Pressure waveforms to assess epidural placement: is there a role on delivery suite?

Failure is simply the opportunity to begin again, this time more intelligently.Henry Ford The timely provision of reliable analgesia during labour remains a primary concern of anaesthetists covering delivery suite. Although approximately 20% of parturients (127,000 women) in England 1 (and up to 61% of mothers in the USA 2) avail themselves of an epidural during labour, providing successful epidural analgesia is an imperfect art. The failure rate of neuraxial labour analgesia has been reported to be in the region of 8–23% 3, and many of these are primary failures, i.e. due to incorrect placement of the epidural catheter 4. Whilst lumbar epidurals per se tend to have a higher success rate than thoracic 5, labour analgesia is often provided at short notice by junior anaesthetists, out-of-hours, to patients rendered relatively non-compliant through pain. Changes in maternal anatomy also contribute to failures: ultrasound assessment to compare women during pregnancy and 9 months post-delivery demonstrated that the soft tissue channel between spinal processes is smaller, the skin-epidural space distance is greater and the epidural space is narrower 6. At best, failed labour analgesia leads to inadequate or delayed analgesia, reduced satisfaction and increased anaesthetic workload. At worst, in patients having epidurals sited because of significant cardio-respiratory disease or other comorbidities, failed analgesia could have significant impact on maternal and fetal outcomes. Therefore, any technique that aims to reduce the incidence of primary failure of labour analgesia is welcome. This issue of Anaesthesia contains an article by Al-Aamri et al. that examines the use of epidural pressure waveforms to achieve this 7. But how does this technique fit into the gamut at our disposal to prevent primary analgesic failure? The most common method to identify the epidural space is the ‘loss of resistance’ (LOR) technique 8, an end-point achieved when the user enters the low-pressure epidural space via the dense homogenous texture of the ligamentum flavum. Whilst this is usually accomplished using tactile feedback from a syringe filled with air or saline, other techniques have been used. The ‘hanging drop’ method relies on a suspended drop of normal saline being ‘sucked’ into the epidural needle on entering the sub-atmospheric pressure of the epidural space. Whilst arguments persist as to whether the epidural space is truly sub-atmospheric, this technique seems best suited to thoracic epidurals performed in the sitting position 9, and is rarely practiced in obstetrics 8. Various spring-loaded syringes have been described 10, 11; those in current use are designed to be filled with saline and attached to a standard epidural needle that has been advanced into the inter-spinous ligament. On further advancement of the needle, the syringe will automatically discharge the saline once the low-resistance epidural space is reached. Results in the obstetric setting have been mixed, with one study finding that inexperienced anaesthetists had an increased incidence of primary failure when compared with traditional saline loss-of-resistance 11. Irrespective of the method used, all these LOR techniques lack specificity, since tissues other than the epidural space may also provide a loss-of-resistance 12. Mid-line gaps in the ligamentum flavum, caused by a failure of its embryologic left and right parts to close, can occur in 35% and 11% of the population at T10–T12 and L3–4 respectively 13, 14. The absence of the typical ‘feel’ of the ligamentum flavum may lead to a reduced or absent LOR. Cysts within the inter-spinous ligaments can also produce a false positive LOR, and whilst they tend to be age related (seen in 85% of cadavers aged 61–79 15), they do also occur in those of child-bearing age 16. Furthermore, large cysts can accommodate up to 20 ml of injected fluid with ease 16. Non-midline needling can produce a falsely reassuring LOR when a paravertebral muscle is entered 12, 17, as may other intermuscular planes such as the quadratus lumborum and psoas muscles 18. Using ultrasound as a pre-procedural tool has been widely studied, with varying outcomes. Ultrasound may be advantageous in estimating distance to the epidural space and detecting the correct puncture site and needle angulation 19, 20; however it is worth noting that the acoustic visibility of the ligamentum flavum, dura mater and epidural space may decrease significantly during pregnancy 6. Two recent meta-analyses of randomised studies have shown a reduction in failed neuraxial blockade using pre-procedural ultrasound compared with traditional palpation, however these included studies from non-obstetric populations, some were reliant on expert operators 21, 22. Two recent randomised studies in obstetrics, conducted after these meta-analyses, failed to show improvements in epidural success using ultrasound 23, 24. Whilst some studies have questioned the benefits of ultrasound in experienced hands 23, for the less experienced anaesthetist there is evidence that using ultrasound imaging for teaching obstetric epidural analgesia results in a higher rate of first-pass success and an improved learning curve 25. Ultrasound may also be useful in those parturients deemed ‘difficult’ due to obesity or abnormal anatomical conditions such as spinal deformity, scoliosis or previous spinal surgery 26. Ultrasound has been used by expert