Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Celebrating 75 Years! Learn More >>

 Archived Content

The National Institute of Mental Health archives materials that are over 4 years old and no longer being updated. The content on this page is provided for historical reference purposes only and may not reflect current knowledge or information.

Sarah H. Lisanby: Transcranial Magnetic Stimulation Safety and Risk


Sarah Lisanby: Hello, I'm Dr. Sarah Lisanby. I direct the Noninvasive Neuromodulation Unit and the Division of Translation Research at the National Institute of Mental Health. Thank you for joining us today for this lecture on the risks and safety of transcranial magnetic stimulation, TMS.

I'm going to start at the end with the take-home points. First of all, the safety profile of TMS is excellent. There are common side effects, but they tend to be minor and easily managed. The serious side effects are rare, and the most serious known side effect is a seizure.

Patients should be screened for seizure risk factors, and seizure can be prevented in various ways through adequate screening, through accurate motor threshold determination, and through proper dosage selection, and finally, through staff training.

The lecture today is meant to provide training to staff so that they can apply TMS in a safe fashion. These are our learning objectives for this lecture.

We're going to cover the adverse effects of TMS, both the common ones and the rare ones. We're going to go over the screening and ways to stratify risk for patients. We're going to cover hearing protection, the diagnosis, and management of seizures and of syncope or fainting, how to recognize and address the needs of special populations that may be at increased risk, how to recognize and address disease-specific conditions and complications.

So, let's start with the adverse side effects or side effects of TMS. The most common side effects of TMS a headache, which typically occurs during or immediately after the procedure. The headache is caused because on its way to the brain, the magnetic field stimulates the scalp muscles causing them to contract, and this can cause a muscle tension-type of headache. Another common side effect is scalp pain or scalp discomfort. This happens again because of the contraction of the scalp muscles, and it occurs during the procedure itself.

TMS can also cause facial pain through stimulation of the facial nerve causing the facial muscles to contract.

It is also reported to cause dizziness or lightheadedness. TMS does also has a risk of hearing loss, which can be prevented with the proper earplug use for the patient and the operator alike, and we're going to be covering that in a separate section.

The most serious, though rare known side effect of TMS, is seizure. The seizure happens during the procedure itself triggered directly by the TMS, and we're going to cover how to diagnose and manage seizures in a separate section. What could be the complications of seizures, and why are we concerned about preventing them?

Well, when a person has a seizure, they can have vigorous movements of their limbs, which can cause bruising, broken bones from the trauma from hitting the floor, or hitting medical equipment.

They can experience aspiration, which is breathing in of the secretions in the mouth into the lungs and aspiration can cause impairment in oxygen levels, and that could cause a lack of blood flow -- of oxygen flow to the brain and could ultimately cause brain damage.

Aspiration can also cause pneumonia.

Seizures can affect the ability to drive and to have a driver's license. The laws are different across different states and localities.

The strategies to mitigate these possible complications of seizure include protecting the body from physical injury and easing the person to the floor, moving equipment away from the patients so that they don't hit their limbs against the equipment, and preventing aspiration by opening the airway. One of the techniques we use to open the airway, which is taught in basic cardiac life support, is the head tilt, chin-lift maneuver.

So, what's the likely worst-case scenario that can happen when TMS causes a seizure? Typically, in TMS induced seizures, the patient had a single isolated seizure that lasted less than a minute and had no adverse sequalae. The patients who have had TMS induced seizures have required an emergency medical evaluation to assess and render emergency care, but they've all been discharged to outpatient care. After a person has a seizure, it's not safe for them to drive themselves home because the seizure can have lasting effects on coordination and movement and so it's important for the person who's had a seizure to be driven home by a responsible adult.

If a seizure happens in the context of a research study, the patient should be removed from the study, and a Significant Adverse Event or SAE should be reported to the Investigation Review Board or IRB. If the study is an FDA regulated trial under an investigational device exemption, then a seizure -- a TMS induced-seizure or SAE needs to be reported to the study sponsor, and the sponsor reports the event to the FDA. TMS has not been reported to cause status epilepticus, which is a seizure that doesn't stop without medical intervention to date, and TMS has not been reported to cause epilepsy, which is a condition where seizures occur spontaneously to date.

So, let's go over some of the strategies to prevent seizure with TMS. Proper screening of subjects to rule out those at increased risk is a very important step in preventing seizures. Patients with seizure risk factors are generally excluded from study participation. In terms of clinical use of TMS, it's a medical judgment about the risk/benefit ratio and the likelihood of the benefits outweighing the risks for each individual patient.

A second step in preventing seizure with TMS is adjusting the dosage to the individual motor threshold, which is the intensity required to elicit a twitch in a hand muscle and that's covered in a separate lecture, to stay within the safety guidelines for TMS dosage, and we're going to be reviewing those later in this talk. This motor threshold determination to set dosage for each individual patient is critical to preventing seizures because seizure risk with TMS is dose-dependent. Giving the correct dosage for each subject is the best way to prevent seizures. A third step in preventing seizure with TMS is to follow and document the timeout procedure, which is the universal procedure where before doing an intervention, you verify that you have the right patient, you're giving them the right dosage, to the right location on the head.

Now let's talk some about the responsibilities of those who are prescribing and operating TMS with respect to safety. So, the first step is prescribing the TMS as a medical intervention. TMS is prescribed by a licensed physician or other professionals with prescriptive authority, such as a nurse practitioner or NP with advanced training in TMS and a written collaborative agreement with a physician experienced in TMS. The prescriber takes responsibility for the medical appropriateness of prescribing TMS for that patient or individual and therefore has, prior to the prescription, performed an assessment of the medical risks and benefits of TMS for that individual patient and that evaluation should be documented in the medical chart. The second step is the operator, the individual who delivers the TMS to the patient.

The device operator can be a licensed physician or a clinically trained and privileged individual operating under the direction of a physician.

