The Journal of Emergency Medicine, Vol. 47, No. 1, pp. 59–64, 2014Copyright � 2014 Elsevier Inc.

Printed in the USA. All rights reserved0736-4679/$ - see front matter

http://dx.doi.org/10.1016/j.jemermed.2012.09.149

RECEIVED: 19 MACCEPTED: 5 Se

Techniquesand Procedures

ULTRASOUND-ASSISTED LUMBAR PUNCTURE INPEDIATRIC EMERGENCY MEDICINE

Stephen Kim, MD and David K. Adler, MD, MPH

Emergency Department, University of Rochester, Strong Memorial Hospital, Rochester, New YorkReprint Address: Stephen Kim, MD, University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555

, Abstract—Background: Ultrasound-assisted lumbarpuncture in the pediatric emergency medicine setting hasnot been well established, but ultrasound could serve as avaluable tool in this setting. Objective: To assess whetherultrasound increases provider confidence in identifying aninsertion point for lumbar puncture. Methods: A feasibilitystudy was conducted using a convenience sample of pediat-ric emergency patients requiring lumbar puncture. Pro-vider confidence in selecting a needle insertion site forlumbar puncture using ultrasound assistance was comparedto provider confidence using traditional landmarks alone. Asimple technique using a linear probe is described. Results:Nineteen patients were included in the study, with the pri-mary end point the mean confidence score (based on afive-point Likert scale) in identifying a needle insertion siteprior to and after using ultrasound. Using the Wilcoxonsigned-rank test, the mean confidence score was 2.89 withthe landmark procedure alone, and 4.79 with ultrasoundassistance, yielding an average score difference of 1.90(95% confidence interval 1.23–2.56; Wilcoxon p < 0.001,paired t-test p < 0.001). Thus, compared to the landmarkprocedure, the use of ultrasound was associated with asignificantly higher average confidence score. Conclusion:The use of ultrasound in the pediatric emergency settingcan be a valuable adjunct with lumbar puncture. � 2014Elsevier Inc.

, Keywords—pediatric; emergency; ultrasound; lumbarpuncture

ay 2010; FINAL SUBMISSION RECEIVED: 20 Februaptember 2012

59

INTRODUCTION

The use of ultrasound-assisted lumbar puncture in the pe-diatric emergency setting has not been well established,but ultrasound could be a valuable tool in this setting.

Lumbar puncture is an essential procedure foranalyzing cerebrospinal fluid in the evaluation for sub-arachnoid hemorrhage, meningitis, sepsis, or fever inneonates and infants <3 months of age. The procedure en-tails palpating the superior iliac crests and then identi-fying midline at the L3–L4 or L4–L5 interspace.Neonates and infants younger than 3 months of age areoften in the lateral decubitus position for the procedure.

In the pediatric setting, the procedure can be anxietyprovoking, not only for the provider, but also for the pa-tient and family. Multiple unsuccessful attempts may bedue to poor positioning, patient movement, abnormalanatomy, body habitus, or accumulation of blood in theepidural space. If unsuccessful, the patient may ulti-mately be referred to interventional radiology forfluoroscopy-guided lumbar puncture (1).

Ultrasound may be beneficial in the pediatric settingby increasing provider confidence in the identificationof key landmarks and minimizing the number of attemptsfor a successful tap. Ultrasound-assisted lumbar puncturealso may alleviate parent anxiety by minimizing thelength of time and number of attempts required to achieve

ry 2012;

Figure 1. Curvilinear probe uses lower frequency to visualizedeeper structures, and linear probe uses higher frequency tovisualize superficial structures.

Figure 2. Transverse view is perpendicular to spine at L4–L5interspace by palpating top of iliac crests and drawing imag-inary line with probe parallel to this line.

60 S. Kim and D. K. Adler

lumbar puncture. Furthermore, ultrasound in the pediatricemergency department (ED) may decrease radiationexposure by obviating the need to undergo fluoroscopy-guided lumbar puncture.

MATERIALS AND METHODS

This is a feasibility study using a convenience sample ofemergency pediatric patients requiring lumbar puncture.Institutional Review Board approval was obtained priorto study initiation. Training in ultrasound-assisted pediat-ric lumbar puncture was provided to pediatric emergencymedicine faculty, emergency medicine residents, pediat-ric emergency medicine fellows, and nurse practitionersthrough a formal didactic session. This didactic sessionincluded a formal PowerPoint (Microsoft Corporation,Redmond, WA) presentation and hands-on workshop. In-clusion criteria were all pediatric patients #18 years ofage who required a lumbar puncture. Exclusion criteriaincluded any patients with signs of increased intracranialpressure, hematologic abnormalities, or signs of infectionclose to the insertion site. Additionally, any patients whowere incarcerated or brought in by foster parents wereexcluded.

