Ultrasound-Assisted Versus Landmark-Guided Spinal Anesthesia in Patients With Abnormal Spinal Anatomy: A Randomized Controlled Trial : Anesthesia & Analgesia

Secondary Logo

Journal Logo

Original Research Articles: Original Clinical Research Report

Ultrasound-Assisted Versus Landmark-Guided Spinal Anesthesia in Patients With Abnormal Spinal Anatomy: A Randomized Controlled Trial

Park, Sun-Kyung MD; Bae, Jinyoung MD; Yoo, Seokha MD; Kim, Won Ho MD, PhD; Lim, Young-Jin MD, PhD; Bahk, Jae-Hyon MD, PhD; Kim, Jin-Tae MD, PhD

Author Information
Anesthesia & Analgesia 130(3):p 787-795, March 2020. | DOI: 10.1213/ANE.0000000000004600
  • Free
  • SDC
  • CME Test
  • Continuing Medical Education



  • Question: Does neuraxial ultrasound facilitate the efficacy of spinal anesthesia in patients with anatomical alterations in the lumbar spine?
  • Findings: The use of ultrasound for spinal anesthesia reduced the number of needle passes required for block success, improved the first attempt success rate, and improved periprocedural pain scores without significantly prolonging total procedure time compared with the landmark-guided technique in patients with abnormal spinal anatomy.
  • Meaning: The use of neuraxial ultrasonography is associated with the improved technical performance of spinal anesthesia in patients with anatomical alterations in the lumbar spine.

Spinal anesthesia has traditionally been performed with a surface landmark–guided technique; however, it can be challenging in patients with anatomical alterations of the lumbar spine, such as scoliosis or previous lumbar spinal surgery.1–4 Patients with scoliosis pose a unique challenge for anesthesia providers because these patients not only have lateral curvatures of the spine but also rotational changes of the vertebral bodies.5,6 In addition, patients who have previously undergone surgery of the lumbar spine can have technical difficulties because surface landmarks may be indistinct or even absent in these patients.7

Minimizing the number of needle manipulations during spinal anesthesia is desirable because multiple insertion attempts are associated with several complications, including epidural/spinal hematoma, neural damage, postdural puncture headache, and infection, as well as patient pain, discomfort, and dissatisfaction.4,5,7–11 A growing body of evidence indicates that ultrasonography of the spine can facilitate the technical performance of spinal anesthesia.1,12–14 Nevertheless, widespread application of neuraxial ultrasound remains limited,13 because the surface landmark–guided technique has been widely considered an efficient approach, especially in expert hands, and safe due to the extremely low rate of complications.15,16 Therefore, there is still a need to define the role of ultrasonography in individual subpopulation settings.13

To date, no randomized trial has compared the efficacy of spinal anesthesia between the ultrasound-assisted technique and landmark-guided technique primarily in patients with anatomical alterations in their lumbar spine. A previous trial reported that the technical difficulty of spinal anesthesia in patients with difficult surface landmarks was significantly decreased with the aid of ultrasonography.7 However, most of the study subjects were obese without abnormalities in their lumbar spine.7 Thus, the utility of ultrasonography in patients with abnormal spinal anatomy remains unclear.7

Therefore, the aim of this study was to determine whether an ultrasound-assisted technique could reduce the number of needle passes required for successful dural puncture in patients with abnormal spinal anatomy compared with the conventional landmark–guided technique. In the course of study, we also compared the procedure time, periprocedural pain and discomfort scores, and the incidence of complications related to spinal anesthesia during the procedure between the 2 techniques in these patient populations.


Study Design

This prospective, randomized, controlled study was approved by the institutional review board of Seoul National University Hospital (No.: 1801-107-917; Chairperson Prof B. J. Park; date of approval, February 27, 2018) and was registered with ClinicalTrials.gov (NCT03459105; principal investigator: Jin-Tae Kim; date of registration, February 27, 2018). We have prepared this article in accordance with the Consolidated Standards of Reporting Trials guidelines.17 The study was conducted between March 2018 and July 2018 at Seoul National University Hospital, Seoul, Korea. All included participants provided written informed consent.

Patient Population

Adult patients with American Society of Anesthesiologists physical status I/II/III scheduled to undergo elective orthopedic surgery under spinal anesthesia were considered for eligibility if they had anatomical abnormalities in their lumbar spine, defined as one of the following: (1) documented mild to severe lumbar scoliosis in preoperative lumbosacral spine x-ray, defined as a Cobb angle ≥10°; and (2) history of lumbar spinal surgery involving L2–L5 vertebrae. Patients with contraindications to spinal anesthesia including allergy to local anesthetics, coagulopathy, or local infection at the puncture site were excluded from the study, as were those unwilling to participate or unable to communicate.