This set of video lectures is meant to provide basic didactic training that is important for both prescribers and operators of TMS devices. The responsibilities of the TMS operators to ensure the safe delivery of TMS are summarized here. First of all, the TMS operator would perform a timeout procedure verifying that they have the right patient, the right procedure, and the right side and the right dosage for that individual. Secondly, the operator ensures proper ear plugins insertion in the patients. So, the patient must wear earplugs at all times during TMS, and the operators should wear earplugs as well. The operators should provide continuous visual monitoring so that they ensure that the patient has not lost consciousness and is not experiencing any abnormal movements. It means that the operator needs to be watching the subject, watching that the coil is remaining in the right place on the scalp, and watching the arms and legs to be sure that there is no abnormal involuntary movement, which might be an indicator of seizure. All operators should receive in-service training on seizure detection and emergency protocols. This video lecture is meant to provide the didactic aspect of that training to support on-site, in-person training on device operation and safety procedures.

All TMS operators should receive training as a first responder. Generally, that's provided as part of the American Heart Association's Basic Life Support training or BLS. Operators should know how to immediately assess -- access clinical coverage and notify the clinician in charge and the clinical care team of any adverse events or significant adverse events that occurred during the procedure. And the operators should know the location of emergency equipment such as a crash cart, oxygen, or suction if this is being performed in a hospital-based setting. If the TMS is being performed in a community-based setting, the operators should know how to activate the emergency medical response system.

Now I'm going to review the frequency of adverse effects and significant adverse events in TMS seen in the Pivotal trials that led to the FDA approval of TMS. This is an overview of the safety of TMS in the industry-sponsored Pivotal trial that led to the FDA clearance of TMS. In that Pivotal trial, which had approximately 300 patients, there were no seizures, deaths, or suicides. There were minimal systemic side effects. There were no adverse effects on cognition. Most of the common adverse events were headache and scalp discomfort, which occurred during the active treatment itself, and less than 5 percent of patients discontinued TMS due to adverse effects. So, in summary, tolerability was extremely high.

After the industry-sponsored trial, the NIMH sponsored a randomized controlled trial, which was independent and had a comparable study design. In the NIMH sponsored trial, there were no seizures, deaths, or suicides, and the systemic side effects were minimal. They were no adverse effects on cognition. Most of the common adverse events were headache and scalp discomfort, and 5.5 percent of the patients receiving active TMS discontinued due to adverse effects, four of those were headaches, and one of those was syncope or fainting.

This table shows the actual frequency -- the percentage of patients who experienced these common adverse events during acute treatment in the industry-sponsored Pivotal trial. You can see that application site pain and application site discomfort were the most common side effects reported in 35, I'm sorry, 36 percent of patients had experienced application site pain, and approximately 11 percent of patients experienced application site discomfort in the active arm.

You can see that muscle twitching also occurred in 21 percent of patients receiving active TMS, other side effects experienced were eye pain, toothache, this can be from the facial nerve, stimulation facial pain, muscle twitching, and pain on the skin.

The occurrence of these common side effects can change over time. Specifically, headache and scalp pain tends to diminish with repeated sessions. On the left part of this slide, you see the time course of the incidence of headache in the randomized controlled trial, and you see that after the first week there was a significant number of reports of headaches, but then the frequency of headache reporting went down with successive days and weeks of treatment. So, the side effect of headaches tends to acclimatize over time. This was also seen with scalp pain, which was on the right part of the slide.

In the green line, you can see that active TMS had a significant increase in scalp and application site pain during the first week, but then that green line goes down, and the frequency is diminished over time.

This table shows a summary of the overall device-related adverse events during the industry-sponsored trial, and you're looking at those that occurred with active TMS. The first column shows the overall incidents, and the far-right column shows the incidence of the side effects that were judged to be device-related, and importantly, overall, the frequency was quite low. You do see a category here towards the bottom called psychiatric disorders, situations such as insomnia, anxiety, changes in libido, and depressive symptoms, and a significant portion of patients did experience insomnia, reported in 35 percent and anxiety in 14 percent; however, only 1 percent of those were judged to be devised related. Sometimes these conditions are part in parcel of the psychiatric condition that is being treated by the TMS, and reports of these side effects can be a consequence of the person's depression, not yet responding to the treatment.

These are the common side effects that we saw in the NIMH sponsored trial. The first column is the frequency occurring with active TMS, and the far right column is the frequency with Sham TMS. You can see that the number one most frequent side effect was headache in 32 percent of people receiving active and 23 percent receiving Sham. The second most common side effect was discomfort at the site of stimulation in 18 percent with active and 10 percent with Sham and then insomnia, close to 8 percent with active and 10 percent with Sham. I will point out that there were no statistically significant differences in the frequency of these common side effects between active in Sham.


A meta-analysis was recently reported of the side effects experienced in TMS randomized sham-controlled trials.

Zis and colleagues pulled data from 93 randomized controlled trials that had active in sham arms, and 29 percent of active and 14 percent of sham patients experienced at least one adverse event. The odds of experiencing an adverse event were 2.6 times higher in active TMS versus Sham. The most common adverse event was headache with 20 percent of people in active and 10 percent in Sham. And so, this meta-analysis is very consistent with the large Pivotal trials that I've just reviewed for you. The second most common adverse event was dizziness occurring in 3 percent with active and 2 percent in Sham. Seizure was a very rare side effect, and in this meta-analysis, seizure was reported in 0.1 percent of active study arms and 0.2 percent in Sham. So, it's an exceedingly rare event.

These are some steps that you can take to manage these commonly occurring adverse events. First of all, in the informed consent process, it's important to set expectations so the patient understands what the common adverse events are and so that they understand what to expect, what TMS will feel like, what it will look like, what it will sound like. It's helpful to reassure patients that the adverse events gradually reduce over time in their severity. If the subject is experiencing headache or scalp pain, you can offer over the counter analgesics prior to treatment in order to prevent the pain from occurring.

You can also perform minor coil positioning adjustments so that you can slightly rotate the coil so that the direction of the induced electric field might be directed away from the facial nerve or the scalp muscles that are causing the pain. In the first few sessions, sometimes it's helpful if the person is experiencing scalp pain to reduce the intensity of the TMS and then gradually increase it back up to the intended treatment intensity as they become acclimatized to the sensation of TMS on their scalp. In some cases, you can consider using topical lidocaine, a topical anesthetic cream at the site, in order to reduce the sensation.

Now let's move to the serious adverse events. I've just reviewed for you the common adverse events, which are relatively minor and mild.

Now we're going to focus on the serious adverse events, which are more rare. They come into different categories starting with the psychiatric events such as worsening of depression or worsening or new emergence of suicidality.

