Inaccurate level of intervertebral space estimated by palpation: The ultrasonic revelation.

BACKGROUND: Spinal cord injury resulting from spinal anesthesia is a rare, but an alarming scenario. The most likely cause is believed to be misjudged level of intervertebral space (IVS). We evaluated the accuracy of palpation method to locate IVS with the ultrasonography.
MATERIALS AND METHODS: A total of 109 patients undergoing spinal anesthesia were included in this observational, double-blind study. First anesthesiologist was asked to mark IVS using palpation method. It was followed by ultrasonographic assessment by another anesthesiologist who was unaware of the level estimated for the mark. We evaluated the accuracy of palpation method in sitting and lateral position as well as the impact of the anesthesiologist's experience (Trainee/Consultant). STATISTICAL ANALYSIS USED: Association between the gender, anthropometric parameters, type of anesthesiologists assessing the IVS, and the level of agreement were identified using Chi-square test. The agreement between palpation method and ultrasound assessment of IVS was analyzed using kappa statistic. P < 0.05 was defined as statistical significance.
RESULTS: The IVS located by palpation method was in agreement with ultrasound location in 37.14% of the patients. There were no statistically significant differences found in terms of demographic data (sex, age, height, weight, or body mass index [BMI]) between agreement and disagreement group. The rate of errors was found to be significantly higher (P = 0.01) among the trainees (74.51%) than the consultants (51.86%). The rate of errors was not different between the sitting and lateral position. The frequency of errors was more common in cephalad direction (53.31.5%) compared to caudal direction (9.52%). The misidentified spaces were as high as three spaces above the intended space while in caudal direction it differed by only one space.
CONCLUSION: The accuracy of palpation method controlled by ultrasound is 37.14% and differs by 1-3 IVS in cephalad direction (53.31%). The accuracy is affected by anesthesiologist's experience but remains unaffected by age, sex height, BMI, and patient positioning.

Introduction

Ideal site for spinal anesthesia is below conus medullaris to avoid trauma to spinal cord. The intercristal line or Tuffier line is an imaginary line joining superior aspect of iliac crest which is known to cross the L4 vertebra or L4-L5 intervertebral disc. This is an important surface landmark to select intervertebral space (IVS) for neuraxial block. In adults, conus medullaris ends at the lower border of L1 verterbra. Since there are variability in position of termination of conus medullaris, selecting lowermost space becomes crucial for practicing safe neuraxial anesthesia.[1]

The most common method of selecting IVS is palpation method which is based on identification of Tuffier line. Palpation method is shown to be least reliable method with accuracy rate being as low as 29%.[2] In the recent years, ultrasonography (USG) has been extensively evaluated for practicing safe neuraxial anesthesia. It is a simple, less time consuming, noninvasive, and easy to learn technology.

Hence in this study, ultrasound is used to evaluate the accuracy of palpation method to estimate the level of the IVS. In addition to that the effect of demographic parameters, patient positioning and anesthesiologists experience were also assessed.

Materials and Methods

This observational single arm study was approved by the Hospital Ethics and Research Committee (Ref no. ECR/215/Inst/Ker/2013) and registered with Clinical Trial Registry India (CTRI/2014/09/005035). One hundred and nine patients undergoing procedures under spinal anesthesia were recruited in this study. Patients aged between 18 and 60 years undergoing lower abdominal or lower limb surgeries were included in this study and written informed consent was taken. Pregnant patients, severely obese patients with body mass index (BMI) >35 kg/m2, spine deformity, or patients who were unable to position themselves for spinal anesthesia were excluded from the study. Patients with poor imaging quality were also excluded from the analysis. Resident anesthesiologist undergoing anesthesia training with minimum 1 year of experience was taken in the trainee group. All consultant anesthesiologists with minimum 5 years of anesthesia experience were taken in consultant group. Two consultant anesthesiologists having more than 5 years of experience in neuraxial ultrasound were allotted to perform the ultrasound assessment throughout the study. They were kept blinded about the marked IVS. Inside of the operation theater, patient was positioned either in sitting or lateral position. An anesthesiologist palpates, marks the chosen interspace, defines the level, and keeps this level in secret. Another anesthesiologist scans with ultrasound the marked interspace and provides his definition of level. Later the levels estimated by palpation and identified by ultrasound are confronted in search for agreement or disagreement. The ultrasound assessment was done using GE Healthcare Venue 40 software and curved array low frequency probe of 2-5 MHz, basing in parasagittal oblique view. Sacrum was taken as a reference landmark, which appears as a horizontal hyperechogenic line [Figure 1]. Probe was then moved in cephalad direction to assess IVS. The hyperechogenic pattern corresponding to the laminae of vertebra while hypoechogenic shadow corresponding IVS forms a saw tooth like pattern [Figure 2]. For accurate counting of interspace, each space corresponding to the center of probe was marked on skin. After ultrasound assessment, neuraxial block was performed in a space unrelated to the studied space in routine manner and surgery was continued. Anesthesiologists were not given immediate feedback of the assessment to avoid self-improvement. On an average, every anesthesiologist marked IVS in maximum 4 patients.

