Motoyama & Davis: Smith's Anesthesia for Infants and Children, 7th ed.


PART THREE – Clinical Management of Special Surgical Problems

Chapter 25 – Anesthesia and Sedation for Procedures Outside the Operating Room

Keira P. Mason,Steven E. Zgleszewski,
Robert S. Holzman



Administrative Requirements for Extramural Locations, 839



Organization and Administration, 839



Personnel, Support, and Logistics,840



Periprocedural Patient Care, 840



Standards of Practice and Quality Assurance,840



Scheduling and Preparation of Patients, 841



Selection of Agents and Techniques, 841



Postanesthesia Care, 843



Resuscitation, 844



Specific Extramural Sites, 844



Radiology, 844



Nuclear Medicine, 848



Radiation Oncology, 849



Clinic and Office Procedures,850



Safety Issues for Patients and Their Anesthesiologists, 851



Use of Contrast Media, 851



Ionizing Radiation, 852



High-Intensity Magnetic Fields, 852



Difficult Airway Management in the Radiology Suite,853



Blood Loss Management Out of the Operating Room,853



Summary, 853

Advances in diagnostic technology, increasing populations of patients surviving critical medical or surgical illness, requests from medical colleagues for support during prolonged or high-risk procedures, and concern about liability exposure have lead to an increased demand for the professional services of anesthesiologists outside the operating room ( Fig. 25-1 ). The limited space and supply of operating rooms in some institutions have further contributed to anesthetic delivery out of the operating room. Nonanesthesiologists have expressed increased interest and willingness in providing sedation and analgesia services outside of the operating room, often using anesthesia codes to bill for their services. Most anesthesiologists do not condone this practice, although no legal action to date has been taken to discourage it. Pediatric radiologists, oncologists, dentists, gastroenterologists, pulmonologists, and others provide the majority of sedation for their procedures. Monitoring techniques, available personnel, and oral intake (NPO) guidelines vary among institutions, and many sedation protocols do not conform to the sedation guidelines of the American Academy of Pediatrics (AAP) ( Committee on Drugs, 1992 ).

Providing and delivering anesthesia outside of the operating room can be challenging and hazardous. The department of anesthesia may not want to commit limited financial resources to providing state-of-the-art anesthesia equipment and monitors in an extramural location with a limited caseload. Consequently, the anesthesia machine and monitors may consist of operating room leftovers and undesirables. Support personnel working in nonsurgical areas may have little experience with anesthesia care, delivery, and needs, particularly in the case of an emergency. Acquiring insurance approval may be difficult because third party insurers do not always understand the justification for an anesthesiologist's services. Likewise, few anesthesiologists are familiar with the specific procedures, nuances, and risks associated with all interventional radiologic, gastroendoscopic, or radiation treatment procedures. Conflicts may arise among physicians, administrative personnel, and third party insurers.


FIGURE 25-1  Average monthly caseload: anesthesia support for radiology at Children's Hospital Boston (November 1999 through April 2003). The overwhelming majority of radiology studies are accomplished without anesthesia or sedation, but a significant minority (>3,000 per year) involve anesthesiologists directly or by consultation. CT, computed tomography; Nuc Med, nuclear medicine; MRI, magnetic resonance imaging.





A good relationship between the extramural department and the department of anesthesiology is critical to providing safe anesthetic coverage. Each department has its own needs, goals, and guidelines. It is ideal to designate a team of anesthesiologists committed to providing extramural anesthesia care and troubleshooting the logistical challenges of providing anesthesia in the various locations. Each member should rotate through the different extramural sites to maintain familiarity with the procedures, to foster a relationship with the physicians and ancillary personnel, and to understand the anesthesia demands unique to each site.

Few extramural locations are configured to deliver anesthetics. Ideally, anesthesiologists should be involved in the early stages of site design to ensure that minimum standards for anesthesia delivery are met and to troubleshoot engineering issues and advocate for adequate space for anesthetic induction and emergence ( Committee on Drugs, 1992 ; American Academy of Pediatrics, 1999 ). Physical plant considerations for magnetic resonance imaging (MRI) site planning have been previously described ( Koskinen, 1985 ). When anesthesia services are requested, these sites may not meet minimum standards ( House of Delegates et al., 1994) and require reengineering to include the minimum requirements stated by the American Society of Anesthesiologists (ASA). The anesthesia machine should be equipped with back-up supplies of E cylinders filled with oxygen and nitrous oxide. If pipeline oxygen is not available, oxygen should be supplied from H cylinders (6,600 L) rather than from the smaller E tanks (659 L).

Scavenging systems should be carefully evaluated in the extramural location. Unlike the operating room, passive scavenging systems may not always be possible. A safe means of active scavenging may be provided by the vacuum at the wall or wall suction canisters. A scavenging system should be dedicated solely to waste gases. Most MRI scanners do not have wall suction because MRI-compatible wall suction is not widely available. If the suction is located outside the MRI suite, a mouse-sized hole may be created in the suite's wall to allow suction tubing to be passed inside ( Koskinen, 1985 ).

Electrical circuitry and lighting in extramural locations may not be up to operating room standards; even if the outlets are grounded and up to hospital grade, there may still be plug incompatibility. Adaptors and conversion plugs should always be available. Although some extramural locations carry a minimal risk of electrical shock or electrocution, these sites do not have line-isolation monitors (LIMs) and do not warn of excess leakage of current. Supplemental lighting for record-keeping, label verification, establishment of intravenous access, and visualization of the patient is critical. Even under the best circumstances, for example, lighting is dim in the MRI scanner and monitoring by simple clinical observation can be limited. Anesthesia personnel may not always remain in the imaging suite, particularly during MRI, computed tomography (CT) scanning, and radiation therapy. Remote television monitoring or hardwiring through reinforced walls can allow remote video display of the patient and monitors within.

A storage area large enough to stock anesthesia equipment and supplies must be easily and quickly accessible. This area should be routinely checked and restocked and kept locked when anesthesia services are not required. The need for redundancy of nondisposable supplies is a matter of philosophy. Are two laryngoscopes sufficient, or should there be a third? Is one electrocardiographic monitor sufficient, or should there be a battery-operated monitor for backup and transport? Drugs should be checked per the usual operating room routine, and expired medications replaced. Gas cylinder supplies must be reliable, especially in areas without piped oxygen. A code cart should be conveniently located in an area known to all physicians and ancillary personnel. This cart should be routinely checked and restocked.

Finally, extramural anesthetizing locations are often distant from the operating room. Patients may need to remain anesthetized during transport to or from the extramural location. For these circumstances, ensured elevator access with key-controlled emergency overrides is a must. All anesthesiologists who deliver extramural services should be familiar with their surroundings. Checklists are invaluable to guarantee consistent patient care, anesthesia monitoring, equipment, documentation, and backup assistance.


Support and medical personnel in nonsurgical areas may not be familiar with the requirements of an anesthesiologist, thereby providing an important educational and training opportunity. Proper training facilitates teamwork and minimizes chaos in critical situations. A standard anesthesia cart at each anesthetizing location should be fully stocked with essential medications, necessary adjuvant equipment, spare self-inflating Ambu bag (Ambu Inc., Linthicum, MD), endotracheal tubes, laryngeal mask airways (LMAs), suction catheters, intravenous supplies, laryngoscope handles and blades, and a variety of oral and nasopharyngeal airways.

A director of anesthesia services at an extramural location can orchestrate, facilitate, and coordinate anesthesia services. This director can also serve as a consultant for the nonanesthesia medical and nursing staff. By being available to answer questions, provide on-site consults, examine patients, and provide backup support or emergency airway expertise, the anesthesiologist can also support a nurse-administered sedation program. Nurses who provide sedation under the supervision of the ordering extramural physician (gastrointestinal, radiology, dental) should be regularly certified in pediatric advanced life support (PALS) and basic life support (BLS). The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) Anesthesia and Sedation Manual sets guidelines for credentialing all personnel (physicians and nurses) who administer sedation. All children scheduled for nursing sedation should receive a prescreen telephone call from a radiology nurse the day before the scheduled scan. Often, these telephone calls are made after business hours to ensure that a parent is home. The nurse reviews the medical history, relays NPO instructions, and reminds the parents to administer the child's routine medications with a sip of clear fluid. The supervising physician must give final approval for sedation, after reviewing the child's medical history and current medical status, before ordering the medications. To minimize the chance of drug delivery error or miscalculation, it is helpful to have preprinted order sheets.

It is crucial to know who to call for assistance if a problem arises. A speed-dial system should be available from all extramural anesthetizing location sites to immediately request emergency backup.

Copyright © 2008 Elsevier Inc. All rights reserved. -

Motoyama & Davis: Smith's Anesthesia for Infants and Children, 7th ed.

Copyright © 2005 Mosby, An Imprint of Elsevier



The practice standards adopted by the ASA in 1986 for basic intraoperative monitoring apply as well to extramural locations. Practice standards and guidelines promulgated by the AAP ( Committee on Drugs, 1992 ; American Academy of Pediatrics, 1999 ) are exceeded by established practice standards in anesthesiology ( AAP, Section on Anesthesiology, 1999 ). Significant variances may exist when nonanesthesiologists provide sedation ( Keeter et al., 1990 ). Practice Standards for Non-Anesthetizing Locations were adopted by the ASA in 1994 ( House of Delegates et al., 1994) .

As recommended by the JCAHO, the department of anesthesia should oversee all sedation protocols and meet on a regular basis to review adverse events, policies, and procedures and to make recommendations for improvements. Ideally, adverse events such as failed or prolonged sedations, paradoxical reactions, hypoxia, emesis, unscheduled admission, and cardiac or respiratory events are identified and entered into a computerized database. In addition, the extramural nurse should call all patients and families within 24 hours to follow up on patient outcome and identify any delayed adverse events.


Appropriate planning for providing anesthesia begins with a familiarity with the procedure. The requesting service orders the procedure and then leaves the logistics of scheduling to the extramural service. The referring physician should not schedule the procedure because he or she may not realize the subtle logistics of the procedure or its duration. Radiologists, in particular, recognize that involvement with anesthesia lengthens their total time commitment to a patient and potentially limits the number of procedures accomplished in a day ( Winter, 1978 ; Cremin, 1990 ). A well-coordinated system to screen patients on the day of procedure is important. Experienced personnel, usually a certified nurse practitioner (CRNP), should be designated to take initial vital signs, review recent medical history, insert intravenous catheter, if necessary, and familiarize the family with the upcoming anesthesia procedure.

Screening patients for extramural procedures may be time consuming and challenging. Many children are chronically ill, nutritionally impaired, and medically complicated. These issues must be carefully addressed through attention to the patient's history, physical examination, previous medical records, outside consultations, and close communication with other medical colleagues. Several consultants may need to confer to fully understand the patient's current state of health. Not every procedure is elective. For example, urgent procedures may be required despite an upper respiratory tract infection (URI), ongoing pneumonia, deteriorating physical status, untreated gastroesophageal reflux, sepsis, or hemodynamic instability. In these situations, consultation among the anesthesiologist, the requesting physician, and the radiologist should confirm urgency. Anesthesia plans should be modified to accommodate the requirements of the procedure (breath-holding for chest CT scan) and the patient's medical condition.

It is not always possible for an anesthesiologist to provide sedation and anesthesia for all children when there is a large volume of cases. A structured nursing sedation program can provide safe and effective sedation. As recommended by the JCAHO, the department of anesthesia at each facility should work with the department of radiology and nursing to develop and oversee a sedation program. In actuality, the sedation program begins with the triaging of patients before scheduling the procedure. A radiology nurse screens the patient by reviewing the past and current medical history and gathering any relevant laboratory or clinical studies. She or he then contacts the family or referring physician for clarification if needed. After this review, the nurse, in the majority of cases, is able to make an appropriate referral for either general anesthesia or procedural sedation. Because MRI is a unique environment, it is more efficient to have the MRI nurses screen the patient before and on the day of the procedure. To ensure consistent decision making, the departments of anesthesia and radiology should develop a set of guidelines and “red flags” ( Table 25-1 ) to help in this triaging process. If there are any questions or additional medical history or studies that need clarification, the nurse and anesthesiologist confer before making the final decision regarding general anesthesia or procedural sedation.

