Secondary Logo

Journal Logo

Obstetric anaesthesia

Spinal ultrasound versus palpation for epidural catheter insertion in labour

A randomised controlled trial

Arzola, Cristian; Mikhael, Rafeek; Margarido, Clarita; Carvalho, Jose C.A.

Author Information
European Journal of Anaesthesiology (EJA): July 2015 - Volume 32 - Issue 7 - p 499-505
doi: 10.1097/EJA.0000000000000119
  • Free



In recent years, anaesthetists have progressively incorporated the use of ultrasound guidance to facilitate the performance of various procedures such as vascular access, peripheral nerve blockade and neuraxial anaesthesia in an attempt to improve their efficacy and safety.1,2 In 2008, the National Institute for Health and Clinical Excellence (NICE) in the United Kingdom issued guidance regarding the use of ultrasound to facilitate the catheterisation of the epidural space.3 Furthermore, a recent systematic review and meta-analysis on the use of ultrasound imaging to facilitate lumbar puncture and epidural catheterisation concluded that it may be a useful adjunct for these procedures, as it can reduce the risk of failed or traumatic lumbar punctures and epidural catheterisations, as well as the number of needle insertions and redirections required to complete the procedure.4

A common limitation in most of the trials published so far on the use of spinal ultrasound is the fact that a single expert investigator5–8 or a selected group of anaesthesists8,9 performed the ultrasound scanning, while an experienced or novice operator executed the neuraxial technique. The expertise of the ultrasound operator is an important factor when comparing ultrasound-assisted and palpation techniques. This important limitation may preclude the generalisability of the existing data, and it remains to be determined whether the same benefits will be observed when clinicians perform both the ultrasound scanning and the neuraxial procedure. In this study, we aimed to determine the impact of ultrasound usage on the ease of insertion of labour epidurals by a group of trainees after a comprehensive teaching programme in ultrasound assessment of the spine.

We hypothesised that the use of preprocedural spinal ultrasound would improve the ease of insertion of labour epidural catheters when compared with the conventional palpation technique.

Materials and methods

Ethical approval for this randomised, controlled trial was provided by the Research Ethics Board of Mount Sinai Hospital, Toronto, Canada (Chairperson Ronald Heslegrave, MD, PhD) in August 2009 (MSH REB #09-0173-E). Written informed consent was obtained from all the participants. The study was conducted at Mount Sinai Hospital, a University of Toronto affiliated teaching hospital and was registered at (NCT00996905). We followed the guidelines for reporting and checklist from CONSORT (Consolidated Standards of Reporting Trials).10

Eligible operators for the study were advanced-level trainees (anaesthesia fellows) and residents in their second year of training in anaesthesia who had no previous experience in the use of spinal ultrasound to facilitate neuraxial anaesthesia. Although second year residents have had previous exposure to neuraxial anaesthesia during their first year of residency, they receive their main training in epidural anaesthesia under close supervision during a 4-month rotation in their second year of training. Anaesthesia fellows, on the contrary, are fully trained anaesthetists pursuing further specialisation in our 1-year obstetric anaesthesia fellowship programme. The purpose of enrolling both residents and fellows was to study operators at different levels of clinical experience. Both groups received comprehensive spinal ultrasound training before patient recruitment.

The inclusion criteria for patients to be recruited into the study were full-term parturients with easily palpable lumbar spines requesting neuraxial labour analgesia. An easily palpable spine was defined as one wherein the spinous processes were easily palpable (as assessed by one of the study investigators), but not necessarily visible on inspection. Patients with contraindications to epidural anaesthesia and those with previous spinal surgery, trauma or obvious deformity were excluded.

Fellows and residents received didactic teaching in the form of reading material and an educational video on ultrasound-facilitated spinals and epidurals,11 which were required to be reviewed prior to a 45-min lecture on the topic. The training proceeded with an interactive teaching component consisting of a 15-min demonstration on a live model with typical sonoanatomy at all lumbar spaces,12 followed by a 2-h hands-on workshop in a small group setting. During the workshop, the trainees were instructed on how to perform a standardised spinal ultrasound assessment for neuraxial procedures. They also received one-to-one feedback on the technical aspects and skills required to perform the following ultrasound scanning tasks:

  1. identification of the sacrum and the five lumbar intervertebral spaces in the paramedian sagittal oblique plane (PSO); and
  2. determination of the midline and the interspace at different lumbar interspaces, the optimal epidural needle insertion point and the distance from the skin to the inner aspect of the ligamentum flavum-dura mater unit in the transverse median plane.

