ICU Patients Need Careful Monitoring in the MRI

Paul M. Kempen, MD, PhD

To the Editor

Dr. Flowerdew’s recent letter regarding physical presence during the administration of anesthesia appropriately addresses the issues involved in radiation therapy patients and also recognizes the need for physical presence to “actually administer the anesthetic.”1 The inherent danger from radiation to the provider exists (should one remain present in the room during treatment) and justifies monitoring from a safe distance. This is NOT true for MRI anesthesia, as no significant or recognized physical danger exists in the scanner room to justify monitoring from the next room. The nature of our invasive profession demands rapid evaluation and direct access to patients, ventilators and monitoring equipment, whenever possible. Clearly, when a qualified provider (i.e., CRNA) is physically present to provide continuous patient care at the bedside, an attending physician can provide direction/supervision from outside the room.

To suggest that intensive care physicians be available for electronic remote consultation can, however, detract attention from the paramount need to have an anesthesia caregiver at the bedside, to continually monitor as per ASA basic standards and guidelines. One might hope that ICU patients in Maine also have ICU nurses in close proximity to the patient, especially with their physicians at great distances. Dr. Flowerdew’s inference, however, does describe the type of contemporary care ICU patients receive in MRI scanners, when anesthesia is not administered by anesthetists: patients are alone in the scanner, on ventilators with very limited capabilities and alarms, often with sedation and vasoactive infusions, and often without end tidal CO2 or all other “routine monitors” (i.e., arterial pressure traces and volumetric pumps) they were afforded previously in the ICU and during transport. Removal of all primary providers from the scanner room further decreases safety, as I have previously reported, especially when the scanner room door is closed and opening this door requires the scan to be interrupted.2

Mark Warner, MD, may have challenged the audience of his 2005 Rovenstine Lecture to look for new ways to deliver anesthesia care as Dr. Flowerdew reported, but I do not think Dr. Warner intended safety to be actively compromised. We as a profession have been intimately involved with standards for safe sedation in all other hospital areas. I personally view the contemporary challenge to be in insuring the safety of all patients in MRI scanners, by promoting specific national standards for this unique and increasingly commonplace procedure, whenever any form of life support or sedation is employed. This requires defining the benchmark of care via our specialty; thus, raising the threshold for all specialties to follow, by “standing near by in the MRI” and with every appropriate monitor used by conventional standard or during prior care/transport.

In the same issue of this Newsletter, the new ASA and AANA guidelines requiring audible alarms for pulse oximetry and CO2 monitors have been introduced and raise a new question: Will/Do current available models reliably provide volumes adequate to overcome the MRI noise and earplugs used while in the scanner rooms?3 Possibly sound amplification, optical alarms, or other new technologies (i.e., with speakers inside protective headsets) may be needed in this special environment to insure a high level of care inside the scanner room. Similarly, if ICU (or anesthetized) patients continue to be monitored from outside the scanner room, will these monitors be required to alarm to the adjacent room, or will they be turned on, or even be utilized/connected when no one is in the scanner room to react to them? We all recognize the shortcomings of ECG monitors during MRI scanning and the inability of audible alarms to ring to the next room through the closed doors of an MRI scanner. Just what are the “national standards” for MRI monitoring, alarm settings, response times, and qualifications of the “monitoring personnel,” especially when an ICU patient on vasoactive infusions is placed on a ventilator in a MRI scanner? I think adequate “standards” remain to be adequately defined for ICU patients. Scanner technicians are not typically trained nurses or otherwise qualified to monitor intensive care patients or sedation from the scanner room. Are gravity infusions via “dial-a-flow” or micro-drip sets with intermittent NIBP measurements used when MRI compatible pumps and pressure transducers are not available? Should this be acceptable now, when modern technology is at hand? Shouldn’t nurses and respiratory technicians be physically present in the scanner to monitor their patients and equipment, just as in the ICU? The radiology physicians are furthermore not typically present or particularly familiar/engaged with life support equipment either, especially in the “off hours,” when scanners run around the clock to amortize their acquisition costs.

Paul M. Kempen, MD, PhD
Pittsburgh, PA


  1. Flowerdew RM. Radiation prevents presence in room. ASPF Newsletter 2006; 21(1):19.
  2. Kempen PM. Stand near by in the MRI. ASPF Newsletter 2005; 20(2):32 and 36.
  3. Notice: ASA and AANA modify standards to include audible alarms. ASPF Newsletter 2006;21(1):2.