The estimated incidence of peripheral nerve injury during regional anesthesia is 0.5% to 10%.1 Most injuries are transient and present as mild paresthesia,2 whereas permanent nerve injury is rare.3,4 Nevertheless, an analysis of American Society of Anesthesiologists closed claims from 1990 to 2007 showed 19% of claims related to regional anesthesia and 22% of claims related to nerve injury. Specifically, 15% of these complications were associated with peripheral nerve blocks.5
Although motor response to electrical stimulation (sensitivity, 74.5%) remains an important means of needle guidance in clinical practice2 and a stimulation current of ≤0.2 mA might suggest intraneural placement of stimulating needle, stimulation currents of >0.2 do not rule out an intraneural position of the stimulating needle.6 Recently, ultrasonographic (US) guidance has increasingly become invaluable in determining accuracy of needle placement during peripheral nerve blockade. However, Koff et al.7 showed that the use of US cannot completely prevent nerve injury but only provide visual evidence of the injury. In addition, continuous visualization of the needle tip may be technically challenging based on experience and anatomical considerations.
Electrical impedance (EI) changes could be useful in identifying intraneural needle placement in regional anesthesia; the clear changes in tissue characteristics between the extraneural and intraneural compartments may be identifiable as the needle traverses these compartments. The availability of continuous EI values in modern commercial nerve stimulators makes it easy to measure but little data exist to aid its interpretation. Recently, Tsui et al.8 have shown a detectable difference in EI measured between the extraneural and intraneural compartments in porcine sciatic nerves. Furthermore, Bardou et al.,9 in a study of 140 peripheral nerve blocks, showed that the EI in the 21 suspected cases of accidental nerve puncture was greater by an average of 4.3%. However, data on impedance measurement after definitive intraneural placement in human peripheral nerves are lacking complicated by the fact that such measurement in live humans may be construed as unethical.
We hypothesized that there is a detectable and significant difference in EI when needle tip progresses from extraneural to intraneural compartments in human peripheral nerves. This investigation studied the neural compartments of freshly amputated patients. The use of freshly amputated human limbs makes it possible to measure the intraneural EI providing a human sample size in lieu of previously described animal models. We analyzed both initial US visualization to measure EI in the intraneural compartment and then reconfirm with EI measurements after open dissection and nerve exposure. In addition, we performed multiple measurements over the same neural segment to see whether EI values remain consistent.
After obtaining institutional review board approval (Montefiore Medical Center, November 5, 2012), patients who were scheduled for lower extremity amputations were approached and enrolled in the study using a written informed consent process. The patients were anesthetized according to the preference and the discretion of the attending anesthesiologist as per the standard protocol for our institution. Immediately after the surgical amputation was completed, the research team obtained the amputated limb, and the rest of the study was conducted in the pathology department.
The amputated limb was placed on a tray, and under ultrasound imaging guidance (M-Turbo, HFL 5–12 MHz linear array; Sonosite Inc., Bothell, WA), an insulated peripheral block needle primed with normal saline (22-gauge, 50-mm StimuQuik insulated needle; Arrow International, Inc., Reading, PA) was passed through the subcutaneous tissue using an in-plane approach until the tip of the needle was visualized 1 to 2 mm outside the nerve bundle (Fig. 1). An independent observer recorded EI and motor response characteristics at this site using a nerve stimulator (2 Hz, 0.1 ms, 0.5 mA) that continually displays EI (Stimuplex HNS 12; B. Braun Medical, Bethlehem, PA). Then, the needle was further advanced under visualization to place the needle tip intraneurally (Fig. 1), and again EI measurements were recorded. Three such measurements were obtained within a 1-cm length of the nerve from the site of the initial entry.
Subsequently, after the needle was removed, the visualized nerve was then exposed by dissection. In this portion of the study, the insulated needle was placed initially in contact with the nerve but in the extraneural compartment. EI and motor response characteristics were recorded. Then the needle was advanced until the needle tip was seen to enter the nerve into the intraneural compartment, and EI values were recorded again. Again, 3 such measurements were done within a 1-cm length of the nerve.
The nerves that were studied were the sciatic nerve in the popliteal fossa in above-knee amputations or the tibial nerve below the calf in below-knee amputations. All measurements were obtained within 45 minutes of amputation.