TABLE 25-1   -- “Red flags” for sedation

Red Flag



Documented by sleep study, strong clinical history, or apnea monitor

Unstable cardiac disease

Cyanotic, depressed myocardial function, significant stenotic or regurgitation lesions

Respiratory compromise

Recent (<8 weeks) pneumonia, bronchitis, asthma, respiratory infection

Craniofacial defect History of a difficult airway

Potential for difficult airway

Active gastroesophageal reflux or vomiting

In poor control, with or without medical or surgical treatment

Hypotonia and lack of head control

Patient may not be able to maintain his or her airway without assistance

Allergies to barbiturates

Usually the mainstay of a sedation protocol; also allergy to other sedatives to be administered

Prior failed sedation

Unable to be sedated or unsuccessful imaging study because of excessive movement


Unlikely to be ablated with sedation



In addition to the usual preanesthetic evaluation issues, chronically ill children often have electrolyte disturbances, coagulation and hematologic abnormalities, and hemodynamic instability. A consent for the administration of general anesthesia or procedural sedation must be obtained.

Is gastroesophageal reflux a contraindication to procedural sedation? Because gastroesophageal regurgitation is common in infants, a detailed clinical history should be taken with regard to the incidence and timing of the regurgitation. If the reflux is predictable (i.e., only associated with mealtimes or soon thereafter), children are usually approved for procedural sedation. NPO guidelines are adjusted to minimize the risk of reflux. For example, if the infant refluxes within 2 hours of solid feeds but never after 3 hours, then the NPO guidelines for this infant may be extended to 6 hours for solids.


The selection of an anesthetic technique in an extramural location depends on the patient's underlying medical condition, age, drug tolerance, and anticipated procedure. The airway management may be influenced by the procedure itself, anticipated postprocedural course (intensive care unit, postprocedural intubation), and past anesthetic course (difficult intubation). The assistance of the department of anesthesia is often sought when sedation administered by the radiologist has failed ( Hubbard et al., 1992 ). Beware that parents and radiologists may have the unrealistic expectation that anesthesia provides ideal conditions and ensures successful completion of the procedure.

Premedication has many purposes—the relief of anxiety, easy separation from parents, sedation, analgesia, amnesia, reduction of salivary and gastric secretions, elevation of gastric pH, and decreased cardiac vagal activity. Medications should be adjusted to the psychological and physiologic conditions of both the patient and family. Parent-present inductions may be offered when the presence of a parent has a calming effect on the child. In the event of potentially detrimental parent anxiety, premedication may be preferable to a parent-present induction ( Kain et al., 2001 ). (See Chapter 7 , Psychological Aspects, and Chapter 8 , Preoperative Preparation.)

Barbiturates may be useful as a sole method of providing sedation. Pentobarbital (Nembutal), for example, has the advantage of providing sedation, minimal respiratory and circulatory depression, and rare adverse events ( Karian et al., 2002 ). Barbiturates have no analgesic properties. They can produce paradoxical reactions, especially in children. No antagonist to barbiturates is available; dosing should be carefully titrated. Intravenous pentobarbital via titration has been used successfully by radiologists while monitoring oral and nasal airflow, oxygen saturation (with a pulse oximeter [SpO2]), end-tidal carbon dioxide, and cardiac rate and rhythm, with transient decreases in SpO2 in up to 7.5% of patients; interventions have included stimulation and head repositioning ( Strain et al., 1988 ; Connor et al., 2003 ). Other studies have described the use of pentobarbital in both the oral and intravenous forms ( Chung et al., 2000 ; Mason et al., 2001 ). For infants younger than 1 year, oral pentobarbital is more successful and carries a lower rate of adverse events compared with chloral hydrate ( Rooks et al., 2003 ). The long half-life of pentobarbital (24 hours) mandates careful and conservative recovery and discharge guidelines. The dosage of pentobarbital is 2 to 6 mg/kg PO, up to 9 mg/kg in patients who are receiving barbiturate therapy.

Sodium thiopental in a mean induction dose of 6 mg/kg and a mean total dose of 8.5 ± 3 mg/kg has been used successfully as the sole anesthetic for CT/MRI in 200 children from 1 month to 12 years of age ( Spear et al., 1993 ). Methohexital has a shorter recovery time than thiopental and is more effective than oral chloral hydrate ( Manuli and Davies, 1993 ). Methohexital-induced seizures in patients with temporal lobe epilepsy have been reported; thiopental or pentobarbital is an alternative for these patients ( Rockoff and Goudsouzian, 1981 ). For patients taking barbiturate-containing anticonvulsant medications, a higher dose limit is generally more successful. Methohexital has also been used intramuscularly for radiotherapy at doses of 8 to 10 mg/kg. The onset time via this route is often twofold to threefold longer than that for rectally administered methohexital ( Jeffries, 1988 ).

Opiates reduce anesthetic and preprocedural and postprocedural analgesic requirements. They are reversible with naloxone. Narcotics may be unnecessary for purely nonpainful diagnostic procedures, but they may be very useful for therapeutic interventions, especially for those patients with postprocedural pain. They are also useful following anthracycline chemotherapy, with documented impaired myocardial function ( Burrows et al., 1985 ). Because narcotics depress the ventilatory response to CO2, this respiratory depression may be of particular concern for children with increased intracranial pressure (ICP). Narcotics may also worsen preexisting nausea and vomiting.

Benzodiazepines have the advantage of anxiolysis with minimal vomiting and cardiorespiratory depression. Intravenous injection of diazepam (Valium) is painful and may lead to thrombophlebitis; midazolam (Versed) is water soluble and therefore may be more suitable for intravenous or intramuscular injection. The elimination half-life of midazolam averages 2.5 hours compared with 20 to 70 hours for diazepam ( Greenblatt et al., 1981 ; Reves et al., 1985 ). Young patients or patients with significant liver disease may have prolonged duration and exaggerated effect of the benzodiazepines.

Preparation of the stomach and aspiration prophylaxis are of particular concern for urgently scheduled cases (outside of NPO guidelines) or when the medical history suggests aspiration risk. If H2-receptor antagonists are used, bronchospasm may occur in asthmatic patients because of the relative increased availability of H1-receptors. H2-blockers may also inhibit metabolism of other concurrently administered medications. Metoclopramide accelerates gastric emptying and increases tone in the lower esophageal sphincter but is associated with a significant incidence of extrapyramidal side effects in children. Ondansetron works synergistically with other agents through its vagal blocking actions in the gastrointestinal tract as well as through its inhibition of the chemoreceptor trigger zone via serotonin receptor antagonism, particularly for patients undergoing radiation therapy with pulses of chemotherapy ( Burnette and Perkins, 1992 ; Figg et al., 1993 ).

Ketamine has been very popular since the 1970s for sedation, analgesia, or anesthesia outside the operating room due to its support of the cardiovascular and respiratory systems. Ketamine-induced nightmares, hallucinations, delusions, and agitation are rare in children ( Sussman, 1974 ; Hostetler and Davis, 2002 ). Mason and others (2002) reported on a ketamine sedation program for use in interventional radiology. In this program, intravenous or intramuscular ketamine was administered in the intervention radiology suite by credentialed radiology nurses and radiologists to patients undergoing selected procedures. This protocol has allowed painful procedures to be tolerated by patients who previously would have required general anesthesia ( Mason et al., 2002 ).

Although propofol does not have a labeled indication for children younger than 3 years, propofol has been used in this age group as a means of providing sedation or anesthesia. Propofol sedation via bolus and continuous infusion for brain MRI (total dosage, 5 mg/kg per hour) has provided successful imaging conditions and allowed patients to meet discharge criteria within 20 minutes ( Vangerven et al., 1992). Fatal metabolic acidosis and myocardial failure associated with lipemic serum have been reported in five children admitted to the intensive care unit for respiratory support for URIs while being sedated with continuous-infusion propofol ( Parke et al., 1992 ). For radiation therapy patients, propofol has been successfully used when administered through a catheter by a syringe pump calibrated to deliver milliliters per hour. Bloomfield and others (1993) suggested that the syringe pump begin at the maximum infusion rate of 99 mL/hr until the patients fell asleep (usually over 1 to 3 minutes); then the infusion rate was gradually decreased during treatment until it was shut off at the end ( Fig. 25-2 ). Patients were typically awake, alert, and taking clear liquids 20 minutes later ( Bloomfield et al., 1993 ).


FIGURE 25-2  Propofol dosing via continuous infusion.



Some patients require general anesthesia because of previous sedation failures, the need for a secure airway, or procedural logistics. Newer, less-soluble anesthetic agents such as sevoflurane and desflurane have pharmacokinetic profiles that compare favorably with propofol in adults ( Van Hemelrijck et al., 1991 ); there is little reason to believe that would not be the case with children, although pediatric anesthesiologists usually avoid using desflurane because of its pungency and ability to cause airway irritability. Repeat halothane anesthetics have been associated with veno-occlusive disease in children undergoing total body irradiation, alkylating agent chemotherapy, and preparation for bone marrow transplantation ( Gentet et al., 1988 ; Griswold et al., 1988 ). The ideal volatile agent for repeat anesthetics is unclear; halothane has its potential disadvantages based on reports of hepatitis ( Kenna et al., 1987 ) and bromism ( Goudsouzian et al., 1988 ; Morrison and Friesen, 1990 ). Isoflurane produces no measurable change in liver function tests during daily anesthesia for several weeks for radiotherapy ( Jones et al., 1991 ), but its ability to cause airway irritability limits its usefulness during induction of anesthesia. Since its introduction to clinical practice in the mid 1990s, sevoflurane has become the volatile anesthetic of choice in children. Its lack of airway irritability and its ability to provide children with stable hemodynamic function coupled with its rapid onset and offset make sevoflurane a useful agent for children.

Volatile agent vaporizer performance in the MRI suite has been studied. The output of a Fortec II vaporizer (Fraser Harlake, Orchard Park, NY) varied according to vaporizer location and orientation of the bimetallic strip within the magnetic field ( Kross and Drummond, 1991 ). The movements of the bimetallic ferromagnetic temperature compensator within the MRI magnetic field altered vaporizer output by as much as 91% of the dialed output concentration. Several other vaporizers (Ohio Forane, Ohio Medical Products, Madison; WI; Ohmeda Isotec IV vaporizer, Ohmeda, Steeton, England; and Forane Vapor 19.1, Dragerwerk AG, Lubeck, Germany) were incompatible with the MRI environment because of stronger ferromagnetic internal component content or the location of a ferromagnetic spring within the temperature compensator. Measuring inspired and end-tidal levels of volatile agents when delivering a general anesthetic in the MRI environment may provide reassurance.

Regional anesthesia, although still relatively rare in pediatric operating room practice and even rarer outside the operating room, remains a valid choice in some circumstances. Intercostal nerve blocks may be very useful for lung or rib biopsies, placement of chest tubes, biliary or subphrenic drainage procedures, and insertion of biliary stents. Nerve block of the brachial plexus via the axillary, interscalene, or supraclavicular route has been reported for the brachial approach to catheterization ( Eggers et al., 1967 ; Ross and Williams, 1970 ), and neuraxial block of the lower extremities, for femoral catheterizations and percutaneous approaches to the kidney ( Lind and Mushlin, 1987 ). Spinal anesthesia has been successfully used for repeat painful radiotherapy on lower extremities, in conjunction with regional hyperthermia and limb exsanguination ( Spencer and Barnes, 1980 ).