There are also potential serious adverse events having to do with central nervous system function, such as cognitive changes or tissue damage, which is not likely in a normal brain, but the metal in the head, as we'll review later, could result in additional risk of the TMS heating up tissue and possibly damaged tissue, auditory side effects, such as hearing loss or tinnitus, which is ringing in the ears and seizures. We're going to review each of these.

Let's look at the frequency that these SAEs were reported in the industry-sponsored randomized control trial.

Here I'm showing you the serious adverse events, on the middle column, are the sham arm and on the far-right column are the active. You can see that overall the numbers are quite low for all of these adverse events.

And on the -- in fact, the psychiatric adverse events, which are the bottom rows, were more common in those receiving sham TMS compared to active. This is thought to be due to the fact that they were not improving in terms of their depression.

The active condition, the most serious, or let's say the most frequently observed serious adverse events in the active condition was scalp burn, and that occurred in 1.2 percent of patients. This was because of a particular addition to the coil, which was an electric field shield, and that was used as part of the blinding strategy for the study. That electric field shield malfunctioned and did cause a scalp burn. TMS, when given with a standard TMS coil, without that particular shield, has not been reported to cause scalp burns.

Here we're looking at the frequency of the emergence of suicidal ideation in the industry-sponsored trial.

The gray line is the sham group, and the blue line is the active TMS group. On the Y-axis, you're looking at a shift in the score indicating the percentage of patients who experienced a change in their report of suicidal ideation.

So, you can see that there was an increase in the percentage of patients reporting suicidal ideation over time and that occurred mainly in those receiving sham TMS, not the active arm.

In these trials, we studied the cognitive safety of TMS using sensitive neuropsychological tasks that assess different aspects of cognitive function. Here you're looking at the results of four of those tests that look at short term recall, the delayed recall, the mini-mental status exam, and the auto graphical memory inventory for autograph -- autobiographical retrograde memory. The blue columns are the active TMS, and the gray columns are the Sham TMS, and this shows you that there was no change in any of these measures and no difference between active or sham. This gives us confidence about the neurocognitive safety of this particular depression treatment protocol.

So, in summary, there were no significant neurocognitive adverse effects seen in the depression trials.

However, it should be noted that TMS can interfere with cognitive functions on a short-term basis. This is called the virtual lesion effect. When you stimulate a part of the brain that is responsible for performing a cognitive task, you can disrupt the person's ability to perform that task while you're stimulating, and indeed this virtual lesion function has been useful in studying brain-behavior relationships.

However, lasting adverse cognitive effects have not been demonstrated. The long-term effects on cognition on the developing brain and by that, I mean on children and adolescents, is an area that requires further study.

Let's talk about the evidence on cellular alterations. As a consequence of TMS, TMS induces an electrical field in the brain. However, that peak induced current density is too low to cause damage to tissue or to cause appreciable heating of brain tissue.

There has been no report and no evidence of cellular damage or mutagenesis, which is the induction of cancer in the brain by TMS. The safety TMS in the presence of implanted metal hardware or devices is less well studied.

There is a theoretical risk that TMS could cause heating in metal that's in the head, and therefore that is considered a contraindication.

Now let's talk about the evidence on how frequently seizures have occurred with TMS. A recently reported survey of TMS labs conducted by Lerner, Wasserman, and Tamir surveyed TMS labs and clinics for their experience with seizures between the years of 2012 and 2016. The respondents reported over 30,000 TMS sessions and over those -- over 30,000 TMS sessions, there were 24 seizures, which indicates a frequency of 0.08 per 1000 sessions of TMS. Four of the seizures were in patients who were treated within the safety guidelines and in subjects who did not have known seizure risk factors. This gives a risk of seizure of less than 0.2 per 1000 sessions when patients without seizure risk factors are being treated within the safety guidelines. This rate is extremely low, but it also points out that the risk of seizure is not zero, and therefore it's important to take seizure risk precautions in every patient and also to inform patients that there is a risk of seizure from TMS.

Seizures were more common in patients who were at risk because of brain lesions or epilepsy or where the dose exceeded the safety guidelines. Seizures were more likely in the first few sessions; however, it is also possible for seizures to occur later in a course, particularly if there are clinical changes or changes in medications. I've reviewed for you common and the rare, the mild, and the serious adverse events of TMS.

In this section of the lecture, we're going to go over screening and risk stratification. Let's talk about the pre-TMS workup. Medical evaluation should include a medical history and physical exam. If you are performing TMS in a research setting, a medical history and exam might be part of your screening procedure if you're performing TMS in another setting such as a clinical setting, documenting that the person has had a recent medical evaluation and physical exam by their primary caregiver may be judged adequate by the physician making this evaluation. In the pre-TMS workup, it's important to screen for the contraindications to TMS, which may include metal in the head or epilepsy unless TMS is being used to study epilepsy in a research setting. To evaluate the medication list for seizure lowering agents so the clinician doing the pre-TMS workup should look at the current medications, think about whether they would influence the seizure threshold up or down and consider which if any of those medications should be tapered off or reduced in dose prior to starting TMS.

The workup should include a screening for substance use disorders and also for illicit substance use at all.

Because even without meeting criteria for a substance use disorder, the binge use of alcohol or the acute withdrawal from intermittent use of an illicit substance could affect seizure risk.

If there are concerns about the possibility of illicit drug use in the subject, you may consider doing a urine or plasma drug screen. Evaluating whether the person is pregnant or could be pregnant is an important part of the workup, and you may consider providing a pregnancy test where pregnancy is a possibility and also discussed with the patient procedures to prevent pregnancy after starting the TMS, such as the use of an effective means of medical birth control or abstinence.

No particular labs or brain imaging scans are required for the pre-TMS workup unless it's indicated by the history or special medical conditions or unless it's indicated by the research protocol.

Informed consent should be part of the screening process and include a thorough discussion of what to expect in terms of the physical sensations, the auditory sensations of TMS, the common side effects, the risks of potentially serious adverse events, including seizure.

How do we screen for seizure risk factors? Well, first of all, take a history of whether the person has ever had a seizure and whether anyone in their family has had a seizure, including febrile seizures in childhood. It's important to review whether the person's taking any medications that lower seizure threshold, most psychotropic medications, and psychostimulants do lower seizure threshold, withdrawal from medications that res seizure threshold, like coming off of an anticonvulsant can also put the person at increased risk of seizure and therefore if they've recently been tapered off of an anticonvulsant, it's important to wait several weeks for their threshold to stabilize before starting TMS if clinically appropriate.