Randomization and Minimization of Bias

Patients were randomly assigned to receive spinal anesthesia using the surface landmark–guided (landmark group) or preprocedural ultrasound–assisted (ultrasound group) technique using a computer-generated table of random numbers. Group allocation was concealed by sequentially numbered, sealed opaque envelopes, which were opened only by the attending anesthesiologist immediately before the procedure.

All spinal procedures were performed by 1 of 3 attending anesthesiologists (S.-K.P., S.Y., or J.-T.K.). All of the 3 performers were experienced regional anesthesiologists who skilled with the landmark-guided neuraxial technique, and each having performed more than 30 ultrasound-assisted neuraxial blocks before this study. Block randomization was performed to balance the allocation of each of the 3 anesthesiologists to the 2 intervention arms.

Study Interventions

In both groups, the anesthesiologist carefully reviewed the patient’s history and previous radiologic examinations, and understood the patient’s anatomy before commencing the spinal procedure (Supplemental Digital Content 1, Figure 1, https://links.lww.com/AA/C990). Spinal anesthesia was administered with patients placed in the sitting or lateral decubitus position. Sedatives were not administered before or during the administration of spinal anesthesia.

In the ultrasound group, skin marking was done before the spinal procedure using the 4C-RS curved array probe (frequency, 2.0–5.5 MHz) of a portable ultrasound system (Vivid-i Ultrasound System, GE Healthcare, Little Chalfont, United Kingdom) and a skin marker (Devon, Covidien, Dublin, Ireland). Paramedian sagittal oblique (PSO) and transverse median (TM) images were obtained as previously described.18–21 The preprocedural ultrasound–assisted spinal anesthesia was administered using a paramedian approach, according to the following systematic protocol.22

  1. In the TM view, each spinous process tip was marked on the skin (Figure 1A). The midline was drawn by connecting spinous process tips. The transverse interlaminar views (Figure 1B) were obtained by visualizing the anterior and posterior complexes.18
  2. The probe was rotated to obtain a PSO view approximately 1–2 cm lateral to the midline (Figure 1C). This view was obtained on the convex side of the scoliotic curve from the midline in patients with scoliosis. The sacrum was identified first and subsequently, the transducer was moved cephalad to identify individual interlaminar spaces from L5–S1 to L2–3.18
  3. The interlaminar space, showing the posterior and anterior complexes18 as clearly as possible, was centered on the ultrasound screen (Figure 1D). With the probe positioned to obtain the clearest image, the skin was marked at the midpoints of the probe (Figure 1C). This midpoint of the probe indicated the paramedian needle-insertion point (Figure 2A).
  4. The intervertebral level that provided the largest interlaminar space was selected as the site for the first attempt, and the skin was carefully marked at this level. The skin was additionally marked at other identified intervertebral levels in preparation, and these sites were used only if the first attempt was not successful.
  5. The medial angle of the probe providing the clearest image in the PSO view was estimated and used as the angle of spinal needle insertion. The needle was directed toward the observed angle off the sagittal plane, and there was no cephalad angulation in the first attempt (Figure 2B; Supplemental Digital Content 2, Video 1, https://links.lww.com/AA/C991).
  6. The required depth of needle insertion was estimated based on the distance from the skin to the posterior complex.
  7. The PSO and TM images were graded as good (both posterior and anterior complexes visible), intermediate (posterior or anterior complex visible), or poor (neither complex visible).21,23
Figure 1.:
Ultrasonographic views of the lumbar spine and corresponding skin markings for spinal anesthesia. A, Probe position and skin markings in transverse median view. B, Transverse median view of the lumbar spine. C, Probe position and skin markings in paramedian sagittal oblique view. D, Paramedian sagittal oblique view of the lumbar spine. AC indicates anterior complex; PC, posterior complex.
Figure 2.:
Preprocedural skin markings and spinal needle insertion. A, The potential needle insertion points (white arrows) and neuraxial midline (black arrows) after the completion of the skin markings. B, Paramedian insertion of a spinal needle using preprocedural ultrasound skin markings.

After the skin marking, the ultrasound gel was removed to ensure the needle-insertion site was free of gel. In the ultrasound group, the anesthesiologist did not palpate the surface anatomic landmarks until completion of spinal injection. The distance from the skin to the posterior complex was measured using an ultrasound caliper tool after the completion of spinal procedure because the accurate measurement of it was required not for the performance of spinal anesthesia but for the assessments of the study outcomes. Sonographic images were graded after the completion of the procedure for the same reason. Hence, the duration of time required to measure the depth and grade the image quality was not included in the procedural time in this study.