The primary purpose of this study was to assess ifultrasound increased provider confidence in identifyingan insertion point for lumbar puncture. The patient’s par-ent(s) were given an informational letter in which the pro-vider explained the purpose of the study. Patients andtheir families were given the opportunity to ask any ques-tions regarding the study. Questionnaires were completedby providers both prior to and after the lumbar puncturewas performed. The questionnaire is shown in theAppendix.

Technique

Prior to the lumbar puncture, providers first used thetraditional landmark technique by palpating the superioriliac crests and identifying the midline by palpating thedorsal spinous processes. The providers would then circletheir level of confidence in needle insertion point identi-fication on a Likert scale ranging from one to five(‘‘none’’ to ‘‘excellent’’ level of confidence, respectively).

Our study utilized the linear probe that maximizes thedetail in superficial structures due to its high frequency(Figure 1). The curvilinear probe can be used if the pa-tient is obese because its lower frequency maximizesthe detail in deeper structures (Figure 1).

Secondly, the providers used ultrasound to augmentthe landmark technique in identifying a needle insertionsite. A linear probe (5–10 MHz) was first placed in atransverse orientation (perpendicular to the spine)(Figure 2). Anatomic midline was identified by the hyper-

echoic spinous processes (Figure 3) (2). The probe wasthen placed longitudinally (parallel to the spine) and anappropriate interspace for the procedure was identifiedand marked (Figures 4, 5).

After using ultrasound, the providers would again ratetheir confidence in identifying a needle-insertion point ona Likert scale. Real-time ultrasound was not used duringthe actual lumbar puncture procedure itself, only to assistin identifying the optimal needle insertion site. After thelumbar puncture was performed, the remainder of thequestionnaire was completed by the providers.

RESULTS

A total of 19 patients met the eligibility criteria and wereincluded in the study. The primary end point was meanconfidence score in identifying a needle insertion site

Figure 3. Transverse view with spinous process midline andtransverse processes lateral.

Figure 5. Longitudinal viewof lumbar spinewith interspinousspaces in between the dorsal spinous processes.

Table 1. Patient and Provider Characteristics

Ultrasound LP in Pediatric EM 61

prior to and after using ultrasound. Statistical significancewas assessed using the Wilcoxon signed-rank test (3).This test takes into account the dependence of observa-tions arising from the repeated-measures design (two ob-servations performed on each subject).

Because this test is a nonparametric procedure, it maybe more appropriate for small samples than the para-metric paired t-test (4). Table 1 shows the characteristicsof the patients and of their providers.

Table 2 shows the distribution of the confidence score.The mean confidence score (the primary end point of thestudy) was equal to 2.89 with the landmark procedurealone, and 4.79 with ultrasound assistance to the land-mark technique, yielding an average score difference of1.90 (95% confidence interval 1.23–2.56; Wilcoxonp < 0.001, paired t-test p < 0.001). Thus, compared tothe landmark procedure alone, the use of ultrasoundwas associated with a significantly higher average confi-dence score.

Figure 4. Longitudinal view is parallel to lumbar spine.

Table 3 shows the distribution of the confidence scoreas a categorical variable. Note that the use of ultrasoundresulted in the maximum possible confidence score (ascore of 5 or ‘‘excellent confidence’’) in 84.2% (16/19)of all cases and produced no cases with a score below 3(‘‘average confidence’’). In contrast, the landmark proce-dure was associated with highly variable confidencescores, with 42% falling below the ‘‘average confidence.’’

