Ultrasound guidance for upper and lower limb blocks


Nerve blocks are used to numb all or part of the arms or legs (peripheral blockade) for surgery, or to provide pain relief after the operation, or both. Using ultrasound, anaesthetists can 'see' vital structures below the skin, which should allow them to place the local anaesthetic injection accurately and avoid damaging other tissues or organs. We aimed to assess whether ultrasound has any advantages over other nerve-locating techniques for nerve blocks of the arms or legs in adults.

Study characteristics

The evidence is current up to 27 August 2014. We found 32 studies with 2844 participants. Most studies compared ultrasound with electrical nerve stimulators or compared ultrasound combined with nerve stimulators against nerve stimulators alone. We reran the search in May 2015. We will deal with the 11 studies of interest when we next update the review.

Key results

We combined the results of studies using statistical tests and found that nerve blocks were more likely to be assessed as adequate for surgery and were less likely to need additional anaesthetic when performed using ultrasound guidance or ultrasound guidance combined with other techniques. We also found that there were fewer complications such as 'pins and needles' or accidental punctures of blood vessels. It also took less time to perform the nerve block when ultrasound alone was used.

Quality of the evidence

There was variation in the quality of the studies and authors had not always made sufficient attempts to ensure that the outcome assessors were unaware of what technique had been used for the nerve block. Studies had also often not clearly explained how experienced the people giving the nerve block were. This is particularly important, as ultrasound is still a relatively new technique and some anaesthetists may have limited experience. We rated our evidence for whether the nerve blocks were sufficient and adequate for surgery as of moderate quality, but evidence for our other outcomes was either low or very low.


Our evidence suggests that ultrasound is superior to other techniques for peripheral nerve blocks. However, we are unable to say whether this result depends on the experience of the practitioner in the technique being used.

Authors' conclusions: 

There is evidence that peripheral nerve blocks performed by ultrasound guidance alone, or in combination with PNS, are superior in terms of improved sensory and motor block, reduced need for supplementation and fewer minor complications reported. Using ultrasound alone shortens performance time when compared with nerve stimulation, but when used in combination with PNS it increases performance time.

We were unable to determine whether these findings reflect the use of ultrasound in experienced hands and it was beyond the scope of this review to consider the learning curve associated with peripheral nerve blocks by ultrasound technique compared with other methods.

Read the full abstract...

Peripheral nerve blocks can be performed using ultrasound guidance. It is not yet clear whether this method of nerve location has benefits over other existing methods. This review was originally published in 2009 and was updated in 2014.


The objective of this review was to assess whether the use of ultrasound to guide peripheral nerve blockade has any advantages over other methods of peripheral nerve location. Specifically, we have asked whether the use of ultrasound guidance:

1. improves success rates and effectiveness of regional anaesthetic blocks, by increasing the number of blocks that are assessed as adequate

2. reduces the complications, such as cardiorespiratory arrest, pneumothorax or vascular puncture, associated with the performance of regional anaesthetic blocks

Search strategy: 

In the 2014 update we searched the Cochrane Central Register of Controlled Trials (CENTRAL; 2014, Issue 8); MEDLINE (July 2008 to August 2014); EMBASE (July 2008 to August 2014); ISI Web of Science (2008 to April 2013); CINAHL (July 2014); and LILACS (July 2008 to August 2014). We completed forward and backward citation and clinical trials register searches.The original search was to July 2008. We reran the search in May 2015. We have added 11 potential new studies of interest to the list of 'Studies awaiting classification' and will incorporate them into the formal review findings during future review updates.

Selection criteria: 

We included randomized controlled trials (RCTs) comparing ultrasound-guided peripheral nerve block of the upper and lower limbs, alone or combined, with at least one other method of nerve location . In the 2014 update, we excluded studies that had given general anaesthetic, spinal, epidural or other nerve blocks to all participants, as well as those measuring the minimum effective dose of anaesthetic drug. This resulted in the exclusion of five studies from the original review.

Data collection and analysis: 

Two authors independently assessed trial quality and extracted data. We used standard Cochrane methodological procedures, including an assessment of risk of bias and degree of practitioner experience for all studies.

Main results: 

We included 32 RCTs with 2844 adult participants. Twenty-six assessed upper-limb and six assessed lower-limb blocks. Seventeen compared ultrasound with peripheral nerve stimulation (PNS), and nine compared ultrasound combined with nerve stimulation (US + NS) against PNS alone. Two studies compared ultrasound with anatomical landmark technique, one with a transarterial approach, and three were three-arm designs that included US, US + PNS and PNS.

There were variations in the quality of evidence, with a lack of detail in many of the studies to judge whether randomization, allocation concealment and blinding of outcome assessors was sufficient. It was not possible to blind practitioners and there was therefore a high risk of performance bias across all studies, leading us to downgrade the evidence for study limitations using GRADE. There was insufficient detail on the experience and expertise of practitioners and whether experience was equivalent between intervention and control.

We performed meta-analysis for our main outcomes. We found that ultrasound guidance produces superior peripheral nerve block success rates, with more blocks being assessed as sufficient for surgery following sensory or motor testing (Mantel-Haenszel (M-H) odds ratio (OR), fixed-effect 2.94 (95% confidence interval (CI) 2.14 to 4.04); 1346 participants), and fewer blocks requiring supplementation or conversion to general anaesthetic (M-H OR, fixed-effect 0.28 (95% CI 0.20 to 0.39); 1807 participants) compared with the use of PNS, anatomical landmark techniques or a transarterial approach. We were not concerned by risks of indirectness, imprecision or inconsistency for these outcomes and used GRADE to assess these outcomes as being of moderate quality. Results were similarly advantageous for studies comparing US + PNS with NS alone for the above outcomes (M-H OR, fixed-effect 3.33 (95% CI 2.13 to 5.20); 719 participants, and M-H OR, fixed-effect 0.34 (95% CI 0.21 to 0.56); 712 participants respectively). There were lower incidences of paraesthesia in both the ultrasound comparison groups (M-H OR, fixed-effect 0.42 (95% CI 0.23 to 0.76); 471 participants, and M-H OR, fixed-effect 0.97 (95% CI 0.30 to 3.12); 178 participants respectively) and lower incidences of vascular puncture in both groups (M-H OR, fixed-effect 0.19 (95% CI 0.07 to 0.57); 387 participants, and M-H OR, fixed-effect 0.22 (95% CI 0.05 to 0.90); 143 participants). There were fewer studies for these outcomes and we therefore downgraded both for imprecision and paraesthesia for potential publication bias. This gave an overall GRADE assessment of very low and low for these two outcomes respectively. Our analysis showed that it took less time to perform nerve blocks in the ultrasound group (mean difference (MD), IV, fixed-effect -1.06 (95% CI -1.41 to -0.72); 690 participants) but more time to perform the block when ultrasound was combined with a PNS technique (MD, IV, fixed-effect 0.76 (95% CI 0.55 to 0.98); 587 participants). With high levels of unexplained statistical heterogeneity, we graded this outcome as very low quality. We did not combine data for other outcomes as study results had been reported using differing scales or with a combination of mean and median data, but our interpretation of individual study data favoured ultrasound for a reduction in other minor complications and reduction in onset time of block and number of attempts to perform block.