In the landmark group, spinal anesthesia was administered on the basis of direct palpation of surface anatomic landmarks, and the anesthesiologist was given the discretion to use either a midline or paramedian approach. In the case of patients with scoliosis, a systematic approach was used according to previous recommendations for spinal anesthesia of the scoliotic spine.5 Namely, the type and severity of scoliosis were determined from previous radiologic examinations before commencing the neuraxial procedure, and good patient positioning was encouraged for mild scoliosis. In moderate scoliosis, anesthesia provider considered either a paramedian approach on the convex side of the scoliotic curve or a midline approach with angulation toward the convex side when indicated.5 After completion of the injection, the anesthesiologist reported the grade of the ease of palpation on a 4-point scale (easy, moderate, difficult, or impossible).7,21 Then, the intervertebral level at which the injection was administered was identified with ultrasonography.

In both groups, strict aseptic techniques were used throughout the procedure. Spinal anesthesia was administered with a 25-gauge Quincke bevel spinal needle (TaeChang Industrial Co, Gongju, Korea). After confirming the backflow of clear cerebrospinal fluid, 0.5% hyperbaric bupivacaine was injected intrathecally, with the dose determined at the discretion of the anesthesiologist. No intrathecal adjuvant was used. If dural puncture was not achieved after 5 attempts of separate needle insertion, the operator had options to use alternative techniques (ultrasound group: midline approach/landmark palpation; landmark group: ultrasound-assisted technique). In ultrasound group, if no visible structure was found at all from L5–S1 to L3–4 interspaces, the operator had options to use a landmark-guided technique or to convert into general anesthesia.


The primary outcome was the number of needle passes required to achieve successful dural puncture (defined as the number of forward advancements of the needle, ie, the number of needle redirections without exiting the skin, including the first pass).

Secondary outcomes included the following:

  1. Number of insertion attempts, defined as the number of any separate skin puncture by a needle.
  2. Time taken to identify landmarks (time from the first placement of the probe on the skin to the operator’s declaration of completion of skin marking in ultrasound group; time from the first touch for palpation to completion of palpation in landmark group).
  3. Time taken to administer spinal anesthesia, defined as time from needle insertion to either completion of intrathecal anesthetic administration by the allocated technique or the anesthesiologist’s declaration to use an alternative technique.
  4. Total procedure time, defined as the sum of time taken to identify landmarks and time taken to administer spinal anesthesia.
  5. Incidence of radicular pain, paresthesia, and blood tap in the spinal needle during the neuraxial procedure.
  6. Periprocedural pain score on an 11-point numeric rating scale (0, no pain; 10, worst pain imaginable). Patients were specifically asked to rate the pain in their back during administration of spinal anesthesia, after completion of the spinal procedure.
  7. Periprocedural discomfort score on a 11-point numeric rating scale (0, no discomfort; 10, worst discomfort imaginable). Patients were specifically asked to rate discomfort during the overall neuraxial procedure, including any uncomfortable feeling due to positioning, anxiety, or fear as a whole, after rating their pain scores.
  8. Level of sensory block tested by loss of cold sensation.

The outcomes were evaluated by 2 independent outcome assessors. Due to the presence of the skin markings in the ultrasound group, neither the intraprocedural outcome assessor nor operator could be blinded to group allocation. However, the postprocedure outcome assessor, who measured the level of sensory block and periprocedural pain/discomfort scores, entered the operating room only after the spinal procedure was completed and was blinded to group allocation. The numbers of passes/attempts were recorded until either completion of spinal anesthesia by the allocated method or decision to use an alternative method.

Statistical Analysis

Data were analyzed with SPSS Statistics for Windows (version 23.0, IBM Corp, Armonk, NY) and R software (version 3.1.0, R Foundation for Statistical Computing, Vienna, Austria). All data were analyzed on an intention-to-treat basis. Continuous data were tested for normality using Q–Q plots. Data showing normal distribution were presented as mean (standard deviation [SD]), and data showing a nonnormal distribution were presented as median (interquartile range [range]). These outcomes were analyzed using Student t test or Mann-Whitney U test for intergroup differences, as appropriate. Intergroup differences in the number of needle passes, the primary outcome, were assessed for significance using Mann-Whitney U test. Categorical data were expressed as number (proportion) and assessed using the Pearson χ2 test. Categorical data were analyzed using Fisher exact test when the expected counts were <5 for at least 25% of the cells. The relative risk of binary variables was presented with 95% confidence interval (CI). For the variables including the number of passes/attempts, time variables, and pain/discomfort scores, 10000 bootstrap samples were taken, and CIs for the difference in medians were calculated using the bootstrap method (sampling with replacement). Subgroup analyses included the types of anatomical abnormalities (scoliosis or previous spinal surgery). Two-tailed P values <.05 were considered statistically significant.

Sample Size

The sample size was calculated based on previous data obtained in elderly patients at our institution.22 We assumed the mean number of passes of the landmark-guided technique to be 5.7 (SD 5.4).22 We hypothesized that the use of ultrasound could decrease the number of passes to 1.9 (SD 1.8).22 To achieve a power of 0.8 and an α error of <.05, 20 patients were required in each group. To allow for dropout, 22 patients were randomized to each group. A priori calculation of sample size was based on a 2-sample t test, while the primary outcome was assessed using Mann-Whitney U test because of the observed distribution of the data.