Table 4 shows the distribution of the differences inconfidence score between the two techniques for all studysubjects. Addition of ultrasound had at least some effecton the confidence score in approximately three-quarters(14/19) of all cases. In the remaining five cases(26.3%), ultrasound and the landmark procedure resultedin the same level of confidence. A change in the selected

Variable Distribution

Patient’s age (days)Mean (SD) 132 (228)Median (range) 60 (35–850)

Patient’s weight (kg)Mean (SD) 6.0 (3.4)Median (range) 4.9 (4.0–16.7)

Patient’s gender, n (%)Female 9 (47.4%)Male 10 (52.6%)

Indication, n (%)Meningitis 8 (42.1%)Fever 8 (42.1%)Other 3 (15.8%)

Position, n (%)Right 13 (68%)Left 6 (32%)

Provider’s level of training, n (%)First-year resident 7 (36.8%)Second-year resident 7 (36.8%)Third-year resident 5 (26.3%)

Table 2. Distribution of the Confidence Score: Mean,Median, SD, and Range

Confidence Score Landmark (n = 19) Ultrasound (n = 19)

Mean 2.89 4.79Median 3 5SD 1.24 0.54Range (1–5) (3–5)

Table 4. Distribution of the Confidence Score Differences(Ultrasound–Landmark)

Score Difference No. of Patients %

0 5 26.31 1 5.32 6 31.63 5 26.34 2 10.5

62 S. Kim and D. K. Adler

needle insertion point after the use of ultrasound occurredin 10 of 19 (53%) of all cases. In 3 of 19 cases (16%), thespinal fluid could not be collected on the first attempt,although the specimen was successfully collected onthe second attempt. An unsuccessful attempt was definedas withdrawing the spinal needle completely from theskin and then reinserting it.

DISCUSSION

The most important finding of this study was that ultra-sound assistance was associated with a higher averageconfidence score in identifying a lumbar puncture inser-tion point than the traditional landmark technique alone.This finding helps substantiate the use of ultrasound in thepediatric ED as a tool to assist with the technique of pe-diatric lumbar puncture.

A study from the pediatric radiology literature inves-tigated the use of ultrasound in the pediatric setting incases of failed lumbar puncture to determine if ultra-sound could help identify the cause of procedure failureand to determine whether further intervention, such asfluoroscopy, was warranted (1). This study found that ul-trasound guidance in neonates and infants after failedlumbar punctures can identify abnormal epidural orintrathecal hematomas in three planes, compared to fluo-roscopy in one plane, and prevent the risk of radiationwith fluoroscopy (1).

Our study incorporated the use of a linear probe thathas been utilized in prior studies of ultrasound evaluationof the lumbar spine (5,6). The linear probe is a high-frequency probe with a short focal length, making itoptimal for visualization of structures that are close tothe surface of the skin. Linear probes are widely used in

Table 3. Distribution of the Confidence Score as aCategorical Variable

Confidence Score

Landmark (n = 19) Ultrasound (n = 19)

No. of Patients % No. of Patients %

1 2 10.5 0 02 6 31.5 0 03 6 31.5 1 5.34 2 10.5 2 10.55 3 15.8 16 84.2

emergency ultrasound, so this technique could be utilizedin many institutions.

The use of ultrasound-assisted lumbar puncture in theadult setting has been evaluated in obese patients and hasbeen found to reduce the number of failed attempts whencompared to the landmark technique alone (7). Ultra-sound may be similarly helpful in the pediatric emer-gency setting, as childhood obesity is increasing. Otherstudies in the adult setting have looked at using ultra-sound to identify the impact of patient positioning onthe size of the interspinous space (8). In a recent studyin the pediatric setting, measurements of the interspinousspace were taken with bedside ultrasound, with theconclusion that the interspinous space was maximizedwith the child sitting with hips flexed, as compared toother patient positions (9).

Limitations

Several limitations should be considered when interpret-ing the results of our study. First, our data are limited intheir precision due to a small sample size. In addition,our study focused on residents who may have had limitedexperience with pediatric lumbar puncture. Third, our in-clusion criteria were broad, enrolling all pediatric emer-gency patients # 18 years of age, with no stratificationbased on age. Our study subjects had a mean age of 132days (approximately 4 months and 12 days). Our resultsmight have been different with older pediatric patientsif needle insertion site selection for lumbar puncture us-ing the landmark technique alone was easier due to theirlarger interspinous spaces. Older patients also may moveless than younger children, which would make it easierto perform a successful lumbar puncture regardless ofwhether or not ultrasound assistance was used. An appro-priate next study to build on our results would be torandomize pediatric patients undergoing lumbar punctureto either ultrasound assistance or landmark techniquealone and compare them.