practitioners to help real-time guidance of the needle during the epidural placement 27, 28, although its utility is limited by requiring either a second operator or a spring-loaded syringe, and concerns regarding the risk of ultrasound gel contaminating neuraxial structures remain. However, with the advent of handheld, pocket-sized ultrasound devices for automatically identifying epidural location and depth 29, it seems likely that ultrasound will be increasingly used by obstetric anaesthetists in the future. Irrespective of any argued benefits in terms of labour analgesia and avoidance of sacral sparing, using a needle-through-needle technique can be used as a confirmatory end-point that the epidural needle-tip is positioned in the epidural space. Dural puncture epidural (DPE), where a dural perforation is created with a spinal needle via an epidural needle, but no intrathecal medication is given, may work by allowing some epidurally-delivered drug to pass through the dural puncture into the CSF 30. In a recent study comparing DPE, combined spinal epidural (CSE) and epidural, DPE showed improved block quality over epidurals, with fewer side effects than CSEs 31. However, all needle-through-needle techniques necessitate a low lumbar insertion (to avoid damage to the conus medullaris), which may make siting the epidural more challenging due to an increased distance from skin to the epidural space, and a narrower inter-laminar space and ‘window’ into the epidural space 32. Epidural test doses in obstetrics are routinely used to rule out intrathecal and intravascular placement before giving a therapeutic dose 12, but could also be used to assess that the epidural catheter is effective, as is often done in the surgical setting. Time intervals as low as 5 min have been used to demonstrate bilateral sensory block to ice, but this timing can be highly variable, and it may take up to 15 min in some patients 33. The lack of real-time feedback potentially limits the usefulness of this method to confirm successful epidural placement in obstetrics. CT fluoroscopic studies have proved enlightening by examining various catheter tip positions after mid-line epidural insertion. Upon entering the epidural space, catheters tend to lodge anteriorly and laterally (presumably as a result of being deviated sideways and forwards by the triangular epidural fat pad; imagine the catheter being pushed towards the apex of a Toblerone box). In one study of non-pregnant women undergoing brachytherapy (n = 20), 45% of lumbar epidural catheter tips ended up in the intervertebral foramina and 10% in the paravertebral tissues 34. The authors found that in spite of the variety of catheter tip positions attained, successful bilateral analgesia was achieved. Whilst of interest, the routine use of fluoroscopy for labour epidural placement and confirmation of the catheter tip position would not be feasible due to equipment, manpower and radiation exposure considerations 12. Colour Doppler and M-mode ultrasound can be used to help identify the position of the catheter within the epidural space by making observations whilst the catheter is flushed with saline. In one recent study of surgical patients 35, the path of the catheter could be identified in 89%, and the catheter tip in up to 75%. However this retrospective study relied on advanced ultrasound techniques and one expert operator. Also the identification of the catheter tip required using single-orifice catheters. Electrical stimulation of the epidural space was first described by Tsui in 1998 36 and provides immediate feedback regarding epidural catheter-tip location. The method, analogous to nerve stimulation for regional anaesthetic blocks, involves applying a low frequency signal via a stimulating epidural catheter that includes a conducting wire and a metallic tip. Increasing the current stepwise leads to muscle contractions in limb or trunk myotomes, typically between 1 and 10 mA for correct epidural placement. It also has the benefit of differentiating between epidural and intrathecal catheter placement, as a catheter in the CSF will cause stimulation at a current below 1 mA 37. Despite an early study in labouring women demonstrating 100% specificity and sensitivity at locating the epidural space in 39 patients 38, the technique has not been widely investigated in the obstetric literature. One study in a surgical population that compared electrical stimulation with pressure waveform analysis to confirm epidural placement, found that they were identical in terms of sensitivity (100%) and specificity (80%) 39. Waveform analysis is a method of transducing the pressure in the epidural space, which demonstrates beat-by-beat pressure waveforms (similar to invasive arterial pressure waveforms) from pulsations transmitted through the dura mater. These can be obtained by connecting a transducer to either the epidural needle or the catheter. This method is attractive in its simplicity and lack of specialised equipment, and is the method used by Al-Aamri et al. in their paper in this issue 7. Waveform analysis has been studied and been shown to be beneficial in surgical patients requiring thoracic epidurals. Prangmalee et al. enrolled 160 patients in whom waveform analysis was carried out in a similar fashion to Al-Aamri et al., before elective surgery. They found the technique was 91% sensitive at confirming correct epidural placement, although this necessitated including weak oscillations (that appear under-damped) as well as