Substance use in withdrawal is an important aspect of the screening procedure risk factors. Most substances of abuse can themselves induce a seizure either during intoxication and overdose or during the withdrawal period.

Certain medical conditions can increase the risk of seizure, such as electrolyte imbalance or a glucose imbalance, either low or high. Therefore, if the subject has diabetes, it's important to be sure that their glucose is stabilized and that they are being well medically monitored during the treatment course.

You should screen the subjects for known structural brain lesions, including those resulting from head trauma, such as traumatic brain injury, as well as those resulting from a stroke, a tumor, or multiple sclerosis. If the person is in a research study and they've had a brain imaging scanning done as part of the research protocol, it's important to look at the results of that scan before performing TMS to see if the scan revealed any previously undiagnosed structural brain lesion.

A useful tool for screening procedure risk factors is the TMS adult safety screen or the TASS. The reference for the TASS is shown here, [inaudible], and here are the questions. You can have a printed form that you give the subjects, or you can perform this screen orally.

You ask the patient, have you ever had an adverse reaction to TMS before? Have you ever had a seizure? Have you ever had an EEG or an electroencephalogram? Have you ever had a stroke? Have you ever had a head injury, including from surgery? Do you have any metal in your head outside of the mouth, such as shrapnel, surgical clips, or fragments from welding or metalwork?

This includes metal in the eye as well as the head. Do you have any implanted devices such as cardiac pacemakers, medication pumps, or intracardiac lines? Do you suffer from frequent or severe headaches?

The TASS is continued on this slide.

Have you ever had other brain-related -- another brain-related condition? Have you ever had any illness that caused brain injury? Are you taking any medications? If you are a woman of childbearing age or you're sexually active?

And if so, are you not using a reliable method of birth control? Does anyone in your family have epilepsy or seizures? Do you need further explanation of TMS and its associated risks?

The purpose of the TASS is to alert investigators and clinicians to factors that may affect the risk of TMS.

The decision of whether to proceed with TMS in the clinical or research setting depends on the clinical judgment of the risks and benefits. So the TASS is meant to elicit the information that then the clinician uses to assess the safety of TMS in that condition.

There are some contraindications to TMS, such as non-removable metal objects in or near the head. This excludes dental hardware. It means if the person has things implanted in their mouth, on their teeth, that is not a contraindication. But what is a contraindication with the metal in the skull or in the head, anything that's conductive could have electricity induced in it when it's underneath the TMS coil, anything that's ferromagnetic could not only heat up but also be moved as a consequence of the TMS or any other magnetic sensitive materials that are implanted and are non-removable within 30 centimeters of the treatment coil.

Another contraindication is an implanted device that is activated or controlled by physiological signals even if the device is located outside of the 30-centimeter distance. This includes a deep brain stimulator where the stimulating pack is implanted in the chest, but the electrodes are directly in the brain or a cochlear implant. Another contraindication is patients wearing a wearable cardioverter-defibrillator.

Implanted device -- certain implanting devices may be allowed if they are located in areas that are greater than 30 centimeters away from the TMS coil and if they're not controlled by physiological signals. Examples of this could include staples in an area of the brain -- in an area of the body outside of the head or an implanted insulin pump.

There are some relative contraindications to TMS. Pregnancy.

Although there are some reports of safety and efficacy, this is an area that requires further study, and I will be discussing the risks and benefits in the context of pregnancy in a later section.

Childhood. Although there are some reports of safety and efficacy in children and adolescents, this remains currently off label at the time of this filming. Heart disease, especially if it is unstable and a cardiac pacemaker. Although the pacemaker is far enough away from the TMS coil to be -- not likely to be affected this is currently considered a relative contraindication. And medication pumps and the use of medications that are known to increase seizure risks such as tricyclic antidepressants and antipsychotics.

Although these have been given safely with TMS in clinical settings and in certain research studies.

There are some special considerations such as vagus nerve stimulation in the case of VNS, the devices implanted in the chest, and the electrodes are implanted on the vagus nerve in the neck.

That device is far enough away to not be expected to be directly affected by the TMS, but caution and consideration of alternative treatments are indicated. Of note, ECT or electric convulsive therapy has been shown to be safe in the context of VNS, but the VNS should be turned off during that procedure.

there are also special considerations about tattoos, especially those that might contain a metallic material if they're on the scalp. There are some warnings that are part of the TMS manual, which is the FDA label for TMS. One of the warnings is that the TMS may not be effective, meaning they condition requiring the use of TMS might worsen during the treatment.

Let's talk about how we stratify risk.

How do we sort through the relative risks across different patients? Well, when new stratify risk, we think about patient factors and device factors. Let's start with patient factors. Patient factors include the patient's medical conditions, their psychiatric conditions, their age, what medications they might be on, and whether there is a risk of substance use.

This is a list of some of the medications that patients may be on when they're referred for TMS., and it's important to consider the class of antidepressant medication, whether it's a class that is associated with increased risk of seizure, benzodiazepines, generally speaking, would lower seizure risk, however, if the person is being tapered off of a benzodiazepine that's a situation that might increase risk. Anticonvulsants generally would lower the seizure risk except for in the case of being tapered down or off of the anticonvulsant. Bupropion, Wellbutrin is a medication that itself has been reported to cause seizures, especially in those patients with a history of bulimia. So it's important to be sure to take the history and consider the risks and benefits of remaining on that particular medication. Caffeine has been known to prolong seizures in the context of ECT.

Generally speaking, moderate use of caffeine is not considered to appreciably affect seizure risk, but heavy users who might be acutely withdrawing is something to consider.

Let's move now to device factors and the dosage and how the dosage affects risk. This slide graphically illustrates the parameters of TMS that go into the dosage. Let's start with intensity. So the little rectangle there is meant to illustrate a cartoon of a TMS pulse, and the pulse can be strong, or it can be weak. That's the intensity of the stimulation. You can set the intensity on the device in terms of the percent of maximal stimulator output from zero to 100. The frequency is how fast the pulses are given. The number of pulses per second is called the frequency in Hertz, so the frequency can be either low or high. High frequency, the pulses are spaced close together, low frequency, the pulses are spaced further apart in time. The next parameter is train duration, which is how many -- how long the train goes on. So, the train could be one second, it could be two seconds, it could be eight seconds.