This study was designed as a descriptive study, and no a priori sample size calculations were performed. Descriptive statistical methods were used to describe the study population. For all continuous baseline characteristics, mean ± SD was reported, as were proportions for categorical variables.
To test the differences in the impedance measurements between intraneural and extraneural needle placement, we performed repeated-measures analysis of variance and were compared using PROC GLM in SAS for Windows version 8.2 (SAS Institute Inc., Cary, NC). The model was built with repeated impedance measurements as the dependent variable and location of the needle tip (intraneural versus extraneural) as independent variable. Impedance measurements were analyzed separately for the ultrasound-guided needle placement and open dissection needle placement as well as for above-knee amputation and below-knee amputation specimens.
EI values were collected from a total of 11 amputated lower extremity limbs. Measurement at 1 site (either above knee or below knee) was done in each specimen. One skilled regional anesthesiologist performed all needle placements. Different independent observers performed all impedance measurements, and all measurements were recorded in a consistent manner. The point estimate of intraclass correlation coefficients (ICCs) for the impedance measurements were at least 0.58 for all EI measurements (Table 1). However, because of the small sample size, the width of confidence intervals of the ICC for the intraneural impedance measurements after surgical dissection in the above-knee amputation group was large, and we could not assume consistency.
In the below-knee amputated extremity (tibial nerve, n = 6) specimens, mean ± SD of ultrasound-guided extraneural impedance for all the 18 punctures was 6 ± 1.6 kΩ. In these same specimens, the mean ± SD of extraneural impedance after open dissection for all 18 punctures was 4.8 ± 1.5 kΩ. The mean intraneural impedance values in the below-knee amputated extremities for ultrasound-guided and open dissection needle placements were 10 ± 2 and 8.4 ± 1.8 kΩ, respectively.
In the above-the-knee amputated extremity (sciatic nerve, n = 5) specimens, mean extraneural impedance values for ultrasound-guided and open dissection approaches in needle placement were 25.2 ± 5.3 and 26 ± 2.6 kΩ. The mean intraneural impedance values for ultrasound-guided and open dissection approaches in needle placement in these same specimens were 35.2 ± 7.9 and 36.7 ± 2.5 kΩ, respectively. In summary, the mean values of extraneural EI were significantly lower than the measured EI when the needle traversed intraneurally with both ultrasound-guided and open dissection approaches in needle placement regardless of the level of amputation (Fig. 2).
In the below-knee amputated extremity specimens based on the ultrasound methods, mean value for ultrasound-guided intraneural impedance was 10 ± 2 kΩ compared with an extraneural impedance of 6 ± 1.6 kΩ (P = 0.005). Likewise, the difference after open dissection was also significant when we repeated the analysis based on the same specimens (P = 0.005) but the ICC of the intraneural impedance measurements after surgical discussion exhibited wide confidence intervals (Table 1). Similarly, in the above-the-knee amputated extremity specimens, mean intraneural impedance was 35.2 ± 7.9 kΩ compared with an extraneural impedance of 25.2 ± 5.3 kΩ (P = 0.037).a Again, the difference obtained after open dissection was also significant when we repeated the analysis based on the same specimens (P = 0.0002).
This study demonstrated that there is a clearly detectable difference in EI measurements between the extraneural and intraneural compartments regardless of nerve location in human peripheral nerves. Our results are similar to those in a porcine sciatic nerve model,8 with an increase in EI observed when the needle tip traverses the epineurium and are also consistent with the findings of Bardou et al.9 in patients with suspected accidental nerve puncture during peripheral nerve blockade.
In our study, we tried to simulate normal ultrasound-guided regional anesthetic approaches in needle placement by measuring the EI changes between the extraneural and intraneural compartments followed by confirmation of the measurements on the same nerve segment after exposure by open dissection. The variation observed in absolute values of EI between the above-knee amputated limbs (sciatic nerve: mean EI, 26 kΩ) and below-knee-amputated limbs (tibial nerve: mean EI, 10 kΩ) could be because of differing tissue characteristics at these sites. Indeed, a similar study evaluating the extraneural EI of the median nerve at the axilla and elbow showed a considerable variation in the measured EI between the 2 regions.10 However, in our study, regardless of the site, the EI increased substantially when the needle progressed into the intraneural compartment. The increase was sustained and consistent with both approaches to EI measurement. Thus, we conclude that the observed increases in EI with intraneural placement can likely be extrapolated to any peripheral nerve location and can be reliably observed in ultrasound-guided peripheral nerve blockade.