The standardised spinal ultrasound procedure was performed with a portable ultrasound system equipped with a 5 to 2 MHz curved array probe (M-Turbo; SonoSite Canada Inc., Markham, Ontario, Canada) and the individual in the sitting position. The final phase of the training comprised the use of ultrasound for the placement of 20 labour epidurals by the participating fellows and residents under the supervision of one of the study investigators. Such cases had to be recorded in a logbook and reviewed by one of the investigators who had 7 years of experience in spinal ultrasound (C.A., J.C.). Participants were deemed to be competent only when the aforementioned ultrasound scanning tasks (determination of the insertion point and estimated distance to the epidural space) were performed independently and accurately.

The actual trial was initiated only when the training was completed. Eligible patients requesting labour epidurals were recruited and informed consent was obtained. The trainees were randomly allocated to the procedures without knowing the patient characteristics in advance. The allocation was either to the intervention group (use of preprocedural spinal ultrasound, ultrasound group) or to the control group (conventional palpation technique, palpation group). An equal number of procedures were planned in the intervention and in the control groups for each trainee.

The preprocedural spinal ultrasound in the ultrasound group was performed in a nonsterile manner. Thereafter, all labour epidural procedures were carried out with the patient in the sitting position and under sterile conditions. Following either palpation or ultrasound examination, the epidural space was located with a 17-gauge, 8.9 cm Tuohy epidural needle with markings at 1-cm intervals, using a loss of resistance (LOR) to air or saline method with a midline approach. Subsequently, a 19-gauge multiport, wire-embedded epidural catheter (Arrow FlexTip Plus; Arrow International Inc., Reading, Pa, USA) was inserted approximately 5 cm into the epidural space. The medications used for the procedure included skin infiltration with lidocaine 2% (2 to 3 ml) at the puncture site; after aspiration of the catheter to exclude intrathecal or intravascular placement, a test dose of lidocaine 2% (3 ml), followed 5 min later by bupivacaine 0.125% (10 ml) with fentanyl 50 μg as a loading dose; and finally a patient-controlled epidural infusion was instituted consisting of bupivacaine 0.0625% with fentanyl 2 μg ml−1 (infusion rate 10 ml h−1, on-demand bolus 5 ml with lockout time 10 min, total hourly limit 20 ml).

The primary outcome of this study was the ease of insertion of labour epidurals, which comprised the following: duration of the epidural procedure defined as the time elapsed from the initial Tuohy needle insertion through the skin to the completion of the threading of the epidural catheter; number of interspace levels at which the insertion was attempted; and number of needle passes (defined as any forward needle advancement following a backward movement in order to redirect the needle). The secondary outcomes included total duration of the procedure: time required to complete both the preprocedural spinal ultrasound or the assessment by palpation and the epidural procedure; success rate at the first needle pass; number of attempts to thread the catheter; need to call a staff anaesthetist for help; failure of epidural anaesthesia defined as the need to reinsert a new epidural catheter due to lack of sufficient analgesia within 2 h of the primary insertion; incidence of inadvertent dural puncture and postdural puncture headache; patient satisfaction as determined by a questionnaire completed on the first postpartum day (5-point scale for each item): perception of the length of the epidural procedure, discomfort during the epidural placement, satisfaction with the quality of labour epidural analgesia and development of back pain postpartum at the site of epidural insertion (Appendix 1).

For allocation of the epidural procedures, a statistician created a computer-generated list of random numbers for each operator (fellow or resident), resulting in 10 envelopes per operator in a 1 : 1 group allocation using block sizes of two and four (five procedures with preprocedural spinal ultrasound and five without). The allocation sequence was concealed in sequentially numbered, opaque, sealed envelopes, which were opened by an independent observer only after the patient's eligibility was assessed and written informed consent was obtained. After the intervention assignment, parturients were not amenable to be blinded but were requested not to disclose whether or not they underwent preprocedural spinal ultrasound. A research assistant, blinded to the group allocation, collected the outcome data during the epidural procedure.