So, the train duration in seconds is reflecting how long the train lasts. And then there's the number of seconds between trains. That's called the intertrain interval or the ITI. The intertrain interval can be long, or it can be short.

These different parameters affect risk in different ways. Let's start with the intensity. Generally speaking, seizure risk goes up as the stimulation intensity goes up.

Generally speaking, seizure risk goes up as the frequency goes up. So, when we go from 1 Hertz to 10 Hertz to 20 Hertz, seizure risk increases. Generally speaking, seizure risk increases with the longer train durations from short trains to longer trains, and seizure risk goes up when the intertrain interval goes down when the trains get closer together.

Studies suggest that single-pulse TMS is thought to be safer than repeated trains of TMS pulses. However, single-pulse TMS still does have risks.

For example, there have been very rare reports of seizure with single-pulse TMS, and this is one of those reports.

Now, single-pulse TMS has also been reported to cause syncope, which is fainting, and there's a type of syncope called convulsive syncope where the person looks like they're having a seizure because of motor movements. And in many of these cases, it's been hard to determine whether it was an actual seizure or convulsive syncope. But in any event, even when you're giving single-pulse TMS, it's important to know that seizure risk is a possibility and to be prepared in that eventuality.

Studies suggest that higher frequencies have a greater risk of seizure than lower frequencies and that shorter intertrain intervals have a greater risk than longer intertrain intervals.

It's important to follow the safety guidelines, which I'm going to show on the next slide for dosage or if you're deviating from those safety guidelines which tell you which combination of frequency intensity and train duration is considered safe, it's important to have a medical rationale to deviate from those safety guidelines. In some cases, if you're going above the safety guidelines, approval of an Investigational Review Board, and in some cases, approval of the FDA may be needed. These are the safety guidelines that show you for each combination of frequency and intensity, how long the train duration should be to prevent seizure risk.

So, the frequencies on this table go from one to 25 Hertz, and those are each of the rows. The intensities go from one hundred to 140 percent above the motor threshold. Those are the columns. The numbers within the column reflect the number of seconds for the train duration.

It means, for example, if you're giving a 10 Hertz train, which is the third row here at 100 percent of motor threshold, the maximum safe train duration for that combination is 4.2 seconds.

This table also tells you that if you're giving one Hertz at 100 percent of motor threshold, the very long trains have been given and have been found to be safe. So this conveys the concept that as you go up and intensity and as you go up in frequency, your train duration needs to go down to remain within a safe range.

There's a way to report TMS induced seizures to the FDA. I'm showing you here one of the reports that were submitted to MedWatch. And if you do have a patient that has a TMS induced seizure, either in a research study or in clinical care, it's useful to report that to the community by reporting it on MedWatch. This person, in this report, had major depression and type 2 diabetes, and TMS induced a generalized tonic-clonic seizure, which means the seizure generalized to the entire body and caused rhythmic movements of the limbs.

The seizure started about eight minutes into their 10th TMS session. The seizure started in the right arm, and it spread proximally towards the shoulder and then ultimately to all four extremities. The onset of the seizure is important. The patient was receiving TMS to the left frontal cortex, which is the common depression treatment target.

The left frontal cortex is near the left motor strip, the area of the brain that controls movement in the right side of the body, so the part of the body that is most likely to start having a seizure from TMS on the left hemisphere and the left frontal cortex is actually the right arm. And so, it's important for the TMS operator to have the right arm and the entire body in clear visual view throughout the procedure so that they can detect any early warning signs of seizure. The warning sign of a seizure, in this case, was movement in the right arm.

Some possible causes that were thought to have resulted in this team as induced seizures were operator error.

The coil was left on the left motor cortex instead of being moved to the left frontal cortex. So, they were stimulating the wrong site. The mitigation for this is operator training and the timeout procedure to be sure that you're stimulating the right site for that patient.

It was also considered whether the motor threshold was not accurately determined. If you determine a motor threshold when your coil is not really on the optimal site for activating the muscle in the hand, you might get a falsely high motor threshold. And then when you set the dose for that patient, you think you're stimulating within the safety guidelines, but in fact, you're not. The mitigating procedure for this is operator training in accurate motor threshold determination.

It was also thought, in this case, that the patient was taking concomitant medications that might've affected their seizure risk. The patient was taking a tricyclic antidepressant, was taking serotonin reuptake inhibitors, the patient was taking atypical antipsychotics and a benzodiazepine. This is a mixture of medications that both can -- many of them can increase, and some can reduce seizure threshold, and the net effect of this combination of medications on seizure threshold is difficult to predict. This case highlights the need for training and quality control and the use of timeout procedures. This set of video lectures is meant to provide the didactic aspect of that training.

So we're still talking about risk stratification and the dosage factors. We've gone over the stimulation parameters and the proper use of the safety guidelines. Now I'm going to talk about on labeled dosage and how it's different for each device.

On label means this is the dosage recommended by the FDA when you're operating this particular device and seizure risk when you're stimulating with an off-label dosage, meaning it's outside of what the FDA is recommending, seizure risk, in that case, might be higher.

The on-label dosage differs by the different manufacturer's devices. Let's start with the figure-eight coil, which is the typical coil used by the Neuronetics device, the Magstim device, and MagVenture device.

The recommended dosage that's on label is 120 percent of motor threshold, four-second duration trains with an intertrain interval of 26 seconds, and you're giving 75 trains in each session and 3000 pulses per session.

The recommended dose for the Mag Vita or MagVenture device using Theta Burst Stimulation or TBS with a figure-eight-coil is 120 percent of the motor threshold giving bursts of three pulses at 50 Hertz, which is, and you give those bursts at five bursts per second, which is the theta part. You give two-second trains with an intertrain interval of eight seconds, and the total number of bursts is 20. You give 20 trains for a total stimulation duration time of three minutes and nine seconds.

The on-label dosage for the brain sway deep TMS coil, also called the H-coil is 100 percent of motor threshold, 18 Hertz, two-second trains, 20-second intertrain interval, 55 trains per session, and a total of 1,980 pulses per session. It's important to be sure you're using the right dose for the right manufacturer. The on-label dose will be found in the operator manual.