strong ones with clear pulsatile waveforms 40. Specificity was 84%. When the same group of investigators carried out a randomised comparison (n = 106) of conventional LOR vs. waveform analysis to confirm thoracic epidural placement, epidural failure rates were 24% for conventional technique and only 2% when using waveform analysis to confirm placement 41. Al-Aamri et al.'s paper 7 is only the second study to evaluate epidural pressure waveforms in labouring women. Sebbag et al. conducted a pilot study of 10 parturients in labour, but were unable to detect any pressure waveforms through the needle or via the epidural catheter, in spite of the epidurals providing adequate analgesia 42. This equates to 0% sensitivity (95%CI 0–31%), with no failed epidurals to allow a calculation of specificity. The authors concluded that the lack of pressure waveforms was attributable to physiological differences in the lumbar epidural space of parturients such as weaker arterial pulsation, increased tissue compliance and increased epidural pressure, hindering waveform transmission. Al-Aamri et al. argue that the lack of pressure waveforms was due to not flushing the needle with 5 ml of saline before transducing it. However, Sebbag and colleagues flushed 2 ml through the needle (following loss-of-resistance to saline), and 5 ml through the catheter before taking their measurements (Sebbag I, personal communication); which compares with 3–4 ml flushed down the needle when saline was used for LOR (once the 1–2 ml used for the LOR is accounted for) in the Al-Aamri study. Whilst previous studies have shown slightly better epidural pressure waveforms with 5 ml rather than 2.5 ml flushed through a catheter, not flushing the catheter at all still demonstrated pressure waveforms in the majority of thoracic epidurals 43. Defining ‘epidural failure’ is a grey area, and numerous definitions have been used, ranging from insufficient analgesia to any reason for early discontinuation of epidural analgesia 4. A study in 2013 by Thangamuthu et al. 3 set out to agree a standardised definition of epidural failure using the Delphi method, and then apply this to a cohort of 1521 parturients. They reached a definition of ‘epidural failure’ as one or more of the following: lack of adequate pain relief by 45 min from the start of epidural placement; dural puncture; re-siting or abandoning the epidural; and maternal dissatisfaction with analgesia at the follow up visit. The most common failure criterion in their cohort was inadequate pain relief (64%), with maternal dissatisfaction with analgesia contributing 21%. Had the investigators in the Al-Aamri paper used analgesia at 45 min as their criterion for an effective epidural, the study would have found a sensitivity and specificity of 95.9% and 100% respectively, with a positive predictive rate of 100% for epidural waveform analysis as a means of predicting epidural success. However, given the convenience sample number of 100, coupled with the surprisingly low epidural failure rate (3/100), this would still equate to a 33.3% negative predictive value. Whilst it's hard to infer much from such small numbers, it does follow a pattern seen in other studies: that the absence of epidural pulsations is not evidence of epidural failure. However, it seems that one can be reassured that if pressure waveforms are seen, then the epidural is likely to work. So where does this leave us on the coalface of delivery suite? The fact that pressure waveform analysis is a technique that uses familiar equipment that is widely accessible (i.e. an invasive pressure monitoring set), and may have a high positive predictive value, is very attractive. If you had spent the last 60 min struggling to insert an epidural, before getting a ‘soft’ loss-of-resistance, into someone with spinal abnormalities or significant cardio-respiratory disease; would it be tempting to check for pressure waveforms before abandoning? Probably yes. However, should you abandon this attempt if the pressure waveforms are absent? The evidence would suggest probably not. The role of epidural waveforms analysis in obstetrics is not yet fully elucidated. Its usefulness in elective surgery is clearer, where the confirmation of successful epidural placement before knife-to-skin has distinct advantages. Its advantages on delivery suite where the time, equipment and resources required are potential barriers to the rapid delivery of analgesia to a mother in extremis, are less obvious. Transposing a successful technique from a surgical setting, does not necessarily translate well into obstetrics. Particularly given that the technique as it has been evaluated in Al-Aamri et al.'s study, investigating the placement of the epidural needle and not the final position of the catheter tip, could falsely reassure. The search for the ideal technique to predict epidural success continues. Epidural stimulation shows promise, but requires specialist equipment, and needs further evaluation in the obstetric setting before it can be widely recommended. Epidural pressure waveforms may play a role, but have been inconsistent in labouring women so far, particularly with regards to the significance of an absent waveform. But before we blame the technology too harshly, let us consider the aphorism ‘when there are problems with any regional technique, look for the cause first on the proximal end of the needle’ 44. Our thanks to Mark Scrutton for reviewing the manuscript. No external funding or competing interests declared.