and equipment were pre-prepared, and the anaesthetist did not have to read any written instructions. We speculate that these subtle disparities in the simulation set-up may explain the relatively small differences in the measured times.
In our clinical observations, the ‘ready for surgery time’, estimated as if the elective case was actually category-1, was remarkably variable with a median (IQR [range]) time of 5:56 (4:23 - 7:39 [2:9 - 13:32]). The level of the spinal block at this point relied upon the anaesthetist's judgement as to when the block height was sufficient to start surgery. This may explain some of the large variability in the timings, besides any true variation in block onset according to objective measures. As shown by the responses to the questionnaire, some anaesthetists may insist that the block height is T4 to cold before commencing surgery for a category-1 caesarean section, whereas others may accept a lower block height as long as it is ascending. The threshold for accepting spinal failure resulting in intra-operative pain may also differ between anaesthetists, with some allowing moderate pain controlled by short-acting supplemental analgesia and others not accepting this.
The total spinal times for optimistic (fastest preparation + fastest onset), median (median preparation + median onset) and pessimistic (slowest preparation + slowest onset) were 3:31, 8:52 and 17:22, respectively. Our times are comparable to a clinical observational study by Kinsella et al. who described a case series of 25 ‘rapid sequence spinals’ and found the median (IQR [range]) time to prepare and perform the spinal was 2 (2–3 [1–7]) min and to develop a ‘satisfactory’ block was 4 (3–5 [2–7]) min. The optimistic time calculated in the same way was 3 min, and the pessimistic was 14 min. The total observed time to induce spinal anaesthesia was 8 (7–8 [6–8]) min .
A recent survey found that training in performance of rapid sequence spinal anaesthesia has been offered to 73% of trainee anaesthetists in the Severn region of England . However, 9% of those who had received training would not use the technique, and 22% were unsure if they would use it in clinical practice. At present, the rapid sequence spinal technique is not used nationwide in the UK. There are also fears among trainees that spinal may be slower compared with general anaesthesia, and the perceived lack of sterile technique may be criticised .
Our study has a number of limitations. In this simulation scenario, we analysed only one step, the anaesthetic technique, in the complex process between decision and delivery of the baby . The participating anaesthetists may not have faced the same pressure as in real life, and thus, the time measurements in our study may not be comparable to the clinical situation. We assumed that spinal insertion and intubation of the trachea occurred on the first attempt. In the real category-1 situation, more than one attempt may be required and the time taken will sometimes be longer than that recorded with simulation.
We did not randomise the order of the anaesthetics because the set-up of our study has not been described previously. We expected a learning curve and were interested in what the ‘best possible’ time achievable by each anaesthetist would be. For the majority of anaesthetists, the times for the first and second simulations did demonstrate a learning effect although the absolute time differences were not clinically significant. This may be because all participants were experienced and used to working in emergency situations.
There has been debate about the influence of patients’ height and weight upon the spread of spinal anaesthesia. McCulloch and Littlewood found that block height was higher in obese patients, but the variation of spread in relation to height and weight was unpredictable . In the clinical part of the study, we did not record patients’ physical characteristics or the actual block height when the anaesthetist judged the patient was ‘ready for surgery’ for a category-1 caesarean section. We hoped to minimise the effect of these factors by averaging the data of 100 patients. To minimise the effect of different drug dose, we only studied patients receiving 11.25 mg bupivacaine.
In our questionnaire survey, most anaesthetists recommended a team approach as the most important factor in increasing the speed of both spinal and general. Obtaining experienced help, especially from a fellow anaesthetist, to gain intravenous access and draw up drugs, can save minutes. Preparation time may be reduced by omitting opioid. Using a larger dose of intrathecal bupivacaine, barbotage and the Trendelenburg position may expedite the onset of the spinal block. More controversial is the decision to start surgery when the block height is below T4 but is ascending. In our survey, anaesthetists were willing to accept a lower block height before starting surgery for a category-1 case than for an elective case. However, the acceptable block height was higher for a category-1 case if difficulty in managing the airway was anticipated. In our study, spinal was converted to general anaesthesia in one of the 100 parturients because of inadequate block height. In the case series by Kinsella et al., 3 (12%) out of the 25 parturients complained of intra-operative pain . As intra-operative pain is the most common cause of litigation in obstetric anaesthesia, appropriate selection and discussion with patients are essential if one decides to accept a lower block height in an emergency scenario , and it is especially important to warn the woman that she may experience some discomfort until the baby is born.
Taking 4–8 full vital capacity breaths, rather than traditional three-minute pre-oxygenation before induction of general anaesthesia, is an acceptable option . Overpressure of the volatile anaesthetic immediately after confirmation of tracheal tube position was also recommended by some anaesthetists.
This combined simulation and clinical observation study provides evidence that induction of spinal is slower than general for an emergency caesarean section. The limiting factor in spinal anaesthesia appears to be the unpredictability in the time to obtain adequate surgical block. Anaesthetic technique is one step in the process between decision and delivery. Morgan et al. found that approximately 87% of emergency caesarean sections were predictable . The anaesthetist should attend the obstetric ward round to anticipate possible problems. This can aid establishment of effective epidural analgesia which can be extended to provide anaesthesia should a category-1 caesarean section be required. The anaesthetist should be proactive and involve the whole multidisciplinary team, encouraging team members to work cooperatively to expedite induction of anaesthesia. For example, while the anaesthetist is donning sterile garb for a spinal, the team can move the parturient into theatre, obtain intravenous access, position the patient and apply chlorhexidine spray to the back. Effective teamworking in the emergency obstetric situation is crucial, and training for this using simulation and labour ward drills was recommended by the last two Confidential Enquiries into Maternal Deaths [10, 19].
We would like to thank Michael Papsdorf for his statistical advice.
No external funding and no competing interests declared.
Labat spinal set, post 1930 Set for continuous spinal anaesthesia, with short sharp bevel needle believed to reduce dural trauma. From the Association of Anaesthetists of Great Britain & Ireland Anaesthesia Heritage Centre. ©Association of Anaesthetists of Great Britain & Ireland.