Now that we've talked about dosage let's move on to coil and how the different TMS coils affect seizure risk.

In general, larger coils have a greater risk than smaller coils, larger coils induce stronger electrical fields in the brain, and larger coils stimulate larger regions of the brain. Here are the two standard coil types that are FDA cleared, the figure-of-eight coil, and the H-coil. Each of these coils has a different distribution of the electric field induced in the brain. Here I'm showing you a model using a concentric spherical model of the head, and you can see that the figure-of-eight coil is stimulating a very focal area, and it does not penetrate very deeply. Whereas the H-coil stimulates a broader ring of tissue across broad areas of the head, and it penetrates somewhat more deeply.

There was a seizure reported with the H-coil in the Pivotal trial that led to the FDA clearance of that device.

The overall safety of the H-coil and in the Pivotal trial was excellent. The most common side effect was a headache, just like with the figure-of-eight coil. There were no suicide attempts in that Pivotal trial. There was one case of a seizure, and the overall seizure rate in that trial was 0.43 percent. According to the FDA approved clinical trial protocol for this study, seizure rates up to 5 percent in patients and suicide attempts up to 3 percent were considered acceptable. Therefore, the safety profile of the H-coil was considered acceptable by the FDA.

There are not just two coil types.

There's actually a, a wide variety of coils available and I'm showing you, models of a large number of them, here, and each of them have different electric field distributions. Therefore, if you're using a coil that's not one of the FDA approved coils, it's important to understand that the safety of that particular coil may not be known, and it might be higher than the FDA approved coils.

Where you put the coil on the head influences safety. Generally speaking, the motor cortex is the area that has the highest seizure risk, and that's thought to have a greater risk than in the frontal cortex. In the case that I showed you of a TMS induced seizure, the coil had been accidentally left over the motor cortex that might've contributed to the risk of seizure for that patient.

Now that we've reviewed the adverse effects and how to screen and risk stratify for the safety of TMS, we're going to talk now about hearing protection. TMS is loud. How loud is it? Well, here I'm showing you in decibels how loud TMS, the red arrow shows you that TMS is louder than a pneumatic drill and not quite as loud as a jet plane. So, it is very loud. In fact, it is loud enough to damage hearing. TMS can damage hearing without hearing protection. TMS has damaged hearing in animal studies, and TMS has been reported to affect hearing in patients who were not wearing hearing protection.

TMS can induce tinnitus, which is ringing in the ears without hearing protection, and tinnitus can be very severe and distressing and is very difficult to treat. TMS can exacerbate preexisting hearing loss without hearing protection. So, if the patient has hearing loss, either age-related decline or hearing loss as a result of industrial sound exposure or blast exposure, they may be more at risk of hearing loss from TMS. So, it's doubly important for those patients to be sure you're using adequate hearing protection.

Properly inserted earplugs must be used at all times by subjects and by the operator to protect hearing.

Earplugs do effectively protect hearing during TMS. We performed auditory testing or audiometry before and after TMS. You're seeing here the blue lines are active TMS, and the gray lines are Sham TMS, and what we're showing you here are hearing levels from four weeks to six weeks in the right and left ear. There were no changes as a consequence of TMS in either group.

Now let's move on to how to diagnose and manage seizures and syncope. These are the signs of seizure. A seizure may include unresponsiveness. So, the person does not respond to you when you talk to them or tap their shoulder. It can cause loss of consciousness where they close their eyes, they slump forward, they lose their muscle tone, they appear to be asleep. A seizure can cause tonus, which is stiffness in the muscles in the arms and legs. It can cause clonus, which is repeated, involuntary rhythmic movements of the arms and legs.

A seizure can cause loss of bowel and bladder control, and it can cause disorientation after the person wakes up from the seizure. Not every seizure has all of these signs. It's important to know that some seizures may only cause unresponsiveness or loss of consciousness. Therefore, if those occur, it's important to see if you can rouse the patient, initiate your basic cardiac life support scan to assess breathing and circulation, and activate the emergency medical system.


A generalized tonic-clonic seizure is a seizure that spreads to all parts of the body, and that type of seizure has most of these signs. It starts with unresponsiveness, loss of consciousness. It has muscle stiffness that evolves into convulsive movements of the arms and legs. It may or may not cause loss of bladder control, and it often causes postictal disorientation.

Well, there are certain signs that might look like a seizure but might be caused by something else. What's the differential diagnosis for seizure. A person might have fainted, that's called syncope where they lose consciousness but are not having a seizure. There's a form of syncope called convulsive syncope where after they lose consciousness, they have a brief set of jerks of arms or legs that does not evolve into a tonic-clonic seizure. They may be having a pseudoseizure, which is something that looks like a seizure but does not actually have the electrographic signs of a seizure. This is also called non-epileptic seizure or non-electrographic seizure.

It's important to note that many patients who have pseudoseizures also have true epilepsy. Therefore, if the person has something that you think might be a pseudoseizure, it's important to treat it as if it could be a real seizure.

Something that could look like a seizure could be hypoglycemia, and this highlights the importance of evaluating and detecting predisposing medical risk factors.

How do you diagnose a syncope or faint?

Fainting or syncope is a loss of consciousness without loss of pulse or breathing. So, if a person appears to have lost consciousness, is important to see, are they truly conscious? Can you rouse them? Maybe they've just fallen asleep, and if they are unconscious, you should assess their pulse and their breathing to be sure that it is not a sign of a cardiac arrest or respiratory arrest.

Typically, syncope does not cause involuntary movements except in the case of a convulsive syncope. Syncope is often found in people who have a history of fainting before often with blood drawing. So it's useful to ask the subjects if they've ever fainted in any circumstance, for example, when they were having their blood drawn.

What to do in the case of syncope.

It's important to activate the emergency medical system in the appropriate context. If you're in a hospital setting, it may be calling a code.

If you're in the community, it may be calling 911. It's important to perform the basic life support assessment where you assess airway breathing and circulation so that you can determine whether it is a cardiac arrest or respiratory arrest or simply a faint. It's important to perform an emergency medical assessment to determine the diagnosis, to evaluate was this a seizure versus a syncope or another cause, and then to render emergency support to address reversible causes of the syncope. For example, the person may be dehydrated, and they may need fluids and then refer them to a medical evaluation to address other potential causes.


