American Journal of Emergency Medicine (2007) 25, 331 – 334
www.elsevier.com/locate/ajem
Brief Report
The use of ultrasound to identify pertinent landmarks for lumbar puncture Kirk A. Stiffler MD*, Sharhabeel Jwayyed MD, Scott T. Wilber MD, Angela Robinson DO Department of Emergency Medicine, Summa Health System, Akron, OH, USA Northeastern Ohio Universities College of Medicine, Rootstown, OH, USA Received 18 April 2006; revised 10 July 2006; accepted 14 July 2006
Abstract Objective: This study was conducted to assess the ultrasound’s (US’s) ability to identify pertinent landmarks for lumbar puncture (LP) in patients of various body mass indices (BMIs) and establish spatial relationships of pertinent LP landmarks across BMIs. Methods: In this institutional review board–approved cross-sectional study, we calculated the BMIs of eligible patients and then categorized them as normal (BMI V24.9), overweight (BMI 24.9-30), or obese (BMI z30). We recorded the difficulty in palpating traditional LP landmarks. Identification and measurement of the spatial relationships of the sacrum; spinous processes of lumbar vertebrae L3, L4, and L5; ligamentum flavum; and the spinal canal by US was attempted. Results: Successful identification of pertinent structures (L4-L5 spinous processes and the spinal canal) occurred in 100% of patients with normal BMI, 95% of those who were overweight, and 74% of those who were obese ( P = .011). Difficulty in palpating landmarks was noted in 5% of patients with normal BMI, 33% of those who were overweight, and 68% of those who were obese ( P b .0001). In subjects with difficult-to-palpate landmarks, US identified pertinent structures in 16 of 21 (76%; 95% confidence interval, 53-92). The average distance from skin to ligamentum flavum was 44 mm in those with normal BMI, 51 mm in those who were overweight, and 64 mm in those who were obese ( P b .00001); measurements between spinous processes did not vary by BMI. Overall, there was a moderate correlation (0.62) between BMI and the distance from skin to ligamentum flavum. Conclusion: The usefulness of US in identifying structures for LP is inversely related to BMI. Even with this limitation, US is still able to identify obese patients’ pertinent landmarks almost 75% of the time. In addition, US may be helpful in identifying pertinent structures for LP in those patients with difficult-to-palpate landmarks. In patients who were obese with structures not palpable by hand or identifiable by US, other modalities should be considered. D 2007 Elsevier Inc. All rights reserved.
Presented as poster at the American College of Emergency Physicians National Scientific Assembly, Washington, DC, on September 26, 2005. * Corresponding author. Emergency Medicine Research Center, Summa Health System, Akron City Hospital, PO Box 2090, Akron, OH 44309-2090, USA. Tel.: +1 330 375 7530; fax: +1 330 375 7564. E-mail address:
[email protected] (K.A. Stiffler). 0735-6757/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.ajem.2006.07.010
1. Introduction Lumbar puncture (LP) is an important diagnostic procedure in emergency medicine. Although surface landmark–guided LP is often successful, there are some patients in whom the procedure is difficult. An alternative to surface
332 landmark–guided LP is fluoroscopic guidance, although this requires transport out of the emergency department (ED), as well as radiation exposure for the patient. Furthermore, this usually requires a radiologist to perform the procedure for the experience with fluoroscopy and hospital requirements. Emergency physicians have become increasingly comfortable with ultrasound (US) for diagnostic evaluations as well as procedural assistance [1]. To be helpful for LP, US must be able to identify pertinent landmarks for LP in patients in whom surface landmark–guided LP is difficult. Although little data exist on predictors of difficult LP, we have anecdotally found that body habitus is a predictor of failure of surface landmark–guided LP. Therefore, our aim was to study the ability of US to identify pertinent landmarks used for LP and to establish underlying spatial relationships of those landmarks in patients of various body mass indices (BMIs).
2. Methods 2.1. Study design We performed a prospective cross-sectional study approved by an institutional review board.
2.2. Study setting and population Of the 4 authors of the study, 3 are board-certified emergency medicine attending physicians; the fourth is a third-year emergency medicine resident. All physicians have undergone formal training in US use and have extensive experience in using US in their clinical practice. In addition to reviewing relevant textbooks and literature, all authors underwent training with members of the radiology department to learn the appearance of the relevant lumbar spine sonographic landmarks. We also piloted lumbar spine ultrasonography on a sample of healthy volunteers before the study to ensure that each investigator could correctly identify the appropriate structures and set forth certain criteria to be followed by all investigators, which would allow consistent distance measurements. After the training and the piloting were completed, we enrolled a convenience sample of eligible patients at an ED. We included patients between the ages of 18 and 55 years, who were present during study times, were able to understand the statement of research, and were able to lie in the lateral decubitus position for the duration of US. We excluded those with significant scoliosis, previous back surgeries, pregnant women, and those unable to comply with instructions.