Of 83 patients who were assessed for eligibility, 44 patients were randomized and completed the study (Figure 3). The baseline characteristics were comparable between the 2 groups except for the type of anatomical abnormalities (Table 1). Despite the random allocations, there were more patients with previous surgery in the landmark group (8 patients) than in the ultrasound group (2 patients). The patients with previous surgery (n = 10) had received posterior decompression or lumbar interbody fusions involving 2 or more vertebrae from L3 to S1. Among the patients with scoliosis (n = 34), the mean Cobb angle was 14.4° (SD 7.0°). Thirty-one patients were classified as mild (Cobb angle 11°−25°) and 3 patients had moderate (Cobb angle 25°−50°) scoliosis.

Table 1. - Demographic and Surgical Characteristics
Landmark Group (n = 22) Ultrasound Group (n = 22)
Age (y) 66.5 (13.2) 70.5 (8.8)
Height (cm) 157.4 (8.1) 155.2 (7.6)
Weight (kg) 64.2 (9.2) 62.9 (8.9)
BMI (kg/m2) 25.9 (2.9) 26.1 (3.2)
Sex, female 17 (77.3%) 20 (90.9%)
Abnormalities of the lumbar spine
 Previous spinal surgery 8 (36.4%) 2 (9.1%)
 Scoliosis 14 (63.6%) 20 (90.9%)
  Cobb angle (°)a 17.5 (9.6) 12.3 (3.6)
  Moderate/severe scoliosis 3 (13.6%)/0 (0%) 0 (0%)/0 (0%)
Type of surgery
 Total hip replacement arthroplasty 5 (22.7%) 5 (22.7%)
 Total knee replacement arthroplasty 13 (59.1%) 15 (68.2%)
 Ankle or foot surgery 4 (18.2%) 2 (9.1%)
Duration of surgery (min) 96.1 (27.5) 91.1 (22.7)
Values are mean (standard deviation) or number (proportion).
Abbreviation: BMI, body mass index.
aData from 14 patients in the landmark group and 20 patients in the ultrasound group were analyzed. Moderate scoliosis, Cobb angle 25°–50°; Severe scoliosis, Cobb angle >50°.

Figure 3.:
CONSORT flow diagram of patient recruitment. CONSORT indicates Consolidated Standards of Reporting Trials.

The median number of passes (1.5 vs 6) and needle-insertion attempts (1 vs 2) required to achieve dural puncture were both significantly lower in the ultrasound group than in the landmark group (both, P < .001; Table 2). The rate of successful dural puncture at first pass in the ultrasound group (50.0%) was higher than that in the landmark group (9.1%; P = .007; relative risk 5.5; Table 2). For all patients in the ultrasound group, dural puncture was achieved within 2 needle-insertion attempts. There was no evidence of interprovider differences in number of needle manipulations (Supplemental Digital Content 1, Table 1, https://links.lww.com/AA/C990).

Table 2. - Efficacy Outcomes of Spinal Anesthesia and Periprocedural Pain/Discomfort Scores
Landmark Group (n = 22) Ultrasound Group (n = 22) P Relative Risk or Difference in Medians (95% CI)
Number of passes 6 (2–9.3 [1–15]) 1.5 (1–3 [1–5]) <.001 4.5 (1–8)
Number of attempts 2 (1–4 [1–5]) 1 (1–1 [1–2]) <.001 1 (0–2)
Successful dural puncture at the first pass 2 (9.1%) 11 (50.0%) .007 5.5 (1.4–22.0)
Successful dural puncture within 2 passes 6 (27.3%) 15 (68.2%) .007 2.5 (1.2–5.2)
Successful dural puncture at the first attempt 9 (40.9%) 20 (90.9%) .001 2.2 (1.3–3.7)
Successful dural puncture within 2 attempts 12 (54.5%) 22 (100%) .001 1.8 (1.2–2.7)
Identifying time (s) 34 (26–49 [18–76]) 95 (83–126 [30–305]) <.001 −61 (−83 to −49)
Performing time (s) 118 (48–268 [25–362]) 38 (30–50 [25–151]) <.001 81 (14–175)
Total procedure time (s) 146 (90–295 [53–404]) 141 (115–181 [101–336]) .888 5 (−55 to 100)
Periprocedural pain score (NRS) 5.5 (3–8 [0–9]) 3.5 (1–5 [0–7]) .012 2 (−0.5 to 5)
Periprocedural patient discomfort score (NRS) 4 (2–6.3 [0–9]) 3 (1–5 [0–6]) .114 1 (−2 to 3.5)
Values are median (IQR [range]) or number (proportion). Identifying time, time taken for identifying the landmarks by palpation or ultrasound scan; performing time, time required for performing spinal anesthesia using the allocated method (time to completion of injection or declaration to use alternative methods, and alternative technique was used in 2 patients in the landmark group); total procedure time, the sum of the identifying time, and the performing time. P values are the results of the Mann-Whitney U test for continuous variables and χ2 test or Fisher exact test for incidence variables between the groups.
Abbreviations: CI, confidence interval; NRS, numeric rating scale.