Indwelling central catheters are implanted in the majority of radiotherapy patients and can be utilized for induction and maintenance of anesthesia, blood draws, intravenous fluid administration, and chemotherapy. Dressing changes are oftenaccomplished in conjunction with the sedation or anesthesia. Antiseptic preparation of all injection sites with povidone-iodine (Betadine) and alcohol is critical. At the end of the session, the catheter should be carefully flushed with heparinized saline. An alternative to central lines is the use of a heparin lock peripheral intravenous line, changed weekly, with careful parental instruction. Smoothness of emergence is particularly important after angiographic procedures because of the risk of dislodging a clot or bleeding at the puncture site. Some of these patients have been heparinized without protamine reversal. Unlike in adults, sandbags and weights are not routinely applied to angiographic cannulation sites of children.

A final word—the best choice may be no medication. Occasionally, the loss of self-control with sedation produces dysphoria and some patients fare better when completely awake. Minimal medication may be preferable in patients with complicated and unstable medical conditions who may not tolerate the anesthesia or sedation. Some procedures (unilateral carotid barbiturate injection or Wada test) may require conversation, interaction, and responsiveness of the patient. In these situations, no sedation may be the best alternative.


Recovery criteria and the recovery room environment after procedural sedation or general anesthesia administered in an extramural location must be no different than the postanesthesia care delivered to children after an operative procedure. JCAHO guidelines must be followed at extramural anesthesia recovery sites. Each site must have sources of supplemental oxygen, ability to deliver positive pressure ventilation, available perioperative suction and monitoring equipment, and a nursing staff trained in postanesthesia care. Discharge criteria should be established by an anesthesiologist in conjunction with the extramural service and its nursing staff.

Analgesic requirements after the procedure are extremely variable. Groin puncture may be only mildly annoying for adults, but the inability to move about and the ache of blood dissecting subcutaneously cause considerable discomfort in children. After angiography, all children require a minimum pediatric acute care unit stay of 4 hours to ensure that the puncture site does not bleed or a hematoma does not develop. Ideally, the patient should be pain free and resting supine and motionless to minimize the risk of a groin-puncture bleed or hematoma. Experienced nursing staff can recognize, manage, and call for extra help when they encounter unexpected agitation, delirium, or obtundation.

The anesthesiologist may be asked to participate in the perioperative care of patients for embolization procedures. These patients frequently experience pain or swelling after the procedure. The degree of pain depends on the extent of embolization, agent used for embolization, postembolic swelling, and amount of tissue necrosis. A variety of analgesic techniques are available, and the use of steroids perioperatively, while not directly decreasing pain, may be of benefit in reducing edema and postembolic neuritis. Postembolic swelling influences perioperative airway management for procedures in the head and neck. Pediatric patients in particular may need to remain intubated after such procedures, particularly when visible edema in the floor of the mouth, tongue, hypopharynx or oropharynx, or anterior neck could compromise a patent airway.

Nausea or vomiting may, because of the Valsalva maneuver, increase venous blood pressure, which can aggravate bleeding and swelling in puncture sites or after head and neck procedures. Hypothermia is a risk at some extramural locations because the MRI, CT, and interventional radiology equipment requires a cool environment. Heating lamps and forced-air heaters may be used when safe and appropriate. Finally, with the use of iodine-containing radiocontrast media (RCM) and sclerosing and embolizing agents, consideration must be given to adequate volume resuscitation, the risk of a contrast media reaction, and bladder catheterization for detection of oliguria, polyuria, or hematuria.


Each extramural anesthetizing location is unique with regard to conducting resuscitation. Redundancy of monitoring devices and equipment is important; one should not be limited to a single item that could malfunction at the time of resuscitation. Patients with multiple allergies, shellfish allergies, or atopic disease are at increased risk of exhibiting anaphylaxis to iodine-containing contrast media. These patients may benefit from pretreatment with steroids and antihistamines. Areas with restricted access, MRI suites in particular, should have designated adjacent locations to perform full resuscitation. These areas should be equipped with wall oxygen, suction, and full monitoring and resuscitation capability. A Laerdal (Laerdal Medical Corp., Wappingers Falls, NY) self-inflating silicone bag (with no ferromagnetic working parts) or nonferrous Jackson-Rees circuit (Meridian Medical Technologies, Columbia, MD) should always be kept inside the MRI suite.

The physicians, nurses, anesthesiologists, technologists, and support personnel must know the location of a readily accessible code cart. In addition, a hard board to be placed under the patient during resuscitation should be available. Mock codes should be performed regularly to ensure adequate flow, teamwork, and delineation of responsibilities in the event of an emergency. The MRI scanner poses a special problem; codes should never be conducted in the scanner because as support personnel rush inside to assist, nonremoved ferrous materials become projectiles and impose an even more hazardous situation. Quenching a magnet should not be an alternative. Quenching requires a minimum of 3 minutes to eliminate the magnetic field. In addition, inadequate exhaust during a quench has been known to produce hypoxic conditions in the scanner and has resulted in a patient's death. A “black quench” could melt the MRI coils and require replacement of the scanner, a costly and time-consuming undertaking. Defibrillators are not MRI compatible and may not function properly when exposed to the magnetic field ( Snowdon, 1989 ). In an emergency, the patient should be removed from the scanner to an area outside of the magnetic field. This designated area is a safe place for resuscitation and should have not only a wall oxygen source for a self-inflating (Ambu) bag but also access to appropriate monitors.

Copyright © 2008 Elsevier Inc. All rights reserved. -

Motoyama & Davis: Smith's Anesthesia for Infants and Children, 7th ed.

Copyright © 2005 Mosby, An Imprint of Elsevier



Computerized Tomography Scanning

CT differentiates between high-density (calcium, iron, bone, and contrast-enhanced vascular and cerebrospinal fluid [CSF] spaces) and low-density (oxygen, nitrogen, carbon in air, fat, CSF, muscle, white matter, gray matter, and water-containing lesions) structures. Because the scan time is quick, CT may be preferable for patients who are medically unstable and in need of rapid diagnosis, such as a child being evaluated for abuse, an intracranial hemorrhage, or an abdominal or a thoracic mass. Other indications for emergency CT scans include encephalopathy and a change in neurologic status. In these situations, the issues of a full stomach and increased ICP usually necessitate a rapid sequence induction with tracheal intubation. Head CT scan is often the preferred study in emergency situations where head trauma is involved ( Blankenberg et al., 2000 ).

The actual scanning sequences are short and can range from 10 to 40 seconds. These short scan times enable many children to complete a CT scan without any sedation, especially with parental presence and distraction techniques. When an anesthesiologist is involved, it is often for airway or failed sedation issues or for a medically complicated patient. An important aspect of some CT scans is to visualize the sinuses, ears, inner auditory canal, and temporomandibular bones and to evaluate for choanal atresia or craniofacial abnormalities. These scans may require direct coronal imaging with extreme head extension (off the end of the table between a 40- to 70-degree angle) or absolute immobility for three-dimensional reconstruction.

Any patient who is at risk for cervical instability should be properly screened before neck extension. Children with Down syndrome are at risk for atlantoaxial instability; the incidence of instability varies from 12% to 32% ( Blankenberg et al., 2000 ). Many children with Down syndrome require cervical spine radiographs before entering grade school or participation in the Special Olympics. Usually, the parents are well aware of the radiologic findings. The cervical spine films, however, do not indicate whether the child is at risk for dislocation ( Davidson, 1988 ). Rather, children who exhibit neurologic signs or symptoms such as abnormal gait, increased clumsiness, fatigue with ambulation, or a new preference for sitting games are at risk. In infants, developmental milestones (crawling, sitting up, reaching for objects, etc.) should be examined. Physical signs may include clonus, hyperreflexia, quadriparesis, neurogenic bladder, hemiparesis, ataxia, and sensory loss. The asymptomatic Down syndrome child with radiologic evidence of instability may be approved for procedural sedation, but unnecessary neck movement should be avoided. It is important to recognize that any child who displays neurologic signs or symptoms should not be sedated by either a nurse or an anesthesiologist until neurosurgical or orthopedic consultation is obtained.

Radiologists use Gastrografin (diatrizoate meglumine/diatrizoate sodium [48.29% total iodine]) when evaluating abdominal masses. Gastrografin diluted to a concentration of 1.5% is usually considered a clear liquid. The volume that is administered orally is significant: infants younger than 1 month receive 60 to 90 mL, infants between 1 month and 1 year of age may receive up to 240 mL, and children between the ages of 1 and 5 years receive between 240 and 360 mL. Because sedation or anesthesia should usually be accomplished within a window of 1 to 2 hours after ingestion of the contrast media, most “elective” NPO guidelines would be violated, yet radiologists insist that the scan must be completed while the Gastrografin is still in the gastrointestinal tract. There are no published data on optimal induction or sedation techniques as they relate to aspiration risk in these circumstances. Full-strength (3%) Gastrografin is hyperosmolar and hypertonic. All Gastrografin should be diluted to an isomolar and isotonic 1.5% concentration of neutral pH. There is one case report of 1.5% Gastrografin aspiration in a child ( Friedman et al., 1986 ) with no adverse sequelae; the risk of using a 1.5% concentration of Gastrografin seems low ( Wells et al., 1991 ).

Embolization Procedures

Interventional techniques include nonvascular and vascular intervention ( Towbin and Ball, 1988 ). In vascular interventions, embolization and sclerotherapy have become important techniques for treating vascular malformations, aneurysms, fistulas, and hemorrhage and accomplishing renal ablation. Percutaneous transluminal angioplasty and fibrinolytic therapy are gaining popularity in pediatric institutions. Even in the smallest babies, great success is being reported, and the important contribution that adequate sedation and analgesia can make to ultimate outcome has been recognized ( Diament et al., 1985 ). The basic indications for embolization are occlusion of vascular malformations, management of uncontrollable hemorrhage, medical renal ablation, and presurgical embolization of hypervascular masses.

Vascular malformations are congenital aberrant connections between blood vessels. Vascular malformations may be composed of lymphatic, arterial, and venous connections. These lesions, although present at birth, are often discrete and not clearly visible. As the child grows, the vascular malformation may expand rapidly, growing with the child. This rapid proliferative phase may occur in response to hormonal changes (pregnancy, puberty), trauma, or other stimuli ( Jackson et al., 1993 ). Vascular malformations may be high-flow or low-flow lesions, depending on which vessels are involved. High-flow lesions include arteriovenous fistulas, some large hemangiomas, and arteriovenous malformations (AVMs). Particularly with large lesions, high-output cardiac failure and congestive heart failure with the potential for pulmonary edema should be anticipated and sought in the medical history and physical examination. Low-flow lesions consist of venous, intramuscular venous, and lymphatic malformations. Surgical resection of symptomatic vascular malformations may be hazardous as well as unsuccessful: any vascular element that is not resected may enlarge and cause further problems. For this reason, the combination of invasive angiography and embolization is becoming an alternative to surgical resection.

Because vascular malformations enlarge over time, even those lesions that are asymptomatic may require intervention. Patients with symptoms may experience pain, tissue ulceration, disfigurement, airway or cardiovascular compromise, impairment of limb function, coagulopathy, claudication, hemorrhage, and progressive nerve degeneration or palsy. Because large vascular lesions require multiple embolization procedures, parents and patients are often comforted by seeing familiar anesthesiologists. Especially with these complicated patients, familiarity with the patient is another benefit of having a core group of anesthesiologists in the radiology suites. Vascular embolization is also used as a bridge to surgical resection. Successful embolization and sclerotherapy decrease the size of the malformation and reduce blood flow to the lesion, thereby decreasing surgical risks.