2.3. Study protocol After explaining the study and giving the patient a statement of research, we recorded pertinent information including age and sex, and we measured each patient’s height and weight. Body mass index was calculated using the formula BMI = [weight in pounds/(height in inches
K.A. Stiffler et al.
Fig. 1 2-D ultrasound lumbar spine, longitudinal view. (Picture taken by investigators at Summa Health System during study).
height in inches)] 703. The following BMI categories were used: normal, 24.9 or less; overweight, 25 to 29.9; obese, 30 or more. We then placed each patient in a lateral decubitus position and rated (on a 5-point Likert scale ranging from very easy to very difficult to palpate) the ability to palpate the following landmarks: both anterior superior iliac spines; the spinous process of lumbar vertebrae L3, L4, and L5; and the sacrum. An investigator performed a 2-dimensional US of the patient’s lumbosacral spine area with a Siemens G20 machine using a variable frequency (2-5 MHz) curvilinear transducer (Siemens, New York, NY) (Fig. 1). The curvilinear lower-frequency probe was used to facilitate visualization of several lumbar vertebral levels and the sacrum within 1 field of view. In addition, the lower frequencies allowed deeper penetration in those patients whose higher BMIs may have limited visualization of the needed structures. The transducer was placed on the midsagittal plane of the subject’s back. The probe was subsequently adjusted to visualize the spine directly below the transducer. Depth, gain, and frequency controls were adjusted within the US machine’s capability and as the operator deemed appropriate for each individual patient. In the sagittal plane, the following structures were searched for, and their presence were noted: the spinous processes of L3, L4, and L5; sacrum; ligamentum flavum; and spinal canal containing spinal fluid. Each US was reviewed by 2 investigators to verify correct identification of the relevant structures. As each of the structures of interest was visualized, measurements of the spatial relationships relevant to LP were recorded. The machine’s internal calipers were used to measure distances relevant to LP, including skin-to-spinous processes (in a line perpendicular to the skin), distance between the lumbar spinous processes (L3-L4 and L4-L5), and distance from skin to ligamentum flavum.
2.4. Outcome measures The primary outcome was successful identification of the L4 spinous process, L5 spinous process, and the spinal
Ultrasound identification of LP landmarks canal. The secondary outcome was the distance of the skin to ligamentum flavum in a plane perpendicular to the skin.
2.5. Data analysis Data were entered into a Microsoft Access database (Microsoft, Redmond, WA) and were converted to Stata for statistical analysis. We used v 2 analysis to analyze differences in the primary outcome by BMI category, and analysis of variance to analyze the secondary outcome. We estimated a 100% success in patients with normal BMI, 90% in those who were overweight, and 75% success in those who were obese. A sample size of 20 patients was required in each BMI category to detect a difference in the primary outcome with an a of .05 and a b of .20.
3. Results We enrolled 62 patients: 22 with a normal BMI, 21 overweight, and 19 obese. The mean age was 36 F 11 years, and 66% were female. We successfully identified pertinent structures in 100% (22/22) of those with normal BMI, 95% (20/21) of those who were overweight, and 74% (14/19) of those who were obese ( P = .011). Difficulty in palpating landmarks was noted in 5% (1/22) of those with normal BMI, 33% (7/21) of those who were overweight, and 68% (13/19) of those who were obese ( P b .0001). In 21 subjects with difficult-to-palpate landmarks, US identified pertinent structures in 16 (76%; 95% confidence interval, 53-92). The average distance from skin to ligamentum flavum was 44 mm in normal BMI (SD, 7 mm), 51 mm in overweight (SD, 8 mm), and 64 mm in obese (SD, 13 mm) (overall P b .00001); measurements between spinous processes did not vary by BMI. Overall, there was a moderate correlation of 0.62 (95% confidence interval, 0.43-0.75) between BMI and distance from skin to ligamentum flavum.