Here are the dos and don'ts of what to do in the event of a seizure. This is the emergency management of the seizure. Number one, stop TMS. If the TMS is triggering a seizure, you need to stop the TMS and activate the emergency medical system. Commence your basic life support assessment for airway breathing and circulation.

Record and document the duration and manifestations of the seizure. Exactly how long did it last? What did you see in the body? Was the person conscious or unconscious? Which part of the body was moving, and for how long and how did the seizure progress?

This information will be vital when the physician arrives so that they can evaluate the diagnosis.

It's important to protect the airway.

You learn in basic life support how to tilt the head and lift the chin to keep the airway open so the person can continue breathing on their own. You ease them to the floor and lie them on their side so that their mouth secretions will fall out onto the floor instead of being aspirated into their lungs. If you're trained to do so, you and if you have the proper equipment, you can administer oxygen.

It's important to protect the body from bruising or breaking bones from the motor convulsion by easing them to the floor and moving away heavy medical equipment. It's important to perform a medical evaluation to determine the causes and to commence a care plan. If this is in the context of a research setting, you should remove the patient from the research protocol unless there's a stipulation about what to do in the event of a seizure and report this as a significant adverse event to the Institutional Review Board. If this is an FDA regulated trial, you should report this as a significant adverse event to the sponsor who reports it to the FDA.

Here are the don'ts, what not to do in the event of a seizure. Do not put anything in the patient's mouth other than a suction device and only if you're properly trained to do so. Suction is used to remove airway secretions to prevent aspiration. Do not give any medications other than oxygen unless you are part of the code team, and unless it's in the event of a status epilepticus, which is a seizure lasting more than three minutes, and unless you are prepared to intubate.

Now, the reason why not to give medications to stop the seizure is that anticonvulsant medications can slow down spontaneous respiration and may require the person needing ventilatory support.

What to do after the seizure.

The person should have a medical and neurological evaluation to determine the causes of seizure. The medical evaluation will include a differential diagnosis to determine the etiology. The patients generally are removed from the research protocol unless and until medically cleared. A patient immediately after a seizure should be on fall precaution because they may be disoriented and at increased risk of falls. In an inpatient setting, you can order frequent neuro checks so that the nursing staff goes in and assesses the neurological function of the person.

After a seizure, the person should not drive themselves home on that day until they're medically cleared to return to driving. And you should document the occurrence of this seizure in the medical record and report it to the IRB as a significant adverse event in the context of research and to the FDA sponsor in the context of an investigational device exemption.

Now let's move on to recognizing and addressing the needs of special populations. One special population of particular interest is pregnancy. We know that pregnancy is a period that women can be at increased risk of depression, and sometimes this depression can be quite severe. Left untreated, depression during pregnancy can interfere with the mother's ability to receive prenatal care, and it can affect outcomes both for the baby and the mother. And so, there is a desire for an effective antidepressant treatment that does not cross the placenta as antidepressant medications can. TMS does not cross the placenta. Therefore there's great interest in studying the safety and efficacy of TMS in pregnancy.

But there are cautions to consider in the event of pregnancy. Reproductive hormones, which change dramatically during pregnancy, can affect seizure risk and can affect the chance that TMS could induce a seizure. If a mother has a seizure while pregnant, this could place both her and the fetus at risk of complications. Although electroconvulsive therapy or ECT can be given safely during pregnancy, and ECT induces seizures. ECT induces seizures when the woman is under general anesthesia and being closely medically monitored. TMS, the person receiving TMS is not under anesthesia, they're not receiving the level of medical monitoring that is performed during ECT, and therefore having a TMS induced seizure puts the woman and the fetus at more risk than what is expected in the case of ECT. An unmodified seizure, meaning a seizure-induced without anesthesia to prevent the motor contractions presents a risk to both the mother and the fetus.

Finite element modeling has been conducted in this study to estimate the electric field induced in the uterus by TMS given to the mother's scalp or brain. And this modeling has shown that when you model the intensity of TMS given to the head, the fetus has negligible electric field exposure. The model electric field exposure in the uterus -- in the fetus is significantly lower than the recommended safety exposure. And so theoretically, we do not expect any effects directly from the induced electric field. The risks of TMS during pregnancy are mainly related to the hormonal changes and the risk of inducing a seizure and the potential complications of seizure.

That being said, there have been a number of studies evaluating the safety and efficacy of TMS in pregnancy.

This recently reported, randomized controlled trial by Kim and colleagues that came out in 2019 reported on 22 women who received active or sham TMS during the second or third trimester of pregnancy. They received one Hertz TMS, which is low frequency, which is thought to have a better safety profile and lower seizure risk than the higher frequencies, and this was given to the right dorsal lateral prefrontal cortex based on studies showing that one Hertz to the right DLPFC had antidepressant effects.


In this study. There were three cases of late preterm birth in women receiving active TMS. All other obstetrical and newborn outcomes were normal. This study supported the relative safety of TMS during pregnancy.

This is a small study, however, and more work is needed to identify the more rare potential side effects of TMS and look at long-term outcomes. Other special populations include comorbid psychiatric disorders such as post-traumatic stress disorder, anxiety disorders, and substance use disorders.

Studies have evaluated the safety and efficacy of TMS in PTSD and demonstrated similar safety and tolerability profile as in the context of depression. Studies have also evaluated the anxiolytic effects of TMS in panic disorder and generalized anxiety disorder, and other anxiety disorders, again, finding a comparable safety profile to those I previously reviewed with you in depression.

In the case of substance use disorders, substance use itself increases the risk of seizure. Studies have been evaluating the safety and efficacy of TMS in this context, and this remains an area of active study and promise given the significant morbidity and mortality that substance use disorders represent.

However, this remains as the time -- as of the time of this filming off-label.

There are certain conditions of which have sensory hypersensitivity, meaning the person might be more sensitive to the sound or the tactile sensations of the TMS. Therefore it's useful when you are introducing the patient to the procedure to let them see what it looks like. Let them hear what it sounds like before you actually put the coral on their head. You can start at low intensities when you put the coil on their head and gradually increase to get them acclimatized to these sensations.

There is a risk of suicidality in patients who have severe depression. Studies do not indicate the TMS increases the risk of suicide, but if you have a person who has preexisting suicidality and they're referred for TMS, it's important to consider the relative effectiveness of TMS compared to other interventions that might be more effective in someone who is acutely suicidal.