The sample size calculation was based on data from a pilot study, which assessed labour epidural insertions in women with easily palpable spines. The mean ± SD duration of the procedure was 201 ± 107 s. We decided to test the alternate hypothesis of reducing the mean duration by 60 s (30% reduction) to 140 s with similar standard deviation. Considering a type I error of 0.05, type II error of 0.2 and a two-tailed test, we required 96 patients. After considering 30% attrition, we planned to recruit 128 patients. The assumption of normal distribution was assessed by the Shapiro–Wilk test for normality. The Student's t-test or Wilcoxon Rank Sum test and Pearson's Chi-square or Fisher's exact test were used as appropriate to analyse the continuous and categorical variables, respectively. Furthermore, we performed a multivariable analysis to analyse the duration of the procedures using a mixed-effect model accounting for correlated data and adjusted for operator effect (fellow or resident). We also planned to perform subgroup analysis for residents and fellows in procedural times. Descriptive statistics were calculated using mean ± SD, mean (confidence interval, CI) and median (IQR) for continuous data, and percentages for discrete variables.


The study was carried out between October 2009 and February 2011. We enrolled 17 residents and five fellows who completed the spinal ultrasound training. During the study period, we randomised 128 parturients requesting labour epidurals (Fig. 1). We analysed 84 epidurals performed by residents and 44 epidurals performed by fellows. The comparison of patient characteristics is presented in Table 1 and the procedural times in Fig. 2.

Fig. 1
Fig. 1:
CONSORT flow diagram of the progress through the phases of the trial.
Table 1
Table 1:
Patient characteristics in the palpation and ultrasound groups
Fig. 2
Fig. 2:
Epidural procedural time and total procedural time (palpation or ultrasound assessment plus epidural procedural time) in the palpation and ultrasound groups. Horizontal black lines represent the median, the height of the boxes the interquartile range (IQR), the whiskers the range and the circle an outlier (outside 1.5 IQR).

There was no difference in the median (IQR) epidural insertion time between the ultrasound and palpation groups [174 (120 to 241) versus 180 (130 to 322.5) s, respectively; P = 0.14]. Furthermore, there were no significant differences even after adjustment for the trainees’ level [mean (95% CI) palpation group 171.4 (146.0 to 201.3) s versus ultrasound group 192.8 (167.0 to 225.2) s P = 0.25]. Although the preprocedural assessment in the ultrasound group was clinically short [median (IQR) 100 (70 to 135) s], the total duration of the procedure (ultrasound assessment and epidural insertion) was longer in the ultrasound group [median (IQR) 261 (217.5 to 420.5) versus 180 (130 to 322.5) s; P < 0.001].

In a subgroup analysis considering the trainees’ level of clinical experience, the residents showed a trend towards a shorter epidural procedural time when performing the procedure after ultrasound assessment [mean (95% CI) ultrasound group 199.3 (165.7 to 239.8) s versus palpation group 247.6 (208.5 to 292.9) s; P = 0.06]. The total duration of the procedures also showed a trend to be prolonged when residents used ultrasound as compared with palpation [mean (95% CI) ultrasound group 305.5 (255.2 to 365.8) s versus palpation group 247.6 (208.5 to 292.9) s; P = 0.06].

Other epidural outcomes showed no significant differences between the palpation and ultrasound group, such as the first-pass success rate (50 versus 60%, respectively; P = 0.26); the number of interspace levels attempted (need to attempt a second interspace level: 4/68 versus 2/60, respectively; P = 0.68); the number of needle passes at the first level [median (IQR) 1.5 (1 to 3) versus 1 (1 to 2), respectively; P = 0.43]; and the number of catheter threading attempts at the first level [median (IQR) 1 (1 to 1) versus 1 (1 to 1), respectively; P = 0.93].

The staff anaesthesiologist was called for help in one case per group. There were no failures in epidural labour analgesia, but there was one inadvertent dural puncture by a fellow in the ultrasound group. The patient satisfaction questionnaire revealed no significant differences between groups, with most patients satisfied with the procedure (Fig. 3, Appendix 1).