During embolization of vascular malformations, radiologists often strive to cut off not only the feeding vessels but also the central confluence (nidus). It is at the nidus that much of the arterial shunting occurs. Embolic agents include stainless steel minicoils, absorbable gelatin pledgets and powder, detachable silicone balloons, polyvinyl alcohol foam, cyanoacrylate glue, and ethanol. The choice of agent depends on the clinical situation and the size of the blood vessel. When permanent occlusion is the goal, polyvinyl alcohol foam and ethanol are often used. Both occlude at the level of the arterioles and capillaries. Medium-sized to small arteries may be occluded with coils, which are the equivalent of surgical ligation. Particularly in trauma situations, when only temporary (days) occlusion is the goal, absorbable gelatin pledgets or powder is used ( Coldwell et al., 1994 ).

Large hemangiomas may be associated with the coagulopathy of Kasabach-Merritt syndrome. In this condition, the hemangioma traps and destroys platelets and other coagulation factors, resulting in thrombocytopenia and an increased risk of bleeding. As the hemangioma involutes, the coagulation status improves ( Mulliken and Young, 1988 ). A condition described as systemic intravascular coagulation (SIC) can occur after the embolization of extensive vascular malformations. This condition is marked by an elevated prothrombin time (PT) with a decrease in coagulation factors and platelets.

Absolute ethanol is injected in vascular malformations to promote sclerosis. Ethanol may produce a coagulum of blood and cause endothelial necrosis ( Becker et al., 1984 ). Sclerotherapy or embolization with absolute (99.9%) ethanol increases the risk of a postprocedural coagulopathy ( Mason et al., 2001 ) marked by positive D-dimers, elevated PT, and decreased platelets. Ethanol causes thrombosis because it injures the vascular endothelium. Ethanol also denatures blood proteins. Extensive ethanol injections can cause hematuria, and urinary catheters should be inserted to monitor urine output, diuresis, and hematuria. Especially with children scheduled for day surgery, liberal fluid replacement ensures that the hematuria clears before discharge. Ethanol can cause neuropathy and tissue necrosis if not injected selectively. Using selective catheterization and direct percutaneous puncture, care is taken not to expose normal blood vessels to the ethanol. In addition to the risk of hematuria, ethanol can cause significant serum alcohol levels. Mason and others (2000) note that up to 1 mL/kg of ethanol can be administered and that serum ethanol levels have been greater than the intoxication level of 0.008 mg/dL. Patients with high serum ethanol levels are either sedated or extremely agitated, depending on their particular response to intoxication.

Embolization or balloon occlusion of AVMs, vascular tumors, intracranial aneurysms, and fistulas carries considerable risk of catastrophic results. Such risks include a sudden intracranial hemorrhage, acute cerebral ischemia, and catheter or balloon migration. If the patient is sedated, he or she may require urgent airway management. Procedures that take very long require a urinary catheter, especially if contrast media is used.

Cerebral angiography requires the patient to be motionless and requires exquisite control of ventilation. Anesthetic technique, both in choice of agent and in control of arterial CO2 tension, may affect cerebral blood flow and hence the quality of the scan. Cerebral angiography in children may be performed for the diagnosis or follow-up of Moyamoya disease, and these children should have anesthetic techniques that minimize the risk of transient ischemic attacks and stroke during the procedure ( Soriano et al., 1993 ). (See Chapter 18 , Neurosurgery.) Other considerations include controlled hypercarbia to promote vasodilation and to facilitate access and visualization of the vasculature for the radiologist. In the event of vasospasm or difficult access of small, tortuous vessels, locally administered (through the catheter) nitroglycerin in small doses (25 to 50 mcg) may facilitate visualization and access. Occlusion of the venous portion of the AVMs without complete occlusion of the arterial inflow vessels could result in acute swelling and bleeding. Vascularity reduction through occlusion of major feeder vessels is the goal of embolization of large AVMs before planned surgical excision. This may be accomplished as a staged procedure over several days, involving repeat anesthesia or sedation sessions.

Angiographic imaging may be enhanced through the use of glucagon. Glucagon is efficacious for digital subtraction angiography, visceral angiography, and selective arterial injection in the viscera. Glucagon, when needed, is administered in divided doses of 0.25 mg to a maximum of 1.0 mg intravenously. Risks include glucagon-induced hyperglycemia, vomiting (particularly when administered rapidly), gastric hypotonia, and provocation signs of pheochromocytoma ( McLoughlin et al., 1981 ; Chernish and Maglinte, 1990 ; Jehenson, 1991 ). Children who receive glucagon should be routinely administered prophylactic antiemetics.

The ability to intermittently assess neurologic function and mental status is invaluable during embolization procedures but may not be practical in children because of fear, pain, and movement. General anesthesia permits easier control of blood pressure and ventilation and eliminates concern about patient movement. For children, general anesthesia is often preferred during high-risk procedures that require immobility and periods of breath-holding. Preprocedural assessment should include any history of seizures, bleeding, treatment with anticonvulsants or anticoagulants, neurologic symptoms, and evaluation of ICP status. It is important to determine whether the patient has had any transient ischemic attacks or evidence of cerebrovascular occlusion. Vasodilator agents (calcium channel blockers) and/or nitrate derivatives may have to be administered after embolization. Because many patients are anticoagulated during the procedure, a preoperative coagulation profile should be obtained. A variety of anticoagulants may have to be on hand as well to provide prophylaxis for thrombosis ( Bidabe et al., 1990 )

The potential morbidity associated with embolization is not negligible. AVMs involving the head and neck frequently require cannulation of the external carotid artery branches and the thyrocervical trunk. All patients scheduled for embolization should be typed and cross-matched for blood. Patients who undergo embolizations of AVMs of the head and neck are at risk for stroke, cranial nerve palsies, skin necrosis, blindness, infection, and pulmonary embolism ( Riles et al., 1993 ). It is important to document full return of neurologic status after the patient is extubated.

Ultrasound-Directed Procedures

Needle biopsies and drainage procedures are directed with ultrasound guidance for diagnostic examination (kidney, liver, lung, muscle, unknown mass, unknown fluid). Percutaneous drainage of abscesses, cysts, pancreatic pseudocysts, and other fluid-containing structures can often be accomplished with ultrasound guidance. Ultrasound is useful for placement of difficult central venous catheters (CVLs) and peripherally inserted central catheters. The requirement for general anesthesia versus procedural sedation for ultrasound-guided procedures depends in part on the duration of the procedure, the location involved, the risks associated with the procedure, and any procedural requirements. The need for controlled ventilation with breath-holding may mandate an endotracheal tube and general anesthesia. Associated secondary effects of the end-organ disease must be kept in mind in the overall anesthesia care plan.

Magnetic Resonance Imaging

Atoms with an odd number of protons and/or neutrons are capable of acting as magnets. When they are aligned in a static magnetic field, they can be subjected to radiofrequency (RF) energy that alters their original orientation. With removal of the RF pulse, the nuclei rotate back to their original alignment (relaxation), and the energy released can be detected and transformed into an image. Hydrogen is the atom most often used for imaging, because it is present in most tissues as water and long-chain triglycerides.

MRI is used for the evaluation of central nervous system neoplasms, nonhemorrhagic trauma, and vascular and hemorrhagic lesions ( Barnes, 1992 ). Brain MRI is frequently performed to evaluate developmental delay, behavioral disorders, seizures, failure to thrive, apnea/cyanosis, hypotonia, and mitochondrial/metabolic disorders. Magnetic resonance angiography (MRA) is especially helpful in evaluating vascular flow and can often replace invasive angiography in follow-up evaluations of vascular malformations, interventional therapy, or radiotherapy (Edelman and Warach, 1993a, 1993b [38] [39]). MRA does not involve the injection of intravascular contrast media and thereby avoids any risk of contrast reaction.

The most common cause of image degradation when performing MRI in children is patient movement. Techniques for monitoring anesthetized or critically ill patients during MRI have been described in several excellent reports ( Karlik et al., 1988 ; Peden et al., 1992 ; Tobin et al., 1992 ). It is important to decide at the outset whether the anesthesia support should be within the magnetic field or outside of it; this will determine the configuration and composition of the equipment. Equipment located outside the magnetic field (e.g., outside of the 30- to 50-Gauss (G) line) can consist of standard equipment with long monitoring leads, and ventilation and gas aspiration tubing. The risks are related to disconnection and impaired direct contact monitoring. When placed close to the magnet, the anesthesia machine and its components must be nonferrous, with power supplied by filtered sources. All battery-operated equipment must be securely fixed in position and not moved during the examination because the homogeneity of the magnetic field is affected and the diagnostic images degraded. Most intravenous needles and catheters with metallic hubs are composed of high-grade stainless steel, which is not ferromagnetic. Infusion pumps can be placed outside of the magnetic field; the pump itself may malfunction under the influence of the strong magnetic field. When placed outside the magnetic field, extra-long small-bore tubing is required for infusion. Intravenous or inhalation general anesthesia may be used effectively; there is some suggestion that the MRI signal may be altered under the influence of general inhalation anesthetics. Sedation can be problematic in the MR environment when so little of the patient is directly visible. MRI-compatible stethoscopes and flashlights are also helpful. Intubation in the MRI scanner can be accomplished without investing in MRI-compatible laryngoscopes; the only component of the laryngoscope that is not MRI compatible is the battery in the handle, but lithium-containing batteries are MRI compatible. Be aware that some batteries labeled as “lithium” may be tainted with a ferrous-containing substance. To identify this situation, the anesthesiologist should carefully bring the battery into the MRI scanner to confirm compatibility. Nonferrous laryngoscopes should be clearly identified to minimize the risk of a projectile injury in the scanner.

Anesthetic management of children in the MRI suite is highly dependent on the availability of MRI, compatible monitors, and anesthesia gas machines and proximity of the anesthetic provider to the patient and MRI unit. Anesthetic management also depends on the availability of support personnel, the personal style and comfort level of the anesthesiologist, and, of course, the patient's particular medical history. Requiring a general anesthetic to complete a noninvasive procedure is often a frightening concept for parents. Parents do not realize that although there is no pain or discomfort involved in the procedure, the child may still need a general anesthetic to remain motionless for the scan. It is the rare child who is able to remain motionless after oral midazolam or intramuscular ketamine. One technique for general anesthesia is to perform an inhalation induction followed by placement of an LMA. During the scan, the patient maintains spontaneous ventilation. Lidocaine gel (2%) on the LMA cuff is a useful adjunct, in that the lidocaine gel can decrease the incidence of sore throat ( Keller et al., 1997 ) and retching ( Chan and Tham, 1995 ). A retrospective study of 200 patients demonstrated the usefulness of this approach ( Brimacombe et al., 1995 ). In children with upper respiratory infections, there was a lower incidence of mild bronchospasm, laryngospasm, breath-holding, and major oxygen desaturation (<90%) in the group with LMAs compared with the group with endotracheal anesthesia ( Tait et al., 1998 ). In the MRI suite, temperature monitoring can be accomplished by liquid crystal display (skin temperature).

Anesthesiologists must be aware of many personal items taken for granted—clipboards, pens, watches, scissors, clamps, credit cards, eyeglasses, paperclips, etc. Laryngoscopes and blades are not ferromagnetic, but the batteries contained within the handle are ferromagnetic. As an alternative, a plastic laryngoscope can be powered by a single, paper-covered nonmagnetic 3-V lithium battery or a DC light source ( Karlik et al., 1988 ; Peden et al., 1992 ; Tobin et al., 1992 ). Conventional electrocardiographic monitoring is not possible because as the lead wires traverse the magnetic fields, image degradation occurs; electrocardiography by telemetry is often chosen. Nonferrous pulse oximeters are available, and in some circumstances, fiberoptically cabled pulse oximeters may be shielded by aluminum foil to minimize magnetic field degradation. Burns have resulted from pulse oximetry monitoring in MRI ( Shellock and Slimp, 1989 ; Brow et al., 1993 ).