ECT or electroconvulsive therapy remains the most effective and rapidly acting treatment. Ketamine is now also FDA approved for severe depression, and these rapidly acting treatments like ketamine or ECT might be more appropriate for a patient who is acutely -- at risk of suicide given that TMS has a lower over efficacy compared to those other treatments, and given that TMS takes longer to act.

There is a risk of mania or hypomania induced by TMS. Mania and hypomania have been reported in patients receiving TMS. There have been rare reports in healthy volunteers, and there also have been reports in patients with bipolar disorder or patients with depression with a formerly not diagnosed bipolar disorder. Antidepressants are known to have a risk of mania in the case of depression and bipolar disorder, as does ECT. Therefore, it's not surprising that TMS also carries a risk of inducing mania.

Knowing that TMS can induce mania in people who are at risk, it's important while you're treating the patient to monitor them for any signs of hypomania or mania. Some of those signs may include decreased need for sleep, increase pressure of speech, euphoric mood, grandiose thinking, and so on.

Patients who have migraines or have frequent headaches or other aspects of pain, such as facial pain, may be more at risk for exacerbation of these conditions during TMS, and that should be considered during the risk-benefit ratio.

Children and adolescents are considered to be a special population because they may be more at risk for side effects of all sorts of treatments and because these treatments have been evaluated in adults and have been relatively less studied in pediatric population. Relevant for TMS, children have higher motor thresholds. This is because of the maturation of the cortical spinal track during development. It means that in children when you titrate relative to their motor threshold, their overall intensities are likely to be higher than what we see in adults.

It's also important to understand that children tend to have lower seizure thresholds. They're more prone to seizures. Epilepsy may begin in childhood and so that a combination of the higher motor thresholds and lower seizure thresholds may put children at increased theoretical risk of seizure with TMS.

Another aspect of a risk for children has to do with the anatomy of the external auditory canal or the ear canal.

The shape of the external auditory canal changes over development, and the aspects of the shape of the external auditory canal in children leads to a higher resonant frequency meaning that TMS pulses sound louder to them, and that may be increasing the risk of hearing loss in children. It makes it all the more important in studies in children and adolescents to be sure that adequate hearing protection is being used. In very young infants or children, the adult earplugs may not fit properly in the infant or child. Therefore, it's important to have hearing protection that's proper for the age of the subject that you are treating.

There have been reports such as this one shown here by Croarkin and colleagues of increased incidents of pain in children receiving TMS, and there were three case reports of this shown here and these -- notably, these occurred even with single-pulse TMS, and it has also been reported with RTMS.

The neurocognitive safety of TMS in children and adolescents has been studied to some degree. This study used neurocognitive testing and a figure-eight coil place to the left dorsolateral prefrontal cortex, the subjects received 10 Hertz in a total of 30 treatment sessions and overall from pre to post these cognitive measures did not show any impairments. In fact, some of them showed improvements which are thought to be due to an improvement of the underlying depression.

I would say, however, although the cognitive safety profile looks excellent in children and adolescents, more study should be done because there have been -- these studies are generally smaller than the evidence base we have in adults.


A systematic review of the literature on the safety TMS in children was recently reported by Allen and colleagues in 2017. They concluded that the overall risk from TMS in children and adolescents appears to be similar to that seen in adults, but the number of children exposed to TMS remains low compared to the adult population. They identify 23 RTMS studies in children with preexisting central nervous system disorders, and in those children, the adverse event rate was 3.78 percent.

They identified three studies using theta-burst stimulation with rates of adverse events of close to 10 percent in healthy volunteers and close to 10 percent in children with CNS disorders.

They found studies with three seizures reported with RTMS in patients with CNS disorders that reflects a rate of 0.14 percent per session. The authors recommended applying the adult safety guidelines to children and adolescents until sufficient data on children are available.

Another special population to consider are those with traumatic brain injury or TBI because TBI itself can cause seizures, TBI can cause epilepsy, and TBI can be associated with an increased risk of seizure. Hoy and colleagues in 2019 reported on 21 patients with depression following TBI who are randomly assigned to receive active or sham sequential bilateral RTMS over a period of four weeks. They reported that the TMS sessions in the TBI population were well tolerated, and there were no significant differences in side effect rates between active or sham.

No seizures were reported in this study.

Working memory and executive function were measured, and they were found to be significantly improved. There was no evidence of impairment in neurocognitive function on any of their measures.

Now let's move on to addressing and recognizing disease-specific conditions or potential complications of rTMS.

In depression, a disease specific complication could be an increased risk of seizure result, I'm sorry, increased risk of suicide resulting from a lack of efficacy and disease progression.

Therefore, it's important while you're giving TMS to a severely depressed population to continuously monitor for any worsening or new emergence of suicidal ideation, and if that does occur, continue to assess whether alternative treatment might be indicated.

In bipolar disorder, there's an increased risk of mania, and also patients with bipolar disorder are at an increased risk of suicidal ideation. So, patients undergoing RTMS treatment in the context of bipolar disorder should be frequently medically monitored for the emergence of either suicide risk or hypomania or mania. You can use rating scales to help monitor during the treatment course, such as the Yale Mania Rating Scale, the YMRS, Hamilton depression Rating Scale, or the Beck Depression Inventory.

Substance use disorders. If your patient has a substance use disorder, they may be at increased risk of seizure-related to changes in their substance use pattern. Therefore, it's important while you're treating patients who may have this condition to monitor and continuously assess their compliance with abstinence to assess their self-reports of substance use frequency and to monitor their substance use by using urine drug testing or serum drug testing.

Now I'm going to recap our take-home points. The safety profile of TMS is excellent. The common side effects tend to be minor and are easily managed. The serious side effects are rare and can be prevented through proper training. The most serious known side effect is seizure. Patients should be screened thoroughly for seizure risk factors. A seizure can be prevented through adequate screening through accurate motor threshold determination, through accurate dosage selection, and through appropriate staff training.

This concludes this video lecture on the risks and safety of TMS. We hope that we have enabled you to have the didactic information to support your hands-on training in the safe and effective delivery of TMS in your clinical or research setting. Again, this is Dr. Sarah Lisanby from the National Institute of Mental Health, and thank you for joining us for this lecture.