CONCLUSION

Our study is the first we are aware of to investigate the useof ultrasound assistance to enhance the technique of

PLEASE FILL IN PRIOR TO LUMBAR PUNCTURELevel of Training: _________________________________________How many times have you used the ultrasound machine: ______Patient Age: _____________________________________________Patient Gender: __________________________________________Patient height (if available): ________________________________Patient weight (if available): ________________________________Indication for LP: _________________________________________Rate your confidence in selecting a spinal needle insertion site

using the landmark technique in this pediatric patient:None Average Excellent1 2 3 4 5Rate your confidence in selecting a spinal needle insertion site

after using ultrasound in addition to the landmark technique inthis pediatric patient:

None Average Excellent1 2 3 4 5Did your landmark change after using ultrasound: ______________________________________________________________________PLEASE FILL IN AFTER LUMBAR PUNCTUREPosition of patient for LP: __________________________________Use of anesthetic to LP site: ________________________________Need for conscious sedation: _______________________________Number of attempts for successful LP: ______________________Able to perform successful LP?: ____________________________If unable to complete LP, specify reason: ____________________

Ultrasound LP in Pediatric EM 63

lumbar puncture in the pediatric emergency setting. Ul-trasound can be a useful tool in pediatric lumbar punctureby increasing provider confidence. This study describes asimple technique and demonstrates that ultrasound assis-tance increases the confidence of resident physicians inidentifying an insertion site for lumbar puncture in pedi-atric patients.

REFERENCES

1. Coley BD, Shiels WE 2nd, Hogan MJ. Diagnostic and interventionalultrasonography in neonatal and infant lumbar puncture. PediatrRadiol 2001;31:399–402.

2. Murphy M, Nagdev A, Solomon RC. Focus on: ultrasound-guidedlumbar puncture. ACEP News 2007;September: 23–25.

3. Lehmann EL. Nonparametrics: statistical methods based on ranks.San Francisco: Holden-Day; 1998.

4. Conover WJ. Practical nonparametric statistics. 3rd edn. New York:John Wiley & Sons; 1999.

5. Chen CP, Tang SF, Hsu TC, et al. Ultrasound guidance in caudalepidural needle placement. Anaesthiology 2004;101:181–4.

6. Moon SH, Park MS, Suk KS, et al. Feasibility of ultrasound exami-nation in posterior ligament complex injury of thoracolumbar spinefracture. Spine 2002;27:2154–8.

7. Nomura JT, Leech SJ, Shenbagamurthi S, et al. A randomized controltrial of ultrasound-assisted lumbar puncture. J UltrasoundMed 2007;26:1341–8.

8. Sandoval M, Shestak W, Sturmann K, Hsu C. Optimal patientposition for lumbar puncture, measured by ultrasonography. EmergRadiol 2004;10:179–81.

9. Abo A, Chen L, Johnston P, Santucci K. Positioning for lumbar punc-ture in children evaluated by bedside ultrasound. Pediatrics 2010;125:1149–53.

APPENDIX: QUESTIONNAIRE UTILIZING ALIKERT SCALE TO ASSESS PROVIDERCONFIDENCE USING THE LANDMARKSCOMPARED TO ULTRASOUND-ASSISTED

LANDMARKS

64 S. Kim and D. K. Adler

ARTICLE SUMMARY

1. Why is this topic important?This topic is important because the current use of ultra-

sound in the pediatric emergency department (ED) is ex-panding, so it will potentially serve as a valuable tool toassist in identifying a needle insertion point for pediatriclumbar puncture.2. What does this study attempt to show?

This study describes a simple technique using a linearprobe and evaluates whether ultrasound increases pro-vider confidence in identifying an insertion point for lum-bar puncture in pediatric emergency patients.3. What are the key findings?

The key finding of this study was that, compared to thelandmark procedure alone, using ultrasound to assist inidentifying an insertion point was associated with a sig-nificantly higher average confidence score. Using theWilcoxon signed-rank test for comparison, the mean con-fidence score (based on a five-point Likert scale) was 2.89,with the landmark procedure alone and 4.79 with ultra-sound assistance, yielding an average score differenceof 1.90 (95% confidence interval 1.23–2.56; Wilcoxonp < 0.001, paired t-test p < 0.001).4. How is patient care impacted?

Patient care is impacted because the use of ultrasoundto assist in lumbar puncture in the pediatric ED could in-crease provider confidence in identifying key landmarks,which may prevent radiation exposure by obviating theneed to undergo fluoroscopy-guided lumbar puncture. Pa-tient care also may be improved if the use of ultrasound toassist in lumbar puncture decreases both patient andparent anxiety by minimizing the number of attemptsand length of time required to identify an insertion sitefor the needle.