• 
• 

Three Part Question

In [obese and non-obese adults patients who need a lumbar puncture], should [ultrasound] be [used to localize the right space and improve first attempt success and reduce rate of traumatic tap]?

Clinical Scenario

While working the night shift, a patient is admitted for thunderclap headache 12 hours ago. After a negative head CT, you decide to do a lumbar puncture to rule out subarachnoid haemorrhage. You specifically need a non-traumatic LP and you wonder if localizing the right lumbar space with an ultrasound would reduce the rate of traumatic LP.

Search Strategy

Pubmed via medline was searched, details are as follows :

Reseach strategy: ((((((("Spinal Puncture"[Mesh]) OR spinal puncture*) OR lumbar puncture*) OR spinal tap*) OR lumbar tap*) OR LP)) AND (((((((((((((ultrasonography) OR medical sonography) OR ultrasonic imaging) OR echography) OR ultrasonic diagnos*) OR ultrasound diagnos*) OR ultrasound-assisted) OR echo-guided) OR POCUS) OR Point of care ultrasound) OR "Ultrasonography [Mesh])) OR U/S) = 506 results

Embase was searched, details are as follows :
('lumbar puncture'/exp OR "lumbar puncture" OR "spinal tap" OR "lumbar tap" OR "LP") AND ('ultrasound'/exp OR 'ultrasound' OR 'ultrasonography' OR 'echography' OR 'ultrasonic' OR 'pocus' OR 'point of care ultrasound' OR 'u/s' OR 'echo-guided')
References :
After reading references to relevant articles, 1 was found relevant

Bestbet.org :

One bestbet was done in 2007 with a neutral conclusion but many articles was found after 2007 and an update of the literature was considered necessary.

Clinicaltrials.gov :

4 trial were found on the subject but 2 are still recruiting, 1 was suspended, 1 the status was unknown.

After titles reading and duplicate eliminated : 34 results
After abstract/article reading : 9 articles