Compared with palpation, ultrasonography required significantly more time to establish landmarks (Table 2). However, this difference was offset by the shorter time required for administering spinal anesthesia in the ultrasound group than in the landmark group. As a result, total procedure time did not differ significantly between the 2 groups (Table 2). The periprocedural pain scores were significantly lower in the ultrasound group than in the landmark group. However, the periprocedural discomfort scores showed no significant difference between the groups (Table 2).

In the landmark group, 2 patients required use of an alternative technique (ultrasound-assisted technique) for successful dural puncture (Supplemental Digital Content 1, Table 2, https://links.lww.com/AA/C990). None of the patients in the ultrasound group required use of an alternative technique. There were 4 patients in the landmark group and 2 in the ultrasound group who experienced periprocedural complications (radicular pain, paresthesia, or blood tap in the spinal needle) (Supplemental Digital Content 1, Table 2, https://links.lww.com/AA/C990). These patients were followed up for 24 hours after surgery and none exhibited persistent symptoms. No patients in either group required general anesthesia, and every spinal block was adequate for the entire surgery.

Intergroup differences in the intervertebral level of anesthesia administration are presented in Table 3 (P = .080). Dural puncture was conducted at the L2–3 intervertebral level in a patient in the landmark group who had a history of posterior lumbar interbody fusion at L3–4–5–S1. No patient was administered a spinal block at the L2–3 interspace in the ultrasound group (Table 3).

Table 3. - Comparisons of Block Characteristics Between Groups
Landmark Group (n = 20)a Ultrasound Group (n = 22) P
Dose of intrathecal bupivacaine (mg) 13 (13–13.8 [12–15]) 13 (12.8–13 [12–15]) .652
Paramedian approach 16 (80.0%) 22 (100.0%) .043
Distance from midline to paramedian needle insertion point (cm) 1.5 (1.5–1.8 [1.1–2.0])b 1.6 (1.4–1.7 [1.2–2.0]) .630
Depth of intrathecal space (cm) 4.6 (4.3–5.2 [4.0–6.2]) 4.5 (3.8–4.9 [2.8–6.0]) .164
Interspace level at which dural puncture was done .080
 L2–3 1 (5.0%) 0 (0%)
 L3–4 9 (45.0%) 3 (13.6%)
 L4–5 3 (15.0%) 5 (22.7%)
 L5–S1 7 (35.0%) 14 (63.6%)
Peak dermatome level .707
 T1–T3 7 (35.0%) 6 (27.3%)
 T4–T6 11 (55.0%) 12 (54.5%)
 T7–T10 2 (10.0%) 4 (18.2%)
Values are median (interquartile range [range]) or number (proportion). Distance from midline to paramedian needle insertion point, the distance between the midline of the neuraxis, and the paramedian needle insertion point. P values are the results of Mann-Whitney U test for continuous variables and χ2 test or Fisher exact test for incidence variables between the groups.
aData from only 20 patients in the landmark group were analyzed (the patients in whom the alternative technique was used were excluded).
bData from only 16 patients in whom the paramedian approach was used were analyzed.

In the landmark group, surface landmarks were easily palpated in 31.8% of patients (Supplemental Digital Content 1, Table 3, https://links.lww.com/AA/C990). Landmarks were moderate or difficult to palpate in 68.2% of the subjects, and no patient had impossible landmarks. In the ultrasound group, PSO images were of good quality in 90% of patients and were of better quality than were TM images (Supplemental Digital Content 1, Table 3, https://links.lww.com/AA/C990).

We performed the exploratory subgroup analyses according to the types of abnormalities of lumbar spine to address the potential influences of the differences between scoliotic spine and previously operated spine. Ultrasound-assisted technique was found to be superior to the landmark-guided technique for the primary outcome in both subgroups (Supplemental Digital Content 1, Table 4, https://links.lww.com/AA/C990).


This study demonstrated that neuraxial ultrasonography facilitated the technical performance of spinal anesthesia in patients with abnormal spinal anatomy including scoliosis and previous spine surgery. We observed that ultrasound assistance significantly reduced the number of needle passes required for success and increased the first attempt success rate without significantly prolonging the total procedure time compared with the landmark-guided technique.