Figure 1

Hyperechogenic line corresponding to sacrum and L5-S junction

Figure 2

Laminae of vertebrae

The primary aim of this study was to assess under ultrasonographic control the accuracy of palpation method to identify lumbar IVS. Secondary aim was to evaluate the effect demographic parameters, patient positioning (sitting and lateral) and anesthesiologist's experience on the accuracy of the palpation method.

Statistical analysis

Sample size was estimated using nMaster (sample size software) and basing on the study of Whitty et al.[3] Considering 55% agreement between IVS assessment by USG and palpation, kappa of 0.08, power of 80%, alpha error of 5% and population agreement of 0.5, we estimated that at least 101 subjects were required. We enrolled 109 to account for subjects drop out.

Data were analyzed using SPSS software, version 18 (Chicago). The demographic details of the study population were presented as proportions and means (+standard deviation). Association between the gender, anthropometric parameters, type of anesthesiologists assessing the IVS and the level of agreement were identified using Chi-square test. The agreement between palpation method and ultrasound assessment of IVS was analyzed using kappa statistic. P < 0.05 was defined as statistical significance.

Results

A total of 109 patients (38 women and 71 men) were enrolled in the study. Four patients were excluded from the study due to poor imaging quality. Data regarding 105 patients were analyzed. The IVS located by palpation method was in agreement with ultrasound location in 37.14% (39 out of 105) of the patients [Figure 3]. The errors were more frequent in cephalad direction (53.31%) and less frequent in caudal direction (9.52%). The misidentified spaces usually differed by one space to as high as three spaces above the intended space. In caudal direction, it differed by only one space.

Figure 3

Disagreement of intervertebral space determination between the anatomical and ultrasound method. Negative values denote deviation in caudal direction and positive values in cephalad direction

The demographic characteristics are presented in Table 1. Demographic and anthropometric data of patients in groups agreement and disagreement with IVS determinations based on palpation and ultrasound imaging are listed in Table 2. There were no statistically significant differences.

Table 1

Demographic data of study population

Table 2

Comparison of anthropometric and demographic data in group of patients with consistent versus inconsistent results of IVS

The other parameters likely to influence the assessment are shown is Table 3. The IVS assessment done by trainee and consultant in sitting and lateral positions and their ultrasound interpretations are shown in Tables 4–7. The rate of errors was found to be higher among the trainees (74.51%) than the consultants (51.86%). This difference was statistically significant (P = 0.01). The rate of errors 66.67% in sitting position and 58.82% in lateral position was not significantly different (P = 1).

Table 3

Parameters affecting the agreement between clinical and ultrasound determination of IVS levels

Table 4

Consultant—sitting

Table 7

Trainee—lateral

Discussion

In our study, the accuracy of palpation method to estimate IVS was poor (37.14%) when controlled by ultrasound. The results showed that the inaccuracy of the palpation method was neither due to demographic reason nor to the patient positioning in sitting or in lateral. They also showed that the accuracy or inaccuracy depended on the anesthesiologist's experience.

The accuracy of palpation method to assess IVS is shown to be in between 29% and 64%.[234567] Our findings are consistent with the finding of Schlotterbeck et al. who observed the accuracy of palpation method to be 36.4% in pregnant patients. The gold standard techniques to assess IVS are computerized tomography (CT), magnetic resonance imaging (MRI), X-ray, and fluroscopy. However, the limitations of these techniques (CT, MRI) are patient's need to lie in supine unflexed position and risk of radiation. In clinical practice, the IVS is palpated after the patient is positioned appropriately by flexing hip and neck. Hence these methods (CT, MRI) can lead to bias when the accuracy is compared with palpation. The advantage of ultrasound technique is being a bedside technique and the assessment can be done in the same flexed position given for spinal anesthesia. Hence, the bias of displacement of marker as seen with CT and MRI can be eliminated. The accuracy of ultrasound assessment when compared to other various techniques has been reported as 68-76% but with appropriate training and experience it can enhanced up to 90%.[789] The ultrasound estimation usually differs from CT and MRI by one space unlike in palpation where errors can be up to 4 spaces. Identifying interspinous spaces is reported to be relatively easier than estimating epidural depth and optimal needle insertion point.[10] Relevant anatomical structures for lumbar puncture can be identified quickly in emergency setting.[11] Hence, spinal ultrasound can form an important tool for practicing safe neuraxial anesthesia and for assessment of accuracy of the palpation method in this study.