Fig. 3
Fig. 3:
Patient Satisfaction Questionnaire (5-point scale, see Appendix 1). Perception of the length of the epidural procedure, discomfort during the epidural placement, satisfaction with the quality of labour epidural analgesia and development of postpartum back pain at the site of epidural insertion. Grey boxes represent the ultrasound group, white boxes the palpation group, horizontal black lines represent the median, the height of the boxes the interquartile range (IQR), the whiskers the range and the circle an outlier (outside 1.5 IQR).


Our study showed that preprocedural ultrasound spinal assessment by a cohort of trainees specifically trained in spinal ultrasound assessment, in women with easily palpable lumbar spines, does not improve the ease of insertion of labour epidural catheters. We observed no improvement in time to perform the epidural catheter insertion, number of interspace levels attempted and number of needle passes, as compared with when the insertion was performed using the standard palpation technique.

Our study is in contrast with most studies published so far that have been summarised in a recent meta-analysis.4 The most important difference between our study and all previous studies is that we trained a group of trainees that performed both the ultrasound assessment and the epidural placements, while in the other studies, a selected number of experts would have performed both the ultrasound assessment and the epidural placement, or the information from the ultrasound assessment would have been provided to trainees who then performed the epidural.4,7,13

We have previously demonstrated that the transfer of knowledge for this particular skill is challenging.18 It may be claimed that with more ultrasound experience in our group of trainees, our results could have been improved and resembled the results from experts. This is indeed a possibility. Despite careful and comprehensive spinal ultrasound training in our study, we did not perform a formal assessment of the newly acquired ultrasound skills.

The acquisition of skills in ultrasound assessment of the spine seems to be one of the impediments to a more widespread adoption of the technique. Anaesthetists have not embraced the use of preprocedural ultrasound for neuraxial blocks to same degree as they have for peripheral nerve blocks. This has been in part related to the perceived more challenging imaging of deeply located structures surrounded by a complex bony encasement, and the difficulty of performing procedures under real-time imaging. Some reviews initially suggested that it would be premature to make recommendations for practice based on the existing limited data.11,14,15 The NICE guidance published in 2008, however, while acknowledging that evidence was still limited, already suggested that ultrasound imaging was well tolerated and helpful in achieving correct epidural placement. Arguments from detractors included difficulty in training, equipment availability and increased procedural time.16,17 Cumulated evidence has only recently become available showing that ultrasound imaging promotes a clinically and statistically significant effect in reducing failed and traumatic procedures.4

We intentionally studied patients with easily palpable lumbar spines. Although this choice has made the cohort of women receiving epidural anaesthesia more uniform, it may be argued that more challenging patients could have demonstrated similar benefits to those seen in previous studies.5,6,8

Although we did observe a trend towards a faster epidural catheter insertion within the group of residents when performing the procedure after ultrasound assessment, the total duration of the procedures (ultrasound assessment as well as epidural insertion) showed a trend towards a longer time period. Although this finding may be hypothesis generating, further research is warranted. The knowledge of the estimated distance to the epidural space may have expedited the LOR technique at a stage when trainees are still consolidating their skills. This hypothesis may be in agreement with the findings by Vallejo et al.7 who demonstrated that residents had a decreased failed rate of labour epidurals when they were provided with insertion point, estimated depth to the epidural space and needle direction from a preprocedural ultrasound examination.

Our study has some limitations. First, although some of our results suggest that the total procedural time was not prolonged within the residents group when using ultrasound, these must be interpreted cautiously in the context of a subgroup analysis. A larger sample size might confirm that the use of ultrasound does not prolong the total procedural time while revealing a shorter epidural insertion time. Second, despite the patient population being homogenous and comparable between the intervention and control groups, the operators were a mixed sample of trainees at different levels (fellows and residents) adding further variability to the study despite subgroup analysis. Finally, our patient satisfaction questionnaire is a tool that has not been previously validated.

In conclusion, our study results showed that preprocedural spinal ultrasound assessment performed by a cohort of anaesthesia trainees did not improve the ease of insertion of labour epidural catheters in patients with easily palpable lumbar spines, as compared with the traditional palpation technique based on anatomical landmarks.