The ultrasound-assisted neuraxial anesthesia in patients with anatomical abnormalities has been described in several case reports,5,24–26 but there are limited number of clinical studies in these populations. The previous clinical trials have focused on the patients with poorly-palpated surface landmarks.7,27,28 Although Chin et al7 showed that the ultrasound assistance improved the efficacy of spinal anesthesia compared with the landmark guidance in adults with difficult surface landmarks in their randomized study, most subjects were obese and only a small proportion (21%) had abnormal spinal anatomy. Therefore, the efficacy of the ultrasound-assisted technique compared with landmark-guided technique in patients with altered spinal anatomy remains unknown. For these reasons, we included only patients having anatomical abnormalities in their lumbar spines in this trial to determine the benefit of ultrasonography in such clinical settings.

Our data suggest that the use of ultrasonography can enhance the efficacy of spinal anesthesia in patients with mild scoliosis, of which the incidence was reported as 6% in elderly patients.29 Bowens et al5 previously developed an approach to neuraxial anesthesia for the scoliotic spine, and they recommended the providers to manage mild scoliosis with good positioning. Meanwhile, our results suggest that the providers should consider the ultrasound assistance in patients with mild scoliosis. Considering the anatomy of the scoliotic spine, a paramedian approach on the convex side of the scoliotic curve was recommended,5 and we observed this approach was highly efficient when performed under ultrasound assistance. Furthermore, ultrasound scan allowed the operator to identify the midline accurately and determine the optimal insertion angle in patients with significant axial rotation, in which the medial angulation was small or even absent.30

Patients with previous spine surgery are also at increased risk for technical difficulties.1,18 The spinous process can be indistinct or absent, and tissue adhesion or bone graft can hinder the neuraxial approach.6 Patient positioning can be limited, and skin scarring can disrupt the midline. However, we observed that dural puncture was achieved with a single needle pass in the patients with previous surgery in ultrasound group. Although the anatomical changes of these patients can be easily identified by preoperative simple radiography, the information derived from preoperative radiography may not easily translate to technical performance. In this situation, we observed the information from ultrasound imaging easily translated into an efficient neuraxial technique. Therefore, in accordance with previous reports,7,26,31 our results suggest the utility of the ultrasound assistance in patients with previous spine surgery. However, only a small number of patients with previous surgery were included in this study, and the subgroup analyses were not sufficiently powered. Moreover, there were more patients with previous surgery in the landmark group than in the ultrasound group; thus, our results should be interpreted cautiously.

The overall procedure time is of concern to many clinicians, particularly in busy clinical practice. Previous trials reported that the use of ultrasound increased the overall procedure time because of scanning time.7,28,32 Conversely, we observed no significant difference in total procedure time between the 2 techniques because the identifying time in the ultrasound group of our study was shorter than those in the previous studies.7,28,32 Namely, it was less than approximately 2 minutes in 75% of cases, and <5 minutes in all cases. We postulated that the scanning time could be reduced for 2 reasons. First, the patients had relatively low body mass indices (BMIs), and it might contribute to easier and faster scanning. Second, the ultrasound-assisted paramedian technique in this study involved primarily PSO views, of which scanning was relatively easy because PSO views consistently provide better sonographic images compared with TM views.21 Meanwhile, TM views were obtained for identifying midline in this study, and time taken to obtain TM views was reduced. Furthermore, all patients with scoliosis or previous surgery have unique anatomy, and individual planning using ultrasound was particularly helpful for spinal anesthesia in these populations. Thus, the time taken to perform spinal anesthesia was decreased in the ultrasound group. Moreover, 2 patients in the landmark group required the “rescue” use of ultrasound for the success of spinal anesthesia. The “rescue” ultrasound–assisted procedures took 143 seconds (66 seconds for scanning; 77 seconds for performing) and 185 seconds (120 seconds for scanning; 65 seconds for performing) in these patients, whereas the procedural time of landmark-guided technique until the declaration of conversion to alternative technique was 324 seconds (24 seconds for palpation; 300 seconds for performing) and 328 seconds (30 seconds for palpation; 298 seconds for performing), respectively. Our results therefore encourage the use of ultrasonography in these patients without concerns about prolonging the procedure time.

Although an increasing body of evidence indicates that ultrasound can improve the efficacy of neuraxial techniques, insufficient evidence exists on its safety outcomes owing to very low baseline incidences of the catastrophic complications of neuraxial techniques, namely <1 in 100,000 cases.1,12 However, it has been suggested that neuraxial ultrasound may possibly reduce several mechanisms of injury related to neurologic complications.1 Our findings can further extend the existing data regarding safety outcomes of ultrasound-assisted neuraxial techniques. A patient in the landmark group received spinal anesthesia at the L2–3 level, whereas no patient in the ultrasound group did so. As the conus medullaris is located at the L1 level ranging from T12 to L3,33 inaccurate identification of intervertebral levels could lead to conus medullaris injury.1,7,34,35 Of note, nevertheless, to conclude whether neuraxial ultrasound can improve the safety of spinal anesthesia, a much larger sample is required.