Any wire in the magnet bore that is a sizable portion of a wavelength may absorb a considerable amount of energy from the transmitting coil, and large voltages may build up on the surface of the wire with no discharge path other than free space. If the wire is poorly insulated or partly exposed, the voltage may discharge through space into the skin, causing significant local burns. Precordial stethoscopes made of nonferrous materials are acceptable, but the amount of noise generated during RF pulsing and the length of tubing required render auscultation ineffective. The use of infrared transmission of breath and heart sounds with a special microphone has been described ( Henneberg et al., 1992 ). Average noise levels of 95 dB have been measured in a 1.5-T MRI machine, comparable to noise levels of very heavy traffic (92 dB) or light road work (90 to 110 dB). Exposure to this level of noise has not been considered hazardous if limited to less than 2 hours per day ( Gangarosa et al., 1987 ). There are case reports, however, of both temporary ( Brummett et al., 1988 ) and permanent ( Kanal et al., 1990 ) hearing loss after an MRI scan. Earplugs or MRI-compatible headphones should be offered to all pediatric patients.

Although studies in mice ( Sperber et al., 1984 ) and dogs ( Shuman et al., 1988 ) suggest that exposure to magnetic fields may increase body temperature, it is unlikely that static magnetic fields up to 1.5 T have any effect on core body temperature in adult humans ( Shellock et al., 1986 ). To date, there have been no definitive studies to determine whether magnetic fields increase body temperature in the anesthetized or sedated pediatric patient. RF heating is a potential risk. The specific absorption rate (SAR) is measured in Watts per kilogram and is used to follow the effects of RF heating. The Food and Drug Administration allows an SAR of 0.4 W/kg averaged over the whole body. Ex vivo exposure of large metal prostheses to fields over six times that experienced in MRI have not revealed any appreciable heating ( Davis et al., 1981 ). There are no appreciable problems with RF in magnets less than 2 T ( Anonymous, 1988 ).

There are focal heating concerns with respect to monitoring equipment in the MRI scanner. For example, electrocardiographic leads must not have frayed or exposed wires. Any coils or loops in a conductor can cause tissue burns. There are case reports of first-, second-, and third-degree burns after MRI ( Shellock and Slimp, 1989 ). To prevent patient injury, care must be used to avoid creating a conductive loop between the patient and a conductor (electrocardiographic monitoring/gating leads, plethysmographic gating wire, and fingertip attachment). During the scan, exposed wires or conductors cannot touch the patient's skin and no imaging coil can be left unconnected to the magnet.

The biologic effect of MRI should be considered when offering parent-present induction. To date there are no reports implicating MRI for any defects in chromosomal aberration, spermatogenesis, cell growth, or behavior and memory. Studies in amphibians demonstrate that exposure to a 4-T magnetic field does not cause any defects in embryologic development ( Prasad et al., 1990 ). Most hospital MRI machines are 1.5 T. Despite these studies, pregnant mothers are usually not allowed in the scanner, regardless of their desire to be present for induction.

The characteristics of internal compression volume and respiratory minute volume in a 9-m breathing circuit in children and infants have been studied, and nomograms have been established to provide “long distance” mechanical ventilatory support for patients undergoing MRI ( Neumark et al., 1989 ). Breathing circuits should be extra-long to ensure continuity of the circuit during movement of the gantry in and out of the scanner bore. A ventilator (225/SIMV; Monaghan Medical Corp., Plattsburg, NY) specifically designed for MRI has become available, which may be particularly important for use in patients who would be adversely affected by the length of circuit needed in this circumstance ( Smith et al., 1986 ).

The American College of Radiology has established guidelines to avoid mishaps in the MRI environment ( Kanal et al., 2002 ). Special care should be taken to distinguish ferrous from nonferrous oxygen tanks. Complications and deaths have occurred due to the accidental introduction of a ferrous cylinder into the MRI environment ( Chaljub et al., 2001 ).

Additional MRI safety issues include implanted objects (i.e., cardiac pacemakers), ferromagnetic attraction creating “missiles,” noise, biologic effects of the magnetic field, thermal effects, equipment issues, and claustrophobia. Some stainless steel may contain ferritic, austenitic, and martensitic components (Steels, 1961; Dujovny et al., 1985 ; Persson and Stahlberg, 1985 ). Martensitic alloys contain fractions of a crystal phase known as martensite, which has a body-centered cubic structure, is prone to stress corrosion failure, and is ferromagnetic. Austenite is formed in the hardening process of low carbon and alloyed steels and has ferromagnetic properties. Iron, nickel, and cobalt are also ferromagnetic. For this reason, the components of any implanted device should be carefully researched before entering the magnet. Stainless steel or surgical stainless objects interacting with an external magnetic field may produce translational (attractive) and rotational (torque) forces. Intracranial aneurysm clips, cochlear and stapedial implants, shrapnel, intraorbital metallic bodies, and prosthetic limbs may move and dislodge. Some eye makeup and tattoos may contain metallic dyes and therefore cause ocular, periorbital, and skin irritation ( Scherzinger and Hendee, 1985 ; Prasad et al., 1990 ). Some tissue expanders that are used in reconstructive surgery have a magnetic port to help identify the location for intermittent injections of saline ( Liang et al., 1989 ). Personal experience has also demonstrated that Bivona (Bivona Medical Technologies, Division of UroQuest Medical Corporation, Gary, IN) tracheostomy tubes may pose a risk in the MRI environment. Although not listed on the package insert, there is ferrous material within the Bivona tracheostomy tube itself. This may not only produce rotational and translational motion but may also prove to be a thermal hazard.

The magnetic field may affect the electrocardiogram. The changes in the T wave are not due to biologic effects of the magnetic field but rather to superimposed induced voltages. This effect of the magnetic field on the T wave is not related to cardiac depolarization, because no changes to the P, Q, R, or S wave have ever been observed in patients exposed to fields up to 2 T. There are no reports of MRI affecting heart rate ( Beischer, 1969 ), electrocardiographic recording ( McRobbie and Foster, 1985 ), cardiac contractility ( Gulch and Lutz, 1986 ), or blood pressure ( Tenforde et al., 1983 ). One study, however, found that humans exposed to a 2-T magnet for 10 minutes developed a 17% increase in the cardiac cycle length (CCL). The CCL reverted to preexposure length within 10 minutes of removing the patient from the magnetic field ( Jehenson et al., 1988 ). The implications of this finding are unclear. This change in CCL in patients with normal hearts may be of no consequence. The implications of this finding for patients with fragile dysrhythmias or sick sinus syndrome, however, have yet to be determined.

Cardiac pacemakers present a special hazard in and around the MRI scanner, especially in patients who are pacemaker dependent. Most pacemakers have a reed relay switch that can be activated when exposed to a magnet of sufficient strength ( Pavlicek et al., 1983 ). This activation could convert the pacemaker to the asynchronous mode. There are at least two known cases of patients with pacemakers who died from cardiac arrest while in the MRI scanner. The autopsy of one patient determined that the death was the result of an interruption of the pacemaker in the magnetic environment ( Health, 1989 ). In addition to the risk of pacemaker malfunction, there is the chance that torque on the pacer or pacing leads may create a disconnect or microshock ( Erlebacher et al., 1986 ). Caution should be taken so that the patient with the pacemaker avoids both the MRI suite and the immediate surroundings, to a fringe area of less than 5 G. Heart valves, unlike cardiac pacemakers, are not ferromagnetic and are not a contraindication to MRI. It is critical that everyone entering the vicinity of the MRI scanner fills out a screening form that specifically lists every possible implantable device, alerting the MRI staff to any potential hazards.

Projectiles are a hazard in the MRI suite. In the presence of an external magnetic field, a ferromagnetic object can develop its own magnetic field. The attractive forces that are created between the intrinsic and extrinsic magnetic fields can propel the ferromagnetic object toward the MRI scanner. Placing a magnet outside the MRI scanner is a helpful way to test any objects as to their attraction to a magnet and avoid disasters in the MRI scanner. Some objects that have been attracted to the MRI magnet are a metal fan, pulse oximeter, shrapnel, wheelchair, cigarette lighter, stethoscope, pager, hearing aid, vacuum cleaner, calculator, hair pin, oxygen tank, prosthetic limb, pencil, insulin infusion pump, keys, watches, and steel-tipped/heeled shoes ( Kanal, 1992 ). Large objects may have so much attractive force with the MRI scanner that quenching the magnet may be the only way to release the object once it is attached to the scanner. Quenching the magnet is not without risk; this action fills the scanner with noxious helium gas, mandates extensive follow-up technical support to restart the magnet, and requires a minimum of 48 hours to regenerate the magnetic field.

Some patients experience claustrophobia and have difficulty cooperating during the study. Anxiety reactions ( Granet and Gelber, 1990 ) have been estimated to occur in 4% to 30% of patients ( Melendez and McCrank, 1993 ). Patients with extreme skeletal abnormalities such as advanced scoliosis or flexion contractures may be unable to lie motionless or supine for the duration of the scan; these patients may require general anesthesia for positioning and comfort.


Nuclear medicine is one of the oldest functional imaging disciplines. These scans are useful for identification of epileptic foci in refractory epilepsy, evaluation of cerebrovascular (Moyamoya) disease, and evaluation of cognitive and behavior disorders ( O'Tuama and Treves, 1993 ). Anesthesiologists become involved when the child's medical history suggests that procedural sedation would not be appropriate; to complete these scans, the child must remain motionless for at least 1 hour.

The two most common nuclear medicine studies that require the administration of an anesthetic are single-photon emission computed tomography (SPECT) and positron emission tomography (PET). SPECT scans use single-photon γ-emitting radioisotopes and rotating gamma cameras to produce three-dimensional brain images. SPECT scans involve the use of radiolabeled technetium-99m 99mTc; half-life, 6 hours), which has a high rate of first-pass extraction as well as intracellular trapping in proportion to regional cerebral blood flow ( Chiron et al., 1989 ). This scan is ideal when seeking seizure foci, which are associated with alterations in regional cerebral blood flow and metabolism. This scan often precedes surgical resection of the identified focus. The technetium radionuclide is ideal because it remains intracellular and can be visualized on scan hours after a seizure has occurred. Ideally, the child should be scanned within 1 to 6 hours of the seizure. The radionuclides are physiologically harmless and nonallergenic. Caretakers, however, should wear gloves to minimize contact with radiation-containing secretions.

PET scans use PET and radionuclide tracers of metabolic activity, such as oxygen or glucose metabolism ( Chugani, 1993 ; Griffeth et al., 1993 ). Unlike SPECT scans, PET scans should be performed during the seizure itself. Because of the short half-life of the glucose tracer (110 minutes), the scan is best completed during the seizure or within 1 hour thereafter.


Radiation therapy for children uses ionizing photons to destroy lymphomas, acute leukemias, Wilms' tumor, retinoblastomas, and tumors of the central nervous system. Repeat sessions are typical, requiring reliable motionlessness to precisely aim the beam at malignant cells while sparing healthy cells and remote monitoring with a child in isolation. A planning session in a simulator is often scheduled before the initiation of radiation therapy so that fields to be irradiated can be plotted and marked.

Radiation therapy is usually very brief and nonpainful and may be approached with a variety of plans for rendering the patient temporarily motionless. The key issue is the anesthesiologist's limited access to the patient. Remote video monitoring and electrocardiography and pulse oximetry are crucial. Two or three video cameras are used to look at the monitors, the chest, and the face of the patient. A CVL in young children undergoing a long course of radiation therapy helps immensely. It is important to remember that infants undergoing radiation therapy after a prolonged fast are at risk for hypoglycemia; delayed awakening or tremulousness should prompt a Dextrostix determination.

Fractionated radiation therapy is the principle of dividing the total radiation therapy course into discrete daily sessions, allowing normal tissue repair between sessions while the tumor burden is lessened or destroyed. Hyperfractionated, or multiple-session daily, radiation therapy is a modality reported primarily in adults for head and neck cancer. The rationale for twice-daily fractionation in children is that fractionation to growing bone in rats reduces the growth deficit by 25% to 30%. The hope is that other normal tissues may be similarly spared during growth ( Eifel, 1988 ; Eifel et al., 1990 ). One successful approach has been to give children an initial formula feeding 6 hours before their first treatment and keep them NPO until recovery from their second anesthetic 6 hours after the first ( Menache et al., 1990 ). However, with the current liberalization of NPO guidelines, another approach is to give children clear liquids during their recovery from the first anesthetic until 3 hours before the second anesthetic.