The amperage required to reliably stimulate a nerve change is based on the extraneural EI, as highlighted in the study by Sauter et al.10 The difference in extraneural impedance at different peripheral nerve sites seen in our study emphasizes that use of a cutoff of 0.2 or 0.4 mA in predicting intraneural placement would be unreliable.11,12 However, nerve stimulation in conjunction with measures of EI could improve monitoring and aid in avoiding intraneural needle placement.
Our study has certain limitations. Obviously, the study could not be done in live humans because of the potential for patient risk and was done in amputated extremities. Measurements were optimized as best as possible by completing the study within 45 minutes of amputation. The effects of ischemia on nerve conduction and excitability are variable.13,14 However, we did achieve appropriate motor response to stimulation at the time of obtaining EI values in all the amputated specimens, which shows that the electrical conduction was still intact. Another confounding factor is the presence of diabetic neuropathy in a subset of patients (60%), which may have affected EI values. However, we found no significant difference in the EI measured in patients with and without diabetic neuropathy, consistent with previous observations.9 Finally, all patients received general anesthetic. Anesthetics are thought to act mainly at synapses, and although there is evidence that nerve conduction characteristics may be affected by general anesthesia, there is no known effect on nerve impedance.15,16
It could be argued that the intraneural EI measured using ultrasound guidance may have been affected by the ability to visualize clear penetration of the nerve by the needle. However, this approach resembles clinical situations, and our EI measurements were consistent across readings and comparable with those obtained using open dissection. We believe the statistically insignificant difference in both intraneural and extraneural EI obtained from the tibial nerve between the 2 approaches may be related to enhanced nerve exposure after open dissection that may have caused lower EI values, especially in the extraneural EI.17 Furthermore, the results of intraneural impedance measured after surgical dissection in the above-knee amputation group need to be cautiously interpreted because the true value of ICC may be as low as 0.13.
Finally, although the needle tip was placed in the intraneural compartment for recording intraneural EI values, we are unable to determine whether the placement was intrafascicular or extrafascicular. There is no modality that could be reliably used to make this determination. Although many studies emphasize that permanent nerve injury is mainly caused by intrafascicular injection,18,19 animal studies have shown that extrafascicular injection of ropivacaine also causes focal demyelination and edema to the endoneurium, although it is less than that observed after intrafascicular injection.20 Therefore, the safety of peripheral nerve blocks may be best improved by avoiding intraneural injections altogether.
In conclusion, peripheral nerve injury remains an important complication. Currently, there is no reliable gold standard to predict or prevent intraneural needle placement. Measurement of the EI in conjunction with nerve stimulation may serve as another tool to recognize possible intraneural needle placement during peripheral nerve blockade.
Name: Amaresh Vydyanathan, MBBS, MS.
Contributions: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Amaresh Vydyanathan has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Boleslav Kosharskyy, MD.
Contributions: This author helped analyze the data and write the manuscript.
Attestation: Boleslav Kosharskyy reviewed the analysis of the data and approved the final manuscript.
Name: Singh Nair, MBBS.
Contributions: This author helped design the study and analyze the data.
Attestation: Singh Nair has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Karina Gritsenko, MD.
Contributions: This author helped write the manuscript.
Attestation: Karina Gritsenko reviewed the analysis of the data and approved the final manuscript.
Name: Ryung S. Kim, PhD.
Contributions: This author led and interpreted the statistical analysis and helped write the manuscript.
Attestation: Ryung S. Kim reviewed the analysis of the data and approved the final manuscript.
Name: Dan Wang, MS.
Contributions: This author performed statistical analysis.
Attestation: Dan Wang reviewed the analysis of the data and approved the final manuscript.
Name: Naum Shaparin, MD.
Contributions: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Naum Shaparin has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
This manuscript was handled by: Terese T. Horlocker, MD.
The authors thank Ms. Kea Alexa Moncada, BS, Department of Anesthesiology, Montefiore Medical Center, for her help with the graphs and figures. The authors thank Zeshan Khan, DO (Department of Anesthesiology), for his help with organizing the data.
a We performed visual residual analysis to see whether the model assumption was appropriate and whether there was an outlier. There were overall 4 repeated-measures analyses of variance because there were 2 locations that are below and above knee and 2 methods that are surgical or ultrasound. This led us to checking 4 residual plots and 4 QQ plots. We are submitting these 8 plots as Supplemental Digital Content (http://links.lww.com/AA/B416). The residual plots show the studentized residuals and the sole binary independent variable. There are clearly no outliers. The residual plots show that the assumption of normality and constant variance was appropriate. All 4 residual sets passed the test of normality; all passed using 5% level of significance. There were some differences of the variance on each value of the independent variable, but because of the modest sample size, formal tests of model assumption would not be significant.