This study has limitations. First, neither primary outcome assessors nor performers could be blinded because of the nature of the study. Second, participants had a relatively low BMI, which may undermine the generalizability of our results. Third, as we did not record the ease of palpation in the ultrasound group, we were unable to compare the quality of surface landmarks directly between the 2 groups. Fourth, as the intervertebral level determined by palpation was not recorded in the landmark group, the discrepancy of interspace identification between ultrasound and palpation in this subpopulation could not be assessed. Fifth, the patients with previous surgery accounted for only a small proportion of participants (23%), and there were more patients with previous surgery in the landmark group than in the ultrasound group. As it cannot be excluded that the imbalance of type of abnormalities between the groups might be the probable explanation for the observed differences in this study, further studies are required to confirm our findings. Finally, we did not evaluate the use of a real-time ultrasound guidance because we consider it to be difficult to implement under the varying conditions at different hospitals due to technical complexity.1,12,18,21,36,37

In conclusion, for anesthesiologists with experience in neuraxial ultrasonography, the use of ultrasound can enhance the efficacy of spinal anesthesia in patients with anatomical alterations in the lumbar spine, including scoliosis and previous spinal surgery. We believe that these results can lead to practical suggestions that encourage the use of ultrasound for spinal anesthesia in patients with abnormal spinal anatomy.


Name: Sun-Kyung Park, MD.

Contribution: This author helped design the study, recruit the patients, collect and analyze the data, and draft the final manuscript.

Name: Jinyoung Bae, MD.

Contribution: This author helped recruit the patients and collect the data.

Name: Seokha Yoo, MD.

Contribution: This author helped recruit the patients and collect the data.

Name: Won Ho Kim, MD, PhD.

Contribution: This author helped analyze the data and draft the final manuscript.

Name: Young-Jin Lim, MD, PhD.

Contribution: This author helped collect and analyze the data, and revise the final manuscript.

Name: Jae-Hyon Bahk, MD, PhD.

Contribution: This author helped analyze the data and draft the final manuscript.

Name: Jin-Tae Kim, MD, PhD.

Contribution: This author helped design the study, recruit the patients, collect and analyze the data, and draft the final manuscript.

This manuscript was handled by: Richard Brull, MD, FRCPC.