Stereotactic radiosurgery (gamma knife) is a major advance in the treatment of selected intracranial AVMs and tumors in children ( Loeffler et al., 1990 ). Stereotactic radiosurgery differs from external beam radiotherapy in several important ways. A focused single large fraction of radiation is used instead of smaller, daily fractions. Stereotactic radiosurgery uses relatively weak intensity γ-rays produced by 201 cobalt-60 sources, which intersect at a single point; all 201 beams converge to destroy tumors, vascular malformations, or abnormal tissue sites within the brain. Normal brain tissue surrounding the abnormality is relatively protected from radiation effects. In children and adults, most radiosurgery had originally concentrated on the treatment of small, histologically benign lesions such as vascular malformations, acoustic neuromas, and pituitary adenomas. This focus has been expanded to include malignant tumors such as solitary metastases, ependymomas, glioblastomas, and several other tissue types. For optimal results, the tumor volume is small (≤14 cm3) ( Coffey et al., 1992 ; Lunsford and Linskey, 1992 ).

Stereotactic radiosurgery requires coordination among members of the departments of radiology, radiation therapy, neurosurgery, and anesthesia. The stereotactic procedure begins in either the CT or MRI suite. The child is placed in a stereotactic head frame that is screwed into the cranium. Most adults are able to tolerate this entire procedure with local anesthesia or sedation. The neurosurgeon infiltrates the skin with a local anesthetic before applying the head frame. Adults and older children who tolerate this procedure with sedation alone may vomit as a result of the anxiety, the headache, or the location of the tumor itself. Because the head frame is heavy and cumbersome, it is difficult for the vomiting patient to turn his or her head to protect the airway. Pediatric patients (including most teenagers) typically require a general anesthetic. General anesthetic with tracheal intubation is induced and a nasogastric tube is placed before placement of the head frame. The key for removal of the head frame is taped to the frame itself.

Calculations for dose and the three-dimensional coordinates for the beam may take several hours to compute following the initial radiologic study and head frame placement. A variety of anesthetic techniques are used, including continuous infusion, volatile anesthetics, or combinations of both. Some patients do well with sedation and spontaneous breathing, but younger patients are usually mechanically ventilated. An initial CT scan is followed by computer calculations. Once the calculations are complete, the patient is transferred to the radiosurgery suite for irradiation. After irradiation, the patient is allowed to emerge from the anesthetic ( Loeffler et al., 1990 ). The most common perioperative problem is nausea and vomiting, probably due to sensitivity of the chemoreceptor trigger zone centers to radiation.

Stereotactic radiation therapy is more precise localization of the fractionated radiation dose over the same duration of time as conventional radiation therapy, with the adjunctive use of a head frame. Considerations for the head frame include ease of application, reliability, ability to deliver supplemental oxygen and support the airway with a facemask if needed, and rapid removal of the facial restraint should it become necessary.

Total body irradiation (TBI) is associated with vomiting. Factors involved with postprocedural emesis include patient movement, single-dose (compared with fractionated) radiation, and patient age of greater than 10 years. Sedation has been found to decrease the incidence of vomiting with TBI, as has general endotracheal anesthesia ( Whitwam et al., 1978 ; Westbrook et al., 1987 ). Propofol sedation might be ideal, given its sedative, recovery, and antiemetic properties.

Anesthesia may alter the radiosensitivity of tumors. A biphasic response to tumor radiosensitivity has been found in mice under ketamine and diazepam anesthesia ( Nias and Perry, 1989 ). Although cellular oxygen metabolism is reduced by the anesthetic as well as by hypothermia, the blood supply to the mouse tumors remained intact and therefore enhanced radiosensiti-vity after 25 minutes, with less radiosensitivity at 10 minutes. Oxygen enrichment tended to enhance sensitivity with time as well. A radioprotective effect and shorter tumor growth delay have been found in phenobarbital-anesthetized mice compared with ketamine-anesthetized mice undergoing radiotherapy based on decreased tumor blood volume. Hypoxic cell sensitizers have been found to decrease the protective influence of anesthetics on normal lung tissue of previously anesthetized mice, whereas they do not alter the sensitivity of normal lung tissue in unanesthetized mice ( Down et al., 1983 ).


Endoscopic Procedures

Gastrointestinal endoscopy has become a routine part of patient care, and as such it constitutes the bulk of procedures performed by a pediatric gastroenterologist ( Fox, 1998 ). Depending on the patient and the type of procedure contemplated (therapeutic versus diagnostic), children may require no or minimal to moderately deep sedation or general anesthesia. Minimal sedation may impair cognitive function and coordination while ventilatory and cardiovascular functions are relatively unaffected. However, pediatric patients often are uncooperative and do not tolerate endoscopic procedures with minimal or moderate sedation, necessitating deeper sedation or general anesthesia to successfully accomplish the procedure ( ASA, 2002 ).

The inability to independently maintain ventilatory function and to respond purposefully increases the risk involved with deep sedation. It has been recommended that deep sedation and general anesthesia be performed by an anesthesiologist ( Saint-Maurice, 1992 ; Wolfe and Rao, 1992 ; Hassall, 1993, 1994 [63] [64]; Dillon et al., 1998 ; Bouchut et al., 2001 ; Koh et al., 2001 ). Furthermore, if the majority of these cases could be done in the endoscopy suite rather than the operating room, then multiple advantages can be realized—increased operating room use for surgical procedures and improved efficiency and turnover in scheduling in the smaller and more manageable endoscopy suite. Children with more complex medical problems, anticipated airway difficulties, morbid obesity, or behavioral problems can undergo their procedure in the operating room. Regardless of the site of the procedure, all patients scheduled for endoscopy should be evaluated before the procedure to confirm that they are appropriate candidates. In addition, the anesthetic technique depends on the procedure, the patient, and the skill of the endoscopist as well as the limitations and capabilities of the endoscopy suite (e.g., total intravenous anesthesia techniques in the absence of appropriate scavenging).

Procedural sedation is readily achieved with an intravenous anesthetic combining a sedative (e.g., midazolam), an opioid (e.g., fentanyl, alfentanil, remifentanil), and a hypnotic (e.g., propofol). Spontaneous ventilation without the patient's airway being intubated has been shown to be a safe and effective technique ( Bouchut et al., 2001 ; Koh et al., 2001 ). The majority of complications are respiratory and usually occur during an esophagogastroduodenoscopy (EGD); these complications include apnea, laryngospasm, bronchospasm, and airway obstruction. Most problems resolve after withdrawal of the endoscope and positive pressure ventilation with a tightly fitting mask, but some patients require endotracheal intubation to secure an airway and have the procedure safely completed.


Access to the airway is obviously limited once a transoral endoscope is in place. It is important to maintain spontaneous ventilation during deep sedation because any airway intervention needed typically requires the removal of the endoscope. The two most stimulating portions of the EGD are transoral and transpyloric passage of the endoscope. A smooth endoscope insertion can be aided by topical spray of local anesthesia to the oropharynx to help eliminate coughing and gagging.

A majority of the respiratory complications noted previously occur during EGD, especially in infants and younger children, compared with colonoscopy. It has been suggested that this results from a combination of factors that include the large size of the endoscope and partial airway obstruction resulting in hypoxemia. Abdominal distention secondary to air introduced into the stomach may impair diaphragmatic excursion and lead to hypoventilation. This has led several groups to select 6 months as the age before which general anesthesia with endotracheal intubation is required for the procedure, due to a higher respiratory complication rate in this age group ( Wolfe et al., 1992 ; Koh et al., 2001 ).


Access to the airway is unimpeded during a colonoscopy. Deep sedation can be achieved more readily, and if respiratory problems occur, airway interventions are straightforward to manage. Patients undergoing colonoscopy also experience increased stimulation during certain parts of the procedure, such as traversing the colon to the cecum. At times, abdominal pressure is applied to help guide the colonoscope. The depth of the anesthetic should be adjusted accordingly.

Endoscopic Retrograde Cannulation of the Pancreas

Although many institutions report success of procedural sedation in pediatric patients undergoing ERCP ( Teng et al., 2000 ; Prasil et al., 2001 ), general anesthesia with endotracheal intubation may make the procedure easier to perform, especially if the procedure is long, the patient has significant comorbid diseases, or the procedure is performed with the patient in the prone position.

Psychiatric Interviews

Intravenous sodium amobarbital has a long history as an adjunct to psychotherapy, having found its peak use during World War II and immediately thereafter in aiding soldiers to deal with the stresses and trauma of combat ( Zonana, 1979 ). This technique has had a resurgence for diagnostic and therapeutic interventions in adults, but pediatric reports are rare. Weller and others (1985) reported success in the diagnosis and treatment of prepubertal children and emphasized the “backup” of an anesthesiologist. The induction of a tranquil state until signs of sedation occur, such as slurred speech, a sense of fatigue, difficulty counting backwards, and “basal” vital signs, are not unlike our daily efforts at anxiolysis during monitored anesthesia care. The use of a bispectral index (BIS) monitor may prove to be a useful adjunct as well ( Palmer et al., 2001 ). The psychiatric interview process under these conditions is fascinating to participate in, if only as an observer. Memory retrieval, such as the uncovering of relationships between current psychopathology and earlier traumatic life events, and symptom removal via therapeutic suggestions are examples of interventions facilitated by the pharmacologically induced relaxed state made possible by the anesthesiologist during the interview.

Copyright © 2008 Elsevier Inc. All rights reserved. -

Motoyama & Davis: Smith's Anesthesia for Infants and Children, 7th ed.

Copyright © 2005 Mosby, An Imprint of Elsevier


As anesthesiologists find themselves participating in the care of patients requiring increasingly sophisticated imaging technology, it is appropriate to examine the risks for patients and staff exposed to the types of high energies and contrast agents used.


In a comprehensive review, Goldberg (1984) noted approximately 5% of radiologic examinations with RCM are complicated by adverse reactions, with one third of these being severe and requiring immediate treatment. Reactions occur most commonly in patients between 20 and 50 years of age and are relatively rare in children. The male/female ratio is about 2.5:1, not dissimilar to the gender distribution of other allergies such as latex, aspirin, and neuromuscular blocking agents. With a history of atopy or allergy, the risk of a reaction is increased 1.5- to 10-fold. Reactions vary from mild, subjective sensations of restlessness, nausea, and vomiting to a rapidly evolving, angioedema-like picture accompanied by bronchospasm, arrhythmias, and cardiac arrest. Because of the high osmolar concentration of these agents (often >1,000 mOsm and sometimes >2,000 mOsm), caution should be exercised with patients who have a limited cardiovascular reserve such as patients in congestive heart failure or those with cardiomyopathy. In addition, volume-depleted young children who have been kept NPO for prolonged intervals or who have had bowel preps should be hydrated before RCM administration.

Patients dependent on a full intravascular volume status (patients with sickle cell disease, restricted pulmonary circuit volume with cyanotic congenital heart disease, with arteriovenous shunts, etc.) should be monitored carefully for an initial rise in filling pressures and intravascular volume and subsequent diuresis following an osmolar load. Patients with impaired excretory mechanisms, such as those in renal failure, must be monitored closely following high osmolar loads. Low osmolar RCM are relatively safe with regard to life-threatening reactions, but moderate non–life-threatening reactions that require some treatment occur 0.2% to 0.4% of the time and a severe life-threatening reaction can occur in 0.04% of patients ( Thomsen and Bush, 1998 ).