In our study, patient positioning has not improved accuracy. Theoretically sitting position should improve the accuracy. In sitting position, it is easier to eliminate the effect of lumbar lordosis. It is easy to determine the intercristal line in this position as it gives good grip by palpating both iliac crests. The distribution of body weight is even in sitting position as compared to lateral position where patient is lying on uneven surface. In lateral position, the axis of pelvic brim is deviated from spine resulting in the variation in the position of Tuffiers line while sitting position helps better to widen the lumbar interspaces and thus pushes Tuffiers line caudally to cross L4-L5 space.[12] Larysa and Broadbent demonstrated no difference in sitting and lateral position. In Larysa et al. study, majority of cases were done in sitting position (89%) and they studied exclusively L3-L4 IVS (76%). While Broadbent has compared palpation assessment with MRI which requires patient to lie down in supine unflexed position which in turn could have caused bias. Ours is the first kind of study where ultrasonographic assessment was carried out for each lumbar IVS with an aim to see impact of patient positioning and experience of anesthesiologist's hand.

In our study, anesthesiologist's experience has improved accuracy. These findings were consistent with Larisa et al. who noticed accuracy improves significantly with experienced anesthesiologist's with professional experience of 10 years. Broadbent, Schlotterbeck and Whitty et al. reported that accuracy is unaffected by level of experience. However in all of these studies, the years of anesthesia experience studied varied widely. As per Broadbent et al., the experience may not improve the accuracy as there is no element for self-correction in the professional training because of lack of radiographic confirmation. Even in experienced hand, errors have occurred 3-4 spaces above the space they believed in and it was a common finding in our study also. This highlights the need of the ultrasound as a modality to confirm IVS for every anesthesiologist's irrespective of years of experience.

There is difference in the opinion about the space corresponding the intercristal line. As per traditional teaching, the intercristal line crosses the body of L4 or the L4-L5 interspace. This forms an important landmark while giving spinal anesthesia. However, this landmark is shown to be influenced by age, sex, pregnancy, height, BMI of the patient as well as imaging methods used to identify the space.[13141516] In pregnant patients, it shown to vary in range of L1-L2 to L4-L5 spinal segments while in tall males it crosses L2-L3 spaces. The position of conus medullaris varies from T12-upper to L3-upper segments and elderly females tend to have low conus. Use of atraumatic spinal needle is not completely risk free. The position of cones medullaris in relation to vertebral column does not change after spinal flexion and hence does not offer any protection against spinal cord damage.[17]

In our study, we have not evaluated for lumbosacral anomaly (lumbarization of sacrum and sacralization of lumbar vertebra). The incidence of these anomalies is around 6% and can be easily missed by ultrasound.[18] Spine radiograph combined with MRI is a gold standard technique to diagnose and to count lumbosacral vertebral anomalies. Kim et al. have shown that despite the transitional vertebrae there seems to be margin of safety of two IVS and these authors consider that counting up from lumbosacral junction is fairly valid approach to select the interspace for spinal anesthesia.[18]

Although neurological complications with spinal anesthesia is rare, but there are some disturbing evidence of grievous spinal injury and likely medico legal malpractice. In addition to liberal attitude of anesthesiologist for selecting lumbar IVS, use of atraumatic spinal needle has been cited as a cause of damage to spinal cord.[19] It is not uncommon to anesthesiologist to select most easily palpated lumbar interspace to improve the success rate. Anesthesiologist may tempt to prefer L2-L3 space in order to reduce the dose of spinal anesthetic drug but in this space if more cephalad needle is directed more likely are the chances of needle trauma.[20] Greaves reported a case of serious spinal injury caused by misidentification of the L2-L3 IVS in a patient receiving heparin.[21] Atraumatic needles (pencil point) carries more risk as free flow of cerebrospinal fluid comes after needle is well advanced in the subarachnoid space.

Major drawback with ultrasound is inability to visualize conus medullaris and hence its routine use during spinal anesthesia to improve safety is questionable. Use of MRI, fluoroscopy before spinal anesthesia is deemed unnecessary. Hence, the skill to locate IVS and upmost care to perform lumbar puncture remains most important safety measure.

The limitation of this study is exclusion of high risk population such as severely obese and pregnant patients. The reason we excluded these patients is that obesity would have impaired the quality of ultrasound images and pregnant patients would experience discomfort during forced flexion. However, it is still noteworthy that accuracy is not improved in normal population even in experienced hand. Hence, ultrasound may be used as a safeguard to prevent dural puncture being performed at misidentified dangerous high level. Future studies are required to study if the use of USG can reduce the neuraxial damage caused by spinal anesthesia.

Conclusion

The accuracy of palpation method when controlled by ultrasound is “poor” and differs by 1-3 IVS particularly in cephalad direction. The accuracy is affected by anesthesia experience but remains unaffected by age, sex height, BMI and patient positioning.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Table 5

Consultant—lateral

Table 6

Trainee—sitting