Evaluation of pedicle screw placement with triggered EMG is probably the most commonly-used test in all of neuromonitoring. Unfortunately, due to lack of standardization, this robust test fails to detect medial pedicle cortex breach more often than it should. As a community of neurophysiologists, we can standardize our methods and improve patient outcomes but we need to work together. This post will take a look at common misconceptions, methodological concerns and solutions for performing the most accurate tests of pedicle screw placement.
Doesn’t everyone test screws in the same way…?
Even though the ASNM has a position paper that covers the subject in part (Leppanen, 2004), there is a serious problem with lack of standardization in our field, and I guess it’s some sort of secret. I mean, how often do you hear people talk about lack of standardization in pedicle screw testing? Almost never, right? Everyone knows that there are no “standards” for motor evoked potentials…you see it everywhere in the literature. But, until recently, lack of standardization in pedicle screw testing was a silent problem. Just this year, an awesome book chapter was published to address this very problem (Schwartz, Bhalodia and Vaccaro, 2014). The authors report that stimulation parameters, screw types, anesthesia, and other factors vary widely between labs. Adding to this problem, most groups fail to fully report their methods when they publish their results. How are we supposed to replicate people’s results if we don’t have all the methods? Jeez. Anyway, most of what I’ve written below, I thought it was all common knowledge, but I’ve been around this field long enough to know that’s simply not true. If there were a such thing as “common knowledge” in our field, I probably wouldn’t write this blog. Anyway, I hope that what I’ve written below is informative and I hope it helps to improve your outcomes. The text below assumes that you already know at least a little about pedicle screw testing.
But the surgeon made me do it that way…
Before we get into this, let’s talk about the elephant in the room. The surgeon can be the most difficult obstacle that you’ll confront in your endeavor to perform an accurate test of pedicle screw placement. Obviously, this isn’t always the case, but we know that there are some surgeons that want things done their way without question. So, even though he/she invited you to perform the test, the surgeon may refuse to perform it properly, even forcing you to use substandard methods. For example, the surgeon may say, “Just put it (intensity) on 8 mA and don’t say word unless you see a response.” In the sections that follow, you’ll see why this approach is so bad, but let’s tackle this problem first.
So, why would the surgeon invite you to the OR just to perform a test incorrectly? Lots of reasons…obviously there is a medico-legal reason for performing the test, but maybe they don’t trust the test, maybe they don’t care, or maybe they don’t know any better. I certainly can’t presume to know, but whatever the reason, what are you supposed to do? If you readily comply, then you are committing malpractice. If the case goes to court, you will be found at fault for not telling the surgeon that his/her methods are flawed. After all, you are the expert in neurophysiology, not the surgeon. However, if you argue with the surgeon, then he/she is likely to yell at you and make you feel belittled in front of the whole OR. Your day will be hell. I know it’s frustrating.
Here’s my take on the subject: I’m not there to please the surgeon. Rather, I’m there to advocate for the patient. I’m a neurophysiologist, I monitor the nervous system, I’m very good at what I do, and no one is going to compromise that. I don’t argue with the surgeon. Instead, I initiate an educated conversation with the surgeon, and I do it before the procedure (not at the time of testing). In my mind, we are establishing a plan for patient care. If he/she wants to turn it into an argument, fine. I don’t yell..I just communicate my points. If the surgeon still won’t comply with my approach, then I do what the surgeon asks; HOWEVER, I document everything that I said in the medical record, and I document what the surgeon said. If the case goes to court, then it will be shown that I did my due diligence.
Now that we’ve cleared the air and acknowledged our biggest obstacle, let’s talk about some common misconceptions in pedicle screw testing.
Myth #1: The EMG test will uncover any pedicle breach.
The test is designed to evaluate the medial pedicle. Surgeons sometimes come to me confused. They say, “The screw was lateralized on CT, but the EMG test was fine, right?” I always have to remind them that the test is not designed to detect lateral breach, only medial breach.
Myth #2: No CMAP up to X mA means the screw placement is good.
Absence of CMAP means absolutely nothing at all. You must increase the intensity of the stimulation until you get a CMAP on every single screw. It not only tells you that the stim/rec system is working, it tells you something important about the nerve…that it’s intact. See below.
Myth #3: All nerve roots will depolarize with about 3.0 mA of current.