4. Discussion Diagnostic LP is commonly performed in the ED to obtain cerebrospinal fluid (CSF) for analysis. It is typically performed via blind surface landmark guidance, as first described by Quincke [2]. To find the most common interspace used in LP (L3-L4 level), an imaginary line is drawn between the patient’s 2 superior iliac crests. Lumbar puncture can be performed in this space or 1 level below this, but should not be performed above this level. The surface landmark technique is reported successful in a high percentage of attempted LPs and has a low rate of complications, which include headache, brain herniation, traumatic injuries, and infections [3-5]. However, surface landmark identification of underlying structures has been shown to be accurate only 30% of the time [6]. Inappropriate or unsuccessful identification of
333 proper landmarks can lead to a reluctance by physicians to perform the procedure, increased difficulty in obtaining CSF if the procedure is performed, as well as higher rates of complications. Currently, there are few alternatives to obtaining CSF if the landmark technique cannot be used or is unsuccessful. In most cases, treatment is either initiated without the CSF, which may delay and obfuscate exact diagnosis, or if available, fluoroscopic-guided LP is performed. Disadvantages of fluoroscopic guidance include its limited availability, its location outside the ED, its inability to directly visualize the spinal canal, and its inherent radiation risks. Other specialties have demonstrated the use of US in localizing intervertebral levels and appropriate epidural spaces for anesthetic catheters [6-8]. The use of US has even been shown useful for documenting causes of failure and for assisting with guidance of a small number of neonatal and infant LPs [9]. Peterson and Abele [10] detailed the technique of US-assisted LP in a case study of 2 patients in the emergency department in whom palpation of landmarks was likely to be difficult. Given the advantages of USs including ease of use, its noninvasive nature, and its increasing availability in emergency departments, using available technology to assist in difficult procedures may be beneficial. Our results indicate that US is indeed useful in identifying pertinent landmarks used for LP. The amount of assistance provided by the US was inversely related to the patient’s BMI. However, even in those patients whose landmarks were rated as difficult to palpate, or whose BMIs were in the obese category, US identified useful landmarks 75% of the time. In this population, surface landmark– guided LP is typically more difficult. Ultrasound may thus provide enough information to allow the operator to proceed with more certainty as to where the appropriate intervertebral level is located. As expected, the only distance that varied between BMI categories was the skin-to-ligamentum flavum distance, with a direct relationship to BMI. Although our study demonstrated that US was able to identify landmarks used for LP, we did not use US in real time to assist with the actual procedure. Furthermore, the population was a convenience sample of adult patients at the ED, not patients undergoing an actual LP. Consequently, the results of this study should be considered preliminary and may not translate into increased rates of success for LP, decreased length of time for completion of procedures, or diminished complication rates for the procedure. Further studies using US for real-time assistance with LP are needed. Operator experience and comfort level with US may alter the overall usefulness when performing LPs.
5. Conclusion The usefulness of US in identifying structures for LP is inversely related to BMI. Even with this limitation, US is
334 still able to identify obese patients’ pertinent landmarks almost 75% of the time. In addition, US may be helpful in identifying pertinent structures for LP in those patients with difficult-to-palpate landmarks. In obese patients with structures not palpable by hand or identifiable by US, other modalities should be considered. The distance from skin to ligamentum flavum is directly related to BMI.
References [1] Tibbles CD, Porcaro W. Procedural applications of ultrasound. Emerg Med Clin North Am 2004;22(3):797 - 815. [2] Quincke HI. Ueber hydrocephalus. X Congress f. innere Medicin (Weisbaden), 1891. [3] Gorelick P, Biller MD. Lumbar Puncture. Techniques, indications, and complications. Postgrad Med 1986;79(8):257 - 68.
K.A. Stiffler et al. [4] Thoennissen J, Herkner H, Lang W, Domanovits H, Laggner AN, Mullner M. Does bed rest after cervical or lumbar puncture prevent headache? A systematic review and meta-analysis. CMAJ 2001; 165(10):1311 - 6. [5] Holdgate A, Cuthbert K. Perils and pitfalls of lumbar puncture in the emergency department. Emerg Med (Fremantle) 2001;13(3):351 - 8. [6] Furness G, Reillly MP, Kuchi S. An evaluation of ultrasound imaging for identification of lumbar intervertebral level. Anaesthesia 2002;57: 277 - 80. [7] Kadziolka R, Asztely M, Hanai K, Hansson T, Nachemson A. Ultrasonic measurement of the lumbar spinal canal. J Bone Joint Surg 1981; 63-B, 504-7. [8] Grau T, Leipold RW, Conradi MD, Martin E, Motsch MD. Ultrasound imaging facilitates localization of the epidural space during combined spinal and epidural anesthesia. Reg Anesth 2001;26:64 - 7. [9] Coley B, Shiels W, Hogan M. Diagnostic and interventional ultrasonography in neonatal and infant lumbar puncture. Pediatr Radiol 2001;31:399 - 402. [10] Peterson M, Abele J. Bedside ultrasound for difficult lumbar puncture. J Emerg Med 2005;28(2):197 - 200.