Monday, February 12, 2018

Sedating Patients For Imaging Studies

An article in this week's JAMA hit me like I was still on my old inpatient job.  It was about the issue of sedating patients for imaging studies.  Quality brain imaging - whether it is an MRI or a CT scan depends on the patient being able to lie very still.  Any movement causes artifact that can obscure critical brain areas of interest.  The reasons for the agitation vary quite a bit based on the population but the agitation is not necessarily any easier to treat.  The setting is often different.

As an example, agitated patients on our neurology service when I was an intern were typically agitated due to brain disease.  One of the first patients I saw was elderly and extremely agitated.  Delirium or psychiatric illness was suspected because of examination limitations.  When I examined the patient in the emergency department (ED) - and did the otoscopic exam - there was a large amount of pus coming out of the left ear.  Subsequent lumbar puncture showed that the diagnosis was pneumococcal meningitis.  Like all agitated neurology patients, the chief resident came by and administered intravenous fentanyl.  The junior resident and I stood by next to the CT scanner in the event the patient became excessively sedated or apneic because of the fentanyl.  Without it the CT scan would have been impossible.  She was subsequently admitted to the ICU and had a very complicated course, but eventually survived and left the hospital.

On the psychiatric side. things are a little bit different.  The indications for brain imaging are all based on psychiatric diagnoses.  The medical status of the patient may be completely unknown, based on their ability to cooperative with a review of systems and physical examination.  In most inpatient psychiatry settings these days the patient has come through the emergency department but the complete diagnostic evaluation is deferred to the inpatient side.  Assessment by the inpatient staff the next day may indicate that brain imaging is needed.  The ability to cooperate may vary from an inability to sit still to overt aggression based on the illness.  It is a common occurrence to get a request from radiology to sedate the patient before they go for a brain imaging study.

The issue from an inpatient psychiatric unit is several fold.  Many of the patients are very vigorous and have no physical illnesses.  The medications used on psychiatric units are not anesthetic agents and they do not work immediately.  Psychiatric units are rarely staffed at a level that several physicians can accompany the patient and give them an agent that would work immediately.  Even if they could - the question would be qualifications to supervise that process. At the minimum, they would need to be qualified to administer that agent and manage a cardiopulmonary arrest.  Finally, there is the hospital wide issue of how much support can psychiatry count on.  Can psychiatry for example request an anesthesiology consult for the purpose of imaging study sedation?

In the case report, an elderly man with a BMI of 39 and an history of stable coronary artery disease presented to the ED with dizziness. He ahd associated hypertension, hyperlipidemia, history of carotid endarterectomy, and obstructive sleep apnea (OSA).  He was treated symptomatically with meclizine and ondansetron but a neurology consultant recommended an MRI scan of the brain. He was not able to tolerate the close confines of the MRI scan and was given 1 mg of lorazepam for anxiety reduction.  Over a period of about 15 minutes in the MRI scanner he became incoherent and eventually unresponsive and a cardiac arrest code was called.

The article reviews the errors made in this case beginning with the administration of lorazepam.  Lorazepam is commonly used on inpatient psychiatric units for detoxification, agitation, and insomnia.  The exact dose in this case is a dose that I have administered many times to patients who were going for imaging studies and it is frequently not enough for that purpose.  In this case the patient has OSA and risk factors such as increased age that place him at higher risk for complications.  In this case the authors suggest the minimal dose and if more is needed to monitor heart rate, pulse oximetry and blood pressure in high risk patients.  I would typically do that by requesting an anesthesiology consult for the purpose of sedating the patient for an MRI scan.

Associated measures of care in  this situation include equipment availability.  They recommend the availability of a fiberoptic bronchoscope in case the patient needs immediate intubation and the intubation is difficult.  They consider it to be a priority in the case of patients who have risk factors for airway loss after sedation.   

The American Association of Anesthesiologists has designated dexmedetomidine as a sedative that does not compromise the cardiorespiratory status of patients.  It is a alpha-2 adrenoreceptor agonist.  I did a search on psychiatric applications of dexmedetomidine and the results of that search can be found here.  The package insert discusses the limited applications of ICU intubation and sedation of non-intubated patients for procedures.   

Communicating the patient's OSA status was also viewed as a key error correction process.  OSA is a highly prevalent condition making it more likely that patients with this condition will be sedated for MRI scans.  The suggest including an OSA section in the MRI checklist.

When I think about how this procedure has been done over the course of my career - it was hardly standardized and apart from my neurology team monitoring critically ill neurology patients inside a CT scanner - little monitoring was done.  About 15 years ago that landscape started to change.  Suddenly anesthesiology consults were much easier to get and much more successful.  That was a great relief compared to a process when additional medications were being requested and nobody was there to monitor the patient.  In a few cases, I called off the scan until adequate monitoring could be established.         

The precautions noted in this case report should be studied by every psychiatrist who finds themselves ordering sedation for MRI scans or other procedures.  It is entirely possible that MRI technology may be available in some hospitals but not the appropriate monitoring staff.

In that case I would recommend forgoing the procedure if all of the recommended staff and equipment is not available. 

George Dawson, MD, DFAPA


1:  Blay E Jr, Barnard C, Bilimoria KY. Oversedation of a Patient With Obstructive Sleep Apnea Prior to Imaging. JAMA. 2018 Feb 6;319(5):495-496. doi: 10.1001/jama.2017.22004. PubMed PMID: 29411034.  

Graphics Credit:

MRI Images are from Shutterstock per their standard licensing agreement,



  1. Great post. I am currently on medicine night float for a Cardiology floor and responsible for covering approximately 35 patients nightly. Many of these patients have OSA in addition to severe CHF, CAD, Conduction Disease, and Pulmonary Parenchymal disease. I have come to the conclusion that no drug is free of risk in this medically ill population and you have to learn to weigh risks rather than find the perfect agent. More articles like this would be very helpful!

    1. Thanks for that input from the trenches.

      It was an extremely useful article. The problem is that there are no established protocols in most hospitals and wide variability in how individual physicians approach the problem. In the population that you describe, the risk/benefit of doing the imaging at all needs to be a carefully thought out decision.