This is sooooo important. Keep in mind that the pedicle screw EMG test is based on the premise that the nerve roots are healthy. It usually takes less than 3.0 mA of current to evoke a CMAP when you stimulate a functionally intact nerve root; however, chronically-compressed nerve roots may have much higher stimulation thresholds (Holland et al, 1998). Add metabolic disorders such as diabetes to the mix, and the thresholds can get higher (Toleikis et al., 2000). Under these circumstances, direct electrical stimulation of nerve roots have CMAP thresholds reported between 18-27 mA (ibid, also Schwartz & Devlin, 2007; Schwartz et al., 2003). I have personally observed this, too, on at least two occasions where I delivered over 20 mA to an exposed nerve root before it depolarized and produced a CMAP. This should make you think about the threshold criteria that you use for breach alert. We’ll come back to this.
Myth #4: You can put your return electrode anywhere on the patient.
A poorly-placed return electrode can cause excessive stimulation artifact and increase the possibility of current shunting. When stimulating lumbar pedicle screws, your return electrode should be in the wound, contralateral and caudal to the screw you are testing. When stimulating thoracic screws, your return electrode should be in the wound and directly contralateral to the screw you are testing. That’s right…you have to move the return electrode for every single screw that you test.
Myth #5: You can put your signal ground anywhere on the patient.
If you put your signal ground in the same muscle from which you are recording, your signals will be dampened or eliminated. The signal ground acts as a common mode reference for your recording electrodes. So, it should see the same “noise” as your recording electrodes, but not the same “signal”. If you are recording CMAPs in response to triggered EMG (or MEPs for that matter), you should put your signal ground in a muscle from which you are not recording. This ensures that the noise is rejected, but the signal is maintained.
Myth #6: The surgeon knows exactly what to do, and is actually doing it.
Don’t make this assumption. Keep your eye on the surgeon. Make sure the wound is suctioned out, that the surgeon is stimulating the shaft of the screw, and that the return electrode is properly placed.
Myth #7: You don’t need to do a train-of-four test.
You should do a TOF test whenever you are testing EMG or MEPs. At its heart, the TOF test gives you documentation that there is sufficient conduction across the neuromuscular junction for reliable monitoring of EMG and MEP. In other words, it tells you whether or not your other test results are even valid. You should know that clearance of neuromuscular blockade is not temporally equal around the quadrants of the body. Schwartz et al (2014) showed that the feet can lag behind the hand by as much as 14 minutes. So, you should do your TOF test from the limbs of importance (legs for thoraco-lumbar cases, hands for cervical cases). It is recommended that you should perform all pedicle screw EMG tests when your TOF ratio is > 70%. Just divide the amplitude of twitch 4 by the amplitude of twitch 1, and then multiply by 100.
Myth #8: Triggered EMG is good enough for testing medial breach.
I once heard a story about a medial breach that was missed with tEMG. The screw tested in the 60 mA range (much higher than all the other screws which tested in the 40s). On CT, the screw was medialized and the patient woke up with severe pain and weakness. How did it test so high? Easy…the screw went through the nerve. So, let me pose the question – How else could we have detected that breach, that iatrogenic injury? Say it with me…. MULTI… MODALITY… MONITORING. The nay-sayers will disagree, but I’m telling you, with combined use of MEPs, SSEP and EMG, the nerve root compromise would have been detected.
What stimulation parameters should I use…?
Many neurophysiologists choose a stimulation frequency that will yield rapid responses for the benefit of the surgeon. For legal reasons, you must be able to view and store every single trace, regardless of whether or not a CMAP is present. Also, you need to be able to show which stimulating intensities evoked CMAPs and which did not. So, if you are stimulating at 5 Hz, you should be saving 5 traces/second and the current intensity associated with each trace. Not all commercially-available systems do this, which sucks, but we’re working to fix it. Also, you don’t want to stimulate as such a high frequency that the nerve is forced into absolute refractory and won’t depolarize. Fortunately, this is not so much of a concern. To be safe, I personally I use 3.1 Hz. It gives the nerve plenty of time to recover, it is plenty fast for the surgeons, and I can record the results of every stimulus pulse delivered.