1. Perlas A, Chaparro LE, Chin KJ. Lumbar neuraxial ultrasound for spinal and epidural anesthesia: a systematic review and meta-analysis. Reg Anesth Pain Med. 2016;41:251–260.
2. Atallah MM, Demian AD, Shorrab AA. Development of a difficulty score for spinal anaesthesia. Br J Anaesth. 2004;92:354–360.
3. Guglielminotti J, Mentré F, Bedairia E, Montravers P, Longrois D. Development and evaluation of a score to predict difficult epidural placement during labor. Reg Anesth Pain Med. 2013;38:233–238.
4. de Filho GR, Gomes HP, da Fonseca MH, Hoffman JC, Pederneiras SG, Garcia JH. Predictors of successful neuraxial block: a prospective study. Eur J Anaesthesiol. 2002;19:447–451.
5. Bowens C, Dobie KH, Devin CJ, Corey JM. An approach to neuraxial anaesthesia for the severely scoliotic spine. Br J Anaesth. 2013;111:807–811.
6. Ko JY, Leffert LR. Clinical implications of neuraxial anesthesia in the parturient with scoliosis. Anesth Analg. 2009;109:1930–1934.
7. Chin KJ, Perlas A, Chan V, Brown-Shreves D, Koshkin A, Vaishnav V. Ultrasound imaging facilitates spinal anesthesia in adults with difficult surface anatomic landmarks. Anesthesiology. 2011;115:94–101.
8. Auroy Y, Narchi P, Messiah A, Litt L, Rouvier B, Samii K. Serious complications related to regional anesthesia: results of a prospective survey in France. Anesthesiology. 1997;87:479–486.
9. Harrison DA, Langham BT. Spinal anaesthesia for urological surgery. A survey of failure rate, postdural puncture headache and patient satisfaction. Anaesthesia. 1992;47:902–903.
10. Horlocker TT, McGregor DG, Matsushige DK, Schroeder DR, Besse JA. A retrospective review of 4767 consecutive spinal anesthetics: central nervous system complications. Perioperative Outcomes Group. Anesth Analg. 1997;84:578–584.
11. Wulf H. Epidural anaesthesia and spinal haematoma. Can J Anaesth. 1996;43:1260–1271.
12. Neal JM, Brull R, Horn JL, et al. The Second American Society of Regional Anesthesia and pain medicine evidence-based medicine assessment of ultrasound-guided regional anesthesia: executive summary. Reg Anesth Pain Med. 2016;41:181–194.
13. Chin A, Crooke B, Heywood L, Brijball R, Pelecanos AM, Abeypala W. A randomised controlled trial comparing needle movements during combined spinal-epidural anaesthesia with and without ultrasound assistance. Anaesthesia. 2018;73:466–473.
14. 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.
15. Lucas DN, Elton CD. Through a glass darkly - ultrasound imaging in obstetric anaesthesia. Anaesthesia. 2016;71:617–622.
16. Cook TM, Counsell D, Wildsmith JA; Royal College of Anaesthetists Third National Audit Project. Major complications of central neuraxial block: report on the Third National Audit Project of the Royal College of Anaesthetists. Br J Anaesth. 2009;102:179–190.
17. Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. J Pharmacol Pharmacother. 2010;1:100–107.
18. Chin KJ, Karmakar MK, Peng P. Ultrasonography of the adult thoracic and lumbar spine for central neuraxial blockade. Anesthesiology. 2011;114:1459–1485.
19. Ghosh SM, Madjdpour C, Chin KJ. Ultrasound-guided lumbar central neuraxial block. BJA Education. 16:213–220.
20. Tawfik MM, Atallah MM, Elkharboutly WS, Allakkany NS, Abdelkhalek M. Does preprocedural ultrasound increase the first-pass success rate of epidural catheterization before cesarean delivery? A randomized controlled trial. Anesth Analg. 2017;124:851–856.
21. Kallidaikurichi Srinivasan K, Iohom G, Loughnane F, Lee PJ. Conventional landmark-guided midline versus preprocedure ultrasound-guided paramedian techniques in spinal anesthesia. Anesth Analg. 2015;121:1089–1096.
22. Park SK, Yoo S, Kim WH, Lim YJ, Bahk JH, Kim JT. Ultrasound-assisted vs landmark-guided paramedian spinal anaesthesia in the elderly: a randomised controlled trial. Eur J Anaesthesiol. 2019;36:763–771.
23. Chin KJ, Ramlogan R, Arzola C, Singh M, Chan V. The utility of ultrasound imaging in predicting ease of performance of spinal anesthesia in an orthopedic patient population. Reg Anesth Pain Med. 2013;38:34–38.
24. McLeod A, Roche A, Fennelly M. Case series: ultrasonography may assist epidural insertion in scoliosis patients. Can J Anaesth. 2005;52:717–720.
25. Minville V, Lavidalle M, Bayoumeu F, Parrant O, Fourcade O. [Ultrasound-assisted spinal anesthesia in obstetric patient with corrected scoliosis]. Ann Fr Anesth Reanim. 2010;29:501–502.
26. Yamauchi M, Honma E, Mimura M, Yamamoto H, Takahashi E, Namiki A. Identification of the lumbar intervertebral level using ultrasound imaging in a post-laminectomy patient. J Anesth. 2006;20:231–233.
27. Grau T, Leipold RW, Conradi R, Martin E. Ultrasound control for presumed difficult epidural puncture. Acta Anaesthesiol Scand. 2001;45:766–771.
28. Wang Q, Yin C, Wang TL. Ultrasound facilitates identification of combined spinal-epidural puncture in obese parturients. Chin Med J (Engl). 2012;125:3840–3843.
29. Leung CC, Yu KS, Chau LF, Sze TS. Management of difficult spinal anaesthesia in a patient with adult lumbar scoliosis. Anaesth Intensive Care. 2002;30:101.
30. Chin KJ, Perlas A, Chan V. The ultrasound-assisted paraspinous approach to lumbar neuraxial blockade: a simplified technique in patients with difficult anatomy. Acta Anaesthesiol Scand. 2015;59:668–673.
31. Geng J, Li M. [Ultrasound-assisted neuraxial anesthesia in a patient with previous lumbar laminectomy and fusion: a case report]. Beijing Da Xue Xue Bao Yi Xue Ban. 2016;48:747–750.
32. Abdelhamid SA, Mansour MA. Ultrasound-guided intrathecal anesthesia: does scanning help? Egypt J Anaesth. 2013;29:389–394.
33. Kim JT, Bahk JH, Sung J. Influence of age and sex on the position of the conus medullaris and Tuffier’s line in adults. Anesthesiology. 2003;99:1359–1363.
34. Reynolds F. Logic in the safe practice of spinal anaesthesia. Anaesthesia. 2000;55:1045–1046.
35. Reynolds F. Damage to the conus medullaris following spinal anaesthesia. Anaesthesia. 2001;56:238–247.
36. Niazi AU, Chin KJ, Jin R, Chan VW. Real-time ultrasound-guided spinal anesthesia using the SonixGPS ultrasound guidance system: a feasibility study. Acta Anaesthesiol Scand. 2014;58:875–881.
37. Chong SE, Mohd Nikman A, Saedah A, et al. Real-time ultrasound-guided paramedian spinal anaesthesia: evaluation of the efficacy and the success rate of single needle pass. Br J Anaesth. 2017;118:799–801.

Supplemental Digital Content

Copyright © 2019 International Anesthesia Research Society