Acknowledgements relating to this article

Assistance with the study: the authors thank Kristi Downey, Perinatal Research Coordinator, Department of Anesthesia and Pain Management, Mount Sinai Hospital, for her invaluable contribution to all stages of this project.

Financial support and sponsorship: this work was supported solely from departmental sources. No other financial support or sponsorship was obtained.

Conflict of interests: none.

Presentation: preliminary data for this study was presented as a poster at the 2011 Annual Meeting of the Canadian Anesthesiologists’ Society (24 to 28 June 2011, Toronto, Ontario, Canada) and at the 2011 Annual Meeting of the Society for Obstetric Anesthesia and Perinatology (13–17 April 2011, Henderson, Nevada, USA).


1. Marhofer P, Willschke H, Kettner S. Current concepts and future trends in ultrasound-guided regional anesthesia. Curr Opin Anaesthesiol 2010; 23:632–636.
2. Ecimovic P, Loughrey JPR. Ultrasound in obstetric anaesthesia: a review of current applications. Int J Obstet Anesth 2010; 19:320–326.
3. National Institute for Health and Clinical Excellence. Ultrasound guided catheterisation of the epidural space: understanding NICE guidance. January 2008. [Accessed February 20, 2013]
4. Shaikh F, Brzezinski J, Alexander S, et al. Ultrasound imaging for lumbar punctures and epidural catheterisations: systematic review and meta-analysis. BMJ 2013; 346:f1720–f1730.
5. Grau T, Leipold RW, Conradi R, Martin E. Ultrasound control for presumed difficult epidural puncture. Acta Anaesthesiol Scand 2001; 45:766–771.
6. Grau T, Leipold RW, Conradi R, et al. Efficacy of ultrasound imaging in obstetric epidural anesthesia. J Clin Anesth 2002; 14:169–175.
7. Vallejo MC, Phelps AL, Singh S, et al. Ultrasound decreases the failed labor epidural rate in resident trainees. Int J Obstet Anesth 2010; 19:373–378.
8. Chin KJ, Perlas A, Chan V, et al. Ultrasound imaging facilitates spinal anesthesia in adults with difficult surface anatomic landmarks. Anesthesiology 2011; 115:94–101.
9. Chin KJ, Perlas A. Ultrasonography of the lumbar spine for neuraxial and lumbar plexus blocks. Curr Opin Anaesthesiol 2011; 24:567–572.
10. Moher D, Hopewell S, Schulz KF, et al. CONSORT 2010 Explanation and Elaboration: updated guidelines for reporting parallel group randomised trials. J Clin Epidemiol 2010; 63:e1–e37.
11. Carvalho JCA. Ultrasound-facilitated epidurals and spinals in obstetrics. Anesthesiol Clin 2008; 26:145–158.
12. Borges BCR, Wieczorek P, Balki M, Carvalho JCA. Sonoanatomy of the lumbar spine of pregnant women at term. Reg Anesth Pain Med 2009; 34:581–585.
13. Grau T, Bartusseck E, Conradi R, et al. Ultrasound imaging improves learning curves in obstetric epidural anesthesia: a preliminary study. Can J Anaesth 2003; 50:1047–1050.
14. Perlas A. Evidence for the use of ultrasound in neuraxial blocks. Reg Anesth Pain Med 2010; 35:S43–S46.
15. Chin KJ, Karmakar MK, Peng P. Ultrasonography of the adult thoracic and lumbar spine for central neuraxial blockade. Anesthesiology 2011; 114:1459–1485.
16. Mathieu S, Dalgleish DJ. A survey of local opinion of NICE guidance on the use of ultrasound in the insertion of epidural catheters. Anaesthesia 2008; 63:1146–1147.
17. Gambling DR. Lumbar ultrasound: useful gadget or time-consuming gimmick? Int J Obstet Anesth 2011; 20:318–320.
18. Margarido CB, Arzola C, Balki M, Carvalho JCA. Anesthesiologists’ learning curves for ultrasound assessment of the lumbar spine. Can J Anaesth 2010; 57:120–126.
© 2015 European Society of Anaesthesiology