1. Liguori GAComplications of regional anesthesia: nerve injury and peripheral neural blockade.J Neurosurg Anesthesiol20041684–6
2. Jeng CL, Torrillo TM, Rosenblatt MAComplications of peripheral nerve blocks.Br J Anaesth2010105suppl 1i97–107
3. Fremling MA, Mackinnon SEInjection injury to the median nerve.Ann Plast Surg199637561–7
4. Auroy Y, Narchi P, Messiah A, Litt L, Rouvier B, Samii KSerious complications related to regional anesthesia: results of a prospective survey in France.Anesthesiology199787479–86
5. Metzner J, Posner KL, Lam MS, Domino KBClosed claims’ analysis.Best Pract Res Clin Anaesthesiol201125263–76
6. Moayeri N, Groen GJDifferences in quantitative architecture of sciatic nerve may explain differences in potential vulnerability to nerve injury, onset time, and minimum effective anesthetic volume.Anesthesiology20091111128–34
7. Koff MD, Cohen JA, McIntyre JJ, Carr CF, Sites BDSevere brachial plexopathy after an ultrasound-guided single-injection nerve block for total shoulder arthroplasty in a patient with multiple sclerosis.Anesthesiology2008108325–8
8. Tsui BC, Pillay JJ, Chu KT, Dillane DElectrical impedance to distinguish intraneural from extraneural needle placement in porcine nerves during direct exposure and ultrasound guidance.Anesthesiology2008109479–83
9. Bardou P, Merle JC, Woillard JB, Nathan-Denizot N, Beaulieu PElectrical impedance to detect accidental nerve puncture during ultrasound-guided peripheral nerve blocks.Can J Anaesth201360253–8
10. Sauter AR, Dodgson MS, Kalvøy H, Grimnes S, Stubhaug A, Klaastad OCurrent threshold for nerve stimulation depends on electrical impedance of the tissue: a study of ultrasound-guided electrical nerve stimulation of the median nerve.Anesth Analg20091081338–43
11. Bigeleisen PE, Moayeri N, Groen GJExtraneural versus intraneural stimulation thresholds during ultrasound-guided supraclavicular block.Anesthesiology20091101235–43
12. Tsai TP, Vuckovic I, Dilberovic F, Obhodzas M, Kapur E, Divanovic KA, Hadzic AIntensity of the stimulating current may not be a reliable indicator of intraneural needle placement.Reg Anesth Pain Med200833207–10
13. Bostock H, Baker M, Grafe P, Reid GChanges in excitability and accommodation of human motor axons following brief periods of ischaemia.J Physiol1991441513–35
14. Kiernan MC, Bostock HEffects of membrane polarization and ischaemia on the excitability properties of human motor axons.Brain20001232542–51
15. Tsui BCThe effects of general anaesthesia on nerve-motor response characteristics (rheobase and chronaxie) to peripheral nerve stimulation.Anaesthesia201469374–9
16. Johnston MF, Simon SA, Ramón FInteraction of anaesthetics with electrical synapses.Nature1980286498–500
17. Gielen FL, Wallinga-de Jonge W, Boon KLElectrical conductivity of skeletal muscle tissue: experimental results from different muscles in vivo.Med Biol Eng Comput198422569–77
18. Hadzic A, Vloka J, Hadzic N, Thys DM, Santos ACNerve stimulators used for peripheral nerve blocks vary in their electrical characteristics.Anesthesiology200398969–74
19. Farber SJ, Saheb-Al-Zamani M, Zieske L, Laurido-Soto O, Bery A, Hunter D, Johnson P, Mackinnon SEPeripheral nerve injury after local anesthetic injection.Anesth Analg2013117731–9
20. Whitlock EL, Brenner MJ, Fox IK, Moradzadeh A, Hunter DA, Mackinnon SERopivacaine-induced peripheral nerve injection injury in the rodent model.Anesth Analg2010111214–20