Search Outcome

After reading all 9 articles and evaluating the best evidence available a total of 4 relevant papers were considered in this review.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Peterson et al. 2014 USA	Adult ongoing LP for any clinical indication. In an emergency department (ED) setting. Randomization to the US group or the palpation landmarks group (PL). The US group: the right space was priory localized with the standard US procedure then the usual LP procedure.	Prospective randomized controlled trial	Number of needle reinsertion	No significant difference (3 in the US group vs 5 in the PL group)	The major weakness of this study is the training time of operator. An independent investigator trained them at the time of enrolment immediately prior to do the procedure. This would underestimate the results. No blinding. Convenience sample.
			Success of LP (return of a non traumatic CSF)	No significant difference (78% in the US group vs 76% in the PL group)
			Pain measured on a VAS	No significant difference (1.1 in the US group vs 3.0 in the PL group
			Patient satisfaction (5 point likert scale)	No significant difference (5 vs 5)
			Median time of procedure	No significant difference (1.6 min in the US group vs 2.0 min in the PL group)
Mofidi et al. 2013 Iran	Adults >18 years old <60, who needed an LP. Exclusion: cerebral lesion, local infection, pregnant women, coagulopathy, previous spinal surgery. In an emergency department setting. Patients were randomized to PL group or to US group.	Prospective randomized clinical trial	Number of attempts	Reduced number of attemps (1 in the US group vs 2 in the PL group, p=-0.047)	No blinding of patients nor LP operator. This could overestimate the results.
			Number of traumatic LP	Reduced number of traumatic tap (5 in the US group vs 18 in the PL group, p=0.024)
			Pain score with a numerical 0 to 10 scale	In the US guided LP, the pain score was significantly lower (4.4 vs 7.4).
			Procedure time	Time (minutes) was significantly lower in the US group 3.3 vs 6.4 min
Shaikh et al. 2013 Canada	Inclusion criteria for randomized control trial: - Randomization - Comparing US marking vs palpation procedure Reported at least one primary outcomes, Clinical trial, observation study and diagnostic studies were included. 14 trials were included, 5 for lumbar puncture, 674 patients in the US group and 660 in the PL group.	Systematic review and meta-analysis	Failed procedure	Reduced failed procedure, 6 in the US group vs 44 in the PL group (RR 0,21 [CI 95% 0.10-0.43]), NNT 16 to reduce one failure	Search strategy seems to be not enough sensible; the grey literature was not searched. 7 studies were done in an anaesthesia settings.
			Traumatic procedure	Reduced traumatic procedure (RR 0.27 [CI 95% 0.11-0.67])
			Number of needle reinsertion	Reduced number of reinsertion needle by a mean of -0.44 (-0.64 to -0.24)
			Number of needle redirection	Reduced number of redirection by a mean of -1.00 per procedure (-1.24 to -0.75, p<0.001)
			Time of the procedure	Time of the procedure could not be evaluated due to heterogeneity.
Nomura et al. 2007 USA	Adults >18 years old, who needed a LP; In an (ED) setting. All patients had a prior ultrasound to localize the right space and marked with an ultraviolet (UV) ink in a separate room, the LP operator then localized the right space by placing an UV ink by palpation landmarks. Randomization occurred after. A different investigator showed the mark to use based on the randomization results to the LP operator who was blinded to the origin of the mark.	Randomized controlled double-blind study	Number of attempts	No significant difference in number of attempts (2 vs 2)	Subgroup analysis very interesting but probably lacks power due to a small sample size. For the US procedure, the localization of the right angle was not included in the procedure.
			Presence of a traumatic tap	No significant difference (RR=1.04 [CI 95% 0.83-1.31])
			Success or failure of procedure	Increased success rate for US (RR 1.32 [CI 95% 1.01-1.72]), 1 failed in the US and 6 in the PL procedure
			Length of operation	No significant difference (15 min in the US group vs 10.5 in the PL group)
			Evaluation of the procedure (ease by the LP operator on a 10 cm VAS scale)	No significant difference (3 in the US group vs 5.2 in the PL group)
			Subgroup analysis for obese patients with a BMI >30	Ease of procedure significantly better (2.7 cm for US vs 6.9 cm for PL), Others results had no significant difference but a trend toward US procedure

Comment(s)

This review showed that with an appropriate training (which could be very brief because the technique is very easy to learn), ultrasound-assisted LP results in a better success rate. An extensive literature is available concerning paediatric populations, which showed a significant advantage to use the ultrasound (USS) prior to LP. Because doing LPs in paediatrics patients is relatively simple with the standard palpation landmark technique, doing a review about adult patients seemed more relevant, especially with obese patients. Except for the first study, this review was well designed and was done in emergency settings, which facilitate external validation. An interesting result is the median time of procedure. We could think that ultrasound-assisted procedure would require more time, but in fact, with better success rate and reduced needle reinsertion, the procedure takes less time or equal amount of time. This is a very important result to consider for emergency physicians. Finally, the USS training usually includes the evaluation of the right angulation of the needle for the procedure. None of these studies used this important technique. This would probably change the results in favour of USS. Future research on this subject, specifically enrolling obese adults as an inclusion criteria, and a study of cost-effectiveness, would be very interesting.

Editor Comment

BMI, body mass index; CSF, cerebrospinal fluid; LP, lumbar tap; NNT, number needed to treat; PL, palpation landmark; RCT, randomised controlled trial; RR, relative risk; USS, ultrasound; VAS, visual analogue scale.

Clinical Bottom Line

Using ultrasound-assisted landmarks prior to LP procedure improves success rate and reduces number of attempts and traumatic tap.

References

1. Peterson MA , Pisupati D , Heyming TW , et al . Ultraound for routine lumbar puncture. Acad Emerg Med 2014;21:130–6.
2. Mofidi M , Mohammadi M , Saidi H , et al . Ultrasound guided lumbar puncture in emergency department: time saving and less complications. J Res Med Sci 2013; 18: 303–7
3. 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
4. Nomura JT, Leech SJ, Shenbagamurthi S et al. A randomized controlled trial of ultrasound-assisted lumbar puncture. J Ultrasound Med 2007;26:1341–8.