RCM contain iodine because its high density and low toxicity make it an ideal agent for visualization and differentiation. The iodine is filtered through the glomeruli and is not reabsorbed by either the glomeruli or the tubules. Because the contrast agent is hypertonic relative to plasma, an initial hypertensive response is usually followed by a hyperosmotic diuresis with the potential for hypotension. Equilibration with the extracellular fluid compartment occurs within 10 minutes, heralded by the onset of diuresis. Special attention should be paid when administering iodine RCM to any child with a history of congestive heart failure. The initial increase in blood volume may precipitate cardiovascular compromise. Patients with hepatic or renal dysfunction should be observed closely for signs of impaired excretion of the RCM. Sickle cell disease presents its own inherent risks associated with RCM.

After administration, the increase in blood osmolarity may precipitate shrinkage, clumping, and, ultimately, sickling of erythrocytes. One theoretical concern in patients with sickle cell disease is sickle crisis and vascular occlusion. Sickled cells are known to align with external magnetic fields to which they are exposed; it is unknown how this theoretical concern compares, for example, with the normal forces of deformation imposed on red cells of patients with sickle cell disease in their normal course through the vascular tree ( Kanal et al., 1990 ).

Gadolinium dithylenetriaminepentaacetic acid (DTPA) is a low osmolar ionic contrast medium used for MRI, with a slower clearance in neonates and young infants than in adults, yielding longer windows for imaging ( Elster, 1990 ). Free gadolinium has a biologic half-life of several weeks, with uptake and excretion taking place in the kidneys and liver. Unfortunately, free gadolinium is quite toxic and is chelated to another structure that restricts the ion and decreases its toxicity. The clinical safety profiles for the three available MRI contrast agents are quite comparable, with the most common reactions being nausea, vomiting, hives, and headache. Local injection site symptoms include irritation, focal burning, or a cool sensation. Transient elevations in serum bilirubin (3% to 4% of patients) have been reported, and a transient elevation in iron for Magnevist and Omniscan (15% to 30% of patients) occurs, which tends to reverse spontaneously within 24 to 48 hours ( Van Wagoner and Worah, 1993 ). Anaphylactoid reactions occur in 1:100,000 to 1:500,000 of doses and are rarer in children.

The earlier literature states that patients who have had anaphylactic reactions to shellfish are at increased risk of anaphylactoid reaction to RCM. The irony of the statement is that it may be correct, but for nonobvious reasons. The original rationale was that shellfish contain high quantities of iodine and therefore it was assumed that there would be a risk of cross-reactivity. However, neither shellfish allergy nor RCM reactions are due to iodine. Atopy per se is a risk factor; the association between atopy, anaphylactic reactions to shellfish, and a possible predisposition to an RCM reaction may indeed be valid.

The treatment of severe allergic reactions, whether anaphylactoid or anaphylactic, is no different than that for any other allergic reaction. Epinephrine, aminophylline, atropine, diphenhydramine, and steroids have all been used to control varying degrees of adverse reactions. A patient who requires RCM administration and who has had a previous reaction to RCM has an increased (35% to 60%) risk for a reaction on reexposure. Pretreatment of these high-risk patients with prednisone and diphenhydramine 1 hour before RCM administration reduces the risk of reactions to 9%; the addition of ephedrine 1 hour before RCM administration further reduces the rate to 3.1% ( Kelly et al., 1978 ; Greenberger, 1984 ).

Allergic reactions rarely occur with oral agents. The incidence of severe anaphylactoid reactions to gastrointestinally administered agents is approximately 1:2,500,000 and the causes remain unknown. There are no pretreatment protocols established for these types of reactions and no well-defined risk factors. Gastrointestinal complications include nausea, vomiting, and diarrhea. One of the factors that may protect against having an allergic reaction is the poor absorption of oral iodinated contrast agents. Indeed, disruption of the gastrointestinal mucosa is recognized as causing an increase in absorption of oral contrast and the urinary excretion of contrast in a gastrointestinal study is a well-recognized sign. Yet rarely, they are associated with bronchospasm, flushing, periorbital edema, pruritis, rash, rhinitis, and urticaria.


Radiation exposure is directly proportional to the duration of the procedure and inversely proportional to the square of the distance from the source. Henderson and others (1991) monitored the radiation exposure of 16 pediatric anesthesia fellows during a 2-month period. Fellows assigned to the cardiac catheterization laboratory had fluoroscopy exposure times of 14 to 85 minutes per case, typically for two or three cases per day. For these anesthesiologists, badge readings ranged from 20 to 180 mrem/month. All noncardiac anesthesia fellows had undetectable (<10 mrem/month) levels. All fellows wore lead aprons, 50% wore a thyroid shield, and one stepped at least 10 feet away from the source during every exposure; this latter fellow had a reading of 30 mrem, despite having spent 26 hours in the catheterization laboratory. The annual maximum permissible dose (MPD) for nonradiation workers (including anesthesiologists) is 100 mrem or 1 milli-Sievert (mSv, Systeme Internationale Units). For comparison, the MPD for radiation workers is 50 mSv annually and 10 mSv times age cumulatively. MPD during pregnancy for radiation workers (per gestation) is 5 mSv.


MRI exposes the patient (and the health care workers surrounding the patient) to a static magnetic field, a rapid switched spatial gradient magnetic field, and RF magnetic fields. The static magnetic field, which causes alignment of unpaired tissue protons, may cause movement of ferromagnetic devices such as vascular clips, ventricular shunt connectors, casings for pacemakers, and control devices for pacemakers. Metallic devices in other areas, particularly when invested with fibrous tissue, are less problematic ( Shellock and Crues, 1988 ; Shellock, 1989 ). As mentioned previously, tissue expanders may have magnetic ports to facilitate identification of the injection site. Despite their low mass, such ports have a potential for torque and movement in the presence of a strong magnetic field; the specific type of tissue expander should be identified before patient entry into the MRI suite ( Liang et al., 1989 ). Assessment of risk in patients with implants or other possibly ferromagnetic devices or objects consists of a careful history that includes penetrating wounds, physical examination to look for scars, and possibly a plain radiograph of the region in question ( Pohost et al., 1992 ). Other concerns have been increased blood pressure, cardiac arrhythmias, and impaired mental function. While these issues have been described or theorized on an experimental basis, little clinical documentation is available.

The magnetic field generates an electrical current that is 2 to 3 orders of magnitude less than that of a defibrillator (10 mA/M2 compared with 1,000 to 10,000 mA/M2). This current strength may nevertheless reprogram a programmable pacemaker and interfere with its function ( Erlebacher et al., 1986 ). Exposure to a strong external magnetic or electromagnetic field can lead to conversion of a demand pulse generator from synchronous to asynchronous mode, damage to the reed switch (which activates the fixed-rate pulse generator), reprogramming of pacemaker parameters, induction of currents in the electrode wires, or displacement of the generator itself. Indeed, it is the sensitivity of some reed switches that has determined the “safety boundary” of magnetic resonance devices as being 5 G (5 × 10-4 T). Patients with implantable defibrillators-cardioverters, implantable infusion (e.g., insulin) pumps, cochlear implants, and neurostimulators are all at risk for having the implant device reprogrammed on exposure to the magnetic field. Defibrillator failure has been reported in the MRI environment ( Snowdon, 1989 ).

As stated previously, RF pulses cause heat production in metallic implants and coiled wires such as electrocardiographic cables or pulse oximeter cables if they are looped and laying on the patient's skin. Patients with compromised thermoregulatory abilities, such as those with cardiac problems or fever or taking certain drugs, may be at particular risk. Included in this group are infants, whose SAR is greater than that of adults because of the greater ratio of body surface area to body mass. SAR refers to the energy absorption (e.g., increasing body temperature) with an increase in the total amount of RF energy absorbed ( Fitzsimmons, 1992 ).

Increased reports of vertigo, nausea, and a metallic taste have been found in a study on human exposure to a 4-T magnetic field (whole body scanner) ( Schenck et al., 1992 ). Fertilized frog embryos exposed to a 4-T magnetic field did not demonstrate any adverse effects on early development ( Prasad et al., 1990 ). An increase in cardiac cycle length of 17% was found in healthy volunteers in a 2-T environment after 10 minutes of exposure, causing speculation about the effect of the 2-T environment on the sinus node ( Jehenson et al., 1988 ). This may be of particular concern in patients with a preexisting arrhythmia history. Of more significant concern in pediatric patients is the potential for hypothermia because of the airflow directed through the scanner cavity and the inability to control room temperature or use radiant warmers. The use of warm intravenous fluid bags, thermal packs, and blankets can decrease heat loss. Excellent reviews of monitoring considerations and equipment choices in the MRI environment as well as patient safety principles are available ( Kanal et al., 1990 ; International Non-Ionizing Radiation Committee of the International Radiation Protection Association, 1991 ;Fitzsimmons, 1992 ; Menon et al., 1992 ; New York Academy of Sciences, 1992 ; Patteson and Chesney, 1992 ; Pohost et al., 1992 ), and the American College of Radiology published a white paper on magnetic resonance safety ( Kanal et al., 2002 ).

MRI and spectroscopy do not use ionizing radiation. Secondary harmful effects, such as magnetic objects becoming projectiles within the magnetic field as they approach the bore of the magnet and potentially causing injury, are a consideration ( Chu and Sangster, 1986 ; Chaljub et al., 2001 ). Patients (and anesthesiologists!) with metallic implants such as vascular clamps, hemostatic clips, dental devices, heart valve prostheses, intravascular coils, filters and stents, ocular implants, orthopedic implants, otologic implants, shrapnel, penile implants, and vascular access ports must be individually evaluated for their risk in the MRI environment ( Cahalan et al., 1987 ; Shellock and Curtis, 1991 ).

As with individual precautions, equipment precautions should be taken for all ferromagnetic objects such as intravenous stands, oxygen and nitrous oxide cylinders, and monitoring equipment. The anesthesia machine, if used in the scanning room, should be outfitted with aluminum gas cylinders and kept in the corner of the room. Anesthesia machines especially designed to be MRI compatible are readily available.


If a child with a known difficult airway requires intubation to complete the scheduled procedure, it is wise to perform the anesthetic induction in the operating room, an area where access to emergency airway equipment is readily available. Regardless of an anesthesiologist's comfort level and familiarity with extramural environments, the same depth of backup coverage is simply not available.

The unrecognized difficult airway is problematic in a remote location; it is important to have LMAs stocked in all extramural anesthesia carts. If a child cannot be intubated or mask ventilated, LMAs can provide a successful alternative. Case reports describe the successful use of LMAs in children with difficult craniofacial anomalies, such as Goldenhar's syndrome ( Fan et al., 1995 ; Haxby and Liban, 1995) and even Pierre Robin association ( Hansen et al., 1995 ). Similarly, a lightwand may facilitate endotracheal intubation in the child with a difficult airway ( Holzman et al., 1988 ).

It is important to recognize that the airway that had not been difficult on induction may become difficult on emergence following sclerotherapy with alcohol and subsequent tissue edema, particularly at the base of the tongue, the neck, or the mediastinum ( Furst et al., 1996 ; Ohlms et al., 1996 ; Fishman, 1999 ). These patients often require several days of airway support and ongoing evaluation in the intensive care unit until airway swelling is no longer a concern.


Transfusion requirements are rare in extramural locations, yet preprocedural anemia, accidental perforation of vascular structures, or medical transfusion requirements such as sickle cell disease or prematurity may require transfusion therapy. Equipment familiar to the anesthesiologist and identical to that available in the operating room is a welcome sight in a life-threatening emergency. Calling for additional help, establishing additional vascular access, and coordinating with the blood bank are crucial. Having a runner available may be critical when there is no designated “circulating nurse.” It may become necessary to involve a surgeon urgently and transport the patient to the operating room, in which case it would be optimal to have another anesthesia team set up the operating room while the patient is being prepared for transport.

Copyright © 2008 Elsevier Inc. All rights reserved. -

Motoyama & Davis: Smith's Anesthesia for Infants and Children, 7th ed.