Also called Pulse Width. No one pays much attention to this parameter, but it is soooo important. A short pulse duration (e.g., 50 µsec) requires much higher current for nerve depolarization than a long duration (e.g., 200 µsec). So, for example, if a you stimulate a pedicle screw with a 200 µsec duration stimulus, maybe you get a CMAP with 10 mA intensity. Change the duration t0 100 µsec, and it could take 15 mA to evoke a CMAP. So, it’s not just the intensity that’s important – you have to understand all stimulus parameters. Make sense? This is really important to understand no matter what you are stimulating. If you want to learn more about the basic math behind this principle, go here. For direct nerve stimulation, I recommend that you never go higher than 50-100 µsec. For pedicle screw stimulation, I recommend a pulse width of exactly 200 µsec.
Start at 0.00 mA and increase the intensity until you get a CMAP. Keep in mind all of the information that I mentioned under the Misconceptions section. Also, remember that healthy bone absorbs much more current than diseased bone. I have seen on many, many occasions where all screws tested above 40 or 50 mA. Increase the intensity until you get a CMAP. If you don’t get something by 100 mA (max output on most machines), something is technically wrong or the screw is lateralized. You should be able to safely and accurately stimulate a pedicle screw and obtain results in 10 seconds per screw.
Criteria for Alert:
As I noted above, you need to look at the whole patient presentation before you interpret the results of pedicle screw stimulation tests. Remember that chronic nerve compression and diabetes can significantly elevate the threshold. Also, bone density can alter the threshold. And don’t forget to check your train-of-four! Finally, look at the results of all screw tests. Did one particular screw test at a much different threshold than all of the others? This is an outlier and should be further investigated. In some cases, you may need to directly stimulate the nerve root in order to determine the depolarization threshold and then extrapolate the criterion for determining medial breach of the pedicle.
Using a pulse width of 200 µsec, Schwartz et al (2014) make the following interpretations:
CMAP Threshold and Interpretation:
< 5 mA
- High probability of medial pedicle wall violation.
- Frank screw penetration in proximity to dura and adjacent nerve root
- Moderate probability of medial pedicle wall violation.
- Screw penetration may not be sufficiently deep to place dura/nerve root “at-risk”. Compare CMAP thresholds to the contralateral side and ipsilateral adjacent levels to determine outliers.
- Low probability of medial pedicle wall violation.
- On average, safe to leave in place. Compare CMAP thresholds to the contralateral side and ipsilateral adjacent levels to determine outliers.
≥ 7 mA
- Extremely low probability of medial pedicle wall violation.
- On average, suggests intact cortical bone. Best practice is to compare CMAP thresholds to the contralateral side and ipsilateral adjacent levels to determine outliers.
Questions, comments, thoughts, suggestions, opinions, etc…..? Feel free to leave a comment below!!
- Devlin VJ, Schwartz DM. Intraoperative neurophysiologic monitoring during spinal surgery. J Am Acad Orthop Surg. 2007 Sep;15(9):549-60. Review.
- Holland NR, Lukaczyk TA, Riley LH 3rd, Kostuik JP. Higher electrical stimulus intensities are required to activate chronically compressed nerve roots. Implications for intraoperative electromyographic pedicle screw testing. Spine (Phila Pa 1976). 1998 Jan 15;23(2):224-7.
- Leppanen RE. Intraoperative monitoring of segmental spinal nerve root function with free-run and electrically-triggered electromyography and spinal cord function with reflexes and F-responses. A position statement by the American Society of Neurophysiological Monitoring. J Clin Monit Comput. 2005 Dec;19(6):437-61.
- Schwartz DM, Bhalodia VM & Vaccaro AR (2014). Neurophysiologic detection of medial pedicle wall violation in the lumbar and thoracic spine. In CM Loftus, J Biller & EM Baron (Eds.), Intraoperative Neuromonitoring (pp. 425-437). New York, NY: McGraw-Hill.
- Schwartz DM, Wierzbowski L, Fan D. Sestokas A. (2003). Surgical Neurophysiologic Monitoring. In. A Vaccaro, R Betz & S Zeidman (Eds.), Principles and Practices of Spine Surgery (pp. 115-126). Philadelphia, PA: Mosby.
- Toleikis, JR, Skelly, JP, Carlvin,AO, Toleikis,SC, Bernard, TN, Burkus, JK, Burr, ME, Dorchak, JD, Goldman, MS and Walsh, TR (2000) The usefulness of electrical stimulation for assessing pedicle screw placements. J. Spinal Disord., 13: 283–289.
The use of automated or semi-automated EMG monitoring devices is dangerous and not recommended. Electrical stimulation of nervous system tissue and recording of EMG activity should be performed by a professional neurophysiologist.