Copyright © 2005 Mosby, An Imprint of Elsevier


We reviewed some general issues as well as some specific situations for anesthetizing children outside the operating room. There is no “correct technique” for delivering anesthesia to patients in these areas. Versatility must be maintained to adapt to many different clinical situations and remote locations. Because of the evolution of specialized equipment and procedures for radiology in particular, it is likely that the involvement of anesthesiologists in caring for children outside of the operating room will increase in years to come.

Copyright © 2008 Elsevier Inc. All rights reserved. -

Motoyama & Davis: Smith's Anesthesia for Infants and Children, 7th ed.

Copyright © 2005 Mosby, An Imprint of Elsevier


American Academy of Pediatrics, Section on Anesthesiology, 1999. American Academy of Pediatrics, Section on Anesthesiology : Guidelines for the pediatric perioperative anesthesia environment.  Pediatrics  1999; 103:512.

American Society for Metals Committee for Wrought Stainless Steels, 1961. American Society for Metals Committee for Wrought Stainless Steels : Wrought stainless steels,  Metals Park, OH, American Society for Metals, 1961.

American Society of Anesthesiologists, ASA Task Force on Sedation and Analgesia by Non-anesthesiologists, 2002. American Society of Anesthesiologists, ASA Task Force on Sedation and Analgesia by Non-anesthesiologists : Practical guidelines for sedation and analgesia by non-anesthesiologists.  Anesthesiology  2002; 96:1004.

Anonymous, 1988. Anonymous : Consensus conference. Magnetic resonance imaging.  JAMA  1988; 259:2132.

Barnes, 1992. Barnes P: Imaging of the central nervous system in pediatrics and adolescence.  Pediatr Clin North Am  1992; 39:743.

Becker et al., 1984. Becker GJ, Holden RW, Clatte EC: Therapeutic embolization with absolute ethanol.  Semin Intervent Radiol  1984; 1:118.

Beischer, 1969. Beischer DE: Vectorcardiogram and aortic blood flow of squirrel monkeys in a strong superconductive electromagnet.   In: Barnothy M, ed. Biological effects of magnetic fields, 2. New York: Plenum; 1969:241.

Bidabe et al., 1990. Bidabe AM, Greselle JF, Gin AM, et al: Protocole de preparation, d—anesthesie et de surveillance en angiographie therapeutique.  Agressologie  1990; 31:280.

Blankenberg et al., 2000. Blankenberg F, Loh N, Bracci P, et al: Sonography, CT, and MR imaging: A prospective comparison of neonates with suspected intracranial ischemia and hemorrhage.  AJNR Am J Neuroradiol  2000; 21:213.

Bloomfield et al., 1993. Bloomfield EL, Masaryk TJ, Caplin A, et al: Intravenous sedation for MR imaging of the brain and spine in children: Pentobarbital versus propofol.  Radiology  1993; 186:93.

Bouchut et al., 2001. Bouchut J, Godard J, Lachaux A: Deep sedation for upper gastrointestinal endoscopy in children.  J Pediatr Gastroenterol Nutr  2001; 32:108.

Brimacombe et al., 1995. Brimacombe J, Tucker P, Simons S: The laryngeal mask airway for awake diagnostic bronchoscopy. A retrospective study of 200 consecutive patients.  Eur J Anaesthesiol  1995; 12:357.

Brose et al., 1989. Brose WG, Samuels SI, Steinberg GK: Cardiorespiratory arrest following initiation of cranial irradiation for treatment of a brain stem tumor.  Anesthesiology  1989; 71:450.

Brow et al., 1993. Brow TR, Goldstein B, Little J: Severe burns resulting from magnetic resonance imaging with cardiopulmonary monitoring. Risks and relevant safety precautions.  Am J Phys Med Rehabil  1993; 72:166.

Brummett et al., 1988. Brummett R, Talbot J, Charuhas P: Potential hearing loss resulting from MR imaging.  Radiology  1988; 169:539.

Burnette and Perkins, 1992. Burnette PK, Perkins J: Parenteral ondansetron for the treatment of chemotherapy and radiation-induced nausea and vomiting.  Pharmacotherapy  1992; 12:120.

Burrows et al., 1985. Burrows FA, Hickey PR, Colan S: Perioperative complications in patients with anthracycline chemotherapeutic agents.  Can Anaesth Soc J  1985; 32:149.

Cahalan et al., 1987. Cahalan MK, Litt L, Botvinick EH, et al: Advances in noninvasive cardiovascular imaging: implications for the anesthesiologist.  Anesthesiology  1987; 66:356.

Chaljub et al., 2001. Chaljub G, Kramer L, Johnson RR, et al: Projectile cylinder accidents resulting from the presence of ferromagnetic nitrous oxide or oxygen tanks in the MR suite.  AJR Am J Roentgenol  2001; 177:27.

Chan and Tham, 1995. Chan S, Tham C: The effects of 2% lignocaine gel on incidence of retching with the use of the laryngeal mask airway.  Anaesthesia  1995; 50:257.

Chernish and Maglinte, 1990. Chernish SM, Maglinte DD: Glucagon: Common untoward reactions—review and recommendations.  Radiology  1990; 177:145.

Chiron et al., 1989. Chiron C, Raynaud C, Dulac O, et al: Study of the cerebral blood flow in partial epilepsy of childhood using the SPECT method.  J Neuroradiol  1989; 16:317.

Chu and Sangster, 1986. Chu WK, Sangster W: Potential impacts of MRI accidents.  Radiol Technol  1986; 58:139.

Chugani, 1993. Chugani H: PET in preoperative evaluation of intractable epilepsy.  Pediatr Neurol  1993; 9:411.

Chung et al., 2000. Chung T, Hoffer F, Connor L, et al: The use of oral pentobarbital sodium (Nembutal) versus oral chloral hydrate in infants undergoing CT and MR imaging—a pilot study.  Pediatr Radiol  2000; 30:332.

Coffey et al., 1992. Coffey RJ, Lunsford LD, Flickinger JC: The role of radiosurgery in the treatment of malignant brain tumors.  Neurosurg Clin North Am  1992; 3:231.

Coldwell et al., 1994. Coldwell D, Stokes K, Yakes W: Embolotherapy: Agents, clinical applications, and techniques.  Radiographics  1994; 14:623.

Committee on Drugs, American Academy of Pediatrics, 1992. Committee on Drugs, American Academy of Pediatrics : Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures.  Pediatrics  1992; 89:1110.

Connor et al., 2003. Connor L, Burrows P, Zurakowski D, et al: Effects of intravenous pentobarbital with or without fentanyl on end-tidal carbon dioxide levels during deep sedation of pediatric patients undergoing MRI scans.  AJR Am J Roentgenol  2003; 181:1691.

Cremin, 1990. Cremin BJ: Sedation for CT examinations in children.  Br J Radiol  1990; 63:316.

Davidson, 1988. Davidson R: Atlanto-axial instability in Down syndrome: A fresh look at the evidence.  Pediatrics  1988; 81:857.

Davis et al., 1981. Davis P, Crooks L, Arakawa M, et al: Potential hazards in NMR imaging: Heating effects of changing magnetic fields and RF fields on small metallic implants.  AJR Am J Roentgenol  1981; 137:857.

Department of Health and Human Services, Food and Drug Administration, 1982. Department of Health and Human Services, Food and Drug Administration : Guidelines for evaluating electro-magnetic exposure risk for trials of clinical NMR systems,  Washington, DC, Food and Drug Administration, 1982.

Diament et al., 1985. Diament MJ, Boechat MI, Kangarloo H: Interventional radiology in infants and children: Clinical and technical aspects.  Radiology  1985; 154:359.

Dillon et al., 1998. Dillon M, Brown S, Casey W: Colonoscopy under general anesthesia in children.  Pediatrics  1998; 102:381.

Down et al., 1983. Down JD, Collis CH, Jeffery PK, et al: The effects of anesthetics and misonidazole on the development of radiation-induced lung damage in mice.  Int J Radiat Oncol Biol Phys  1983; 9:221.

Dujovny et al., 1985. Dujovny M, Kossovsky N, Kossowsky R, et al: Aneurysm clip motion during magnetic resonance imaging: In vivo experimental study with metallurgical factor analysis.  Neurosurgery  1985; 17:543.

Edelman and Warach, 1993a. Edelman RR, Warach S: Magnetic resonance imaging (Part I).  N Engl J Med  1993; 328:708.

Edelman and Warach, 1993b. Edelman RR, Warach S: Magnetic resonance imaging (Part II).  N Engl J Med  1993; 328:785.

Eggers et al., 1967. Eggers GW, Metzgar MT, Plumlee JE: Axillary block and sedation for cardiac catheterization.  Anesthesiology  1967; 28:936.

Eifel, 1988. Eifel PJ: Decreased bone growth arrest with hyperfractionated irradiation in weanling rats.  Int J Radiat Oncol Biol Phys  1988; 15:141.

Eifel et al., 1990. Eifel PJ, Sampson CM, Tucker SL: Radiation fractionation sensitivity of epiphyseal cartilage in a weanling rat model.  Int J Radiat Oncol Biol Phys  1990; 19:661.

Elster, 1990. Elster AD: Cranial MR imaging with Gd-DTPA in neonates and young infants: Preliminary experience.  Radiology  1990; 176:225.

Erlebacher et al., 1986. Erlebacher J, Cahill P, Pannizzo F, et al: Effect of magnetic resonance imaging on DDD pacemakers.  Am J Cardiol  1986; 57:437.

Fan et al., 1995. Fan S, Lee T, Chen L: Long-term propofol infusion and airway management in a patient with Goldenhar's syndrome.  Acta Anaesthesiol  1995; 5:33.

Figg et al., 1993. Figg WD, Graham CL, Hak LJ, et al: Ondansetron: A novel antiemetic agent.  South Med J  1993; 86:497.

Fishman, 1999. Fishman S: Vascular anomalies of the mediastinum.  Semin Pediatr Surg  1999; 8:92.

Fitzsimmons, 1992. Fitzsimmons JR: The design of RF systems for patient safety.  Ann N Y Acad Sci  1992; 649:313.

Fox, 1998. Fox V: Clinical competency in pediatric endoscopy.  J Pediatr Gastroenterol Nutr  1998; 26:200.

Friedman et al., 1986. Friedman B, Hartenberg M, Mulroy J, et al: Gastrografin aspiration in a 3 3/4-year-old girl.  Pediatr Radiol  1986; 16:506.

Furst et al., 1996. Furst S, Burrows P, Holzman R: General anesthesia in a child with a dynamic, vascular anterior mediastinal mass.  Anesthesiology  1996; 84:976.

Gangarosa et al., 1987. Gangarosa R, Minnis J, Nobbe J, et al: Operational safety issues in MRI.  Magn Reson Imaging  1987; 5:287.

Gentet et al., 1988. Gentet JC, Bernard JL, Aimard L, et al: Veno-occlusive disease in children after intensive chemo- and radiotherapy and repeated halothane anesthesias.  Acta Oncol  1988; 27:579.

Goldberg, 1984. Goldberg M: Systemic reactions to intravascular contrast media. A guide for the anesthesiologist.  Anesthesiology  1984; 60:46.

Goudsouzian et al., 1988. Goudsouzian NG, Alifimoff JK, Cote CJ: Isoflurane for radiotherapy in children?.  Anesthesiology  1988; 68:648.

Granet and Gelber, 1990. Granet RB, Gelber LJ: Claustrophobia during MR imaging.  N J Med  1990; 87:479.

Greenberger, 1984. Greenberger P: Contrast media reactions.  J Allerg Clin Immunol  1984; 74:600.

Greenblatt et al., 1981. Greenblatt DJ, Locniskar A, Ochs HR, et al: Automated gas chromatography for studies of midazolam pharmacokinetics.  Anesthesiology  1981; 55:176.

Griffeth et al., 1993. Griffeth L, Rich K, Dehdashti F, et al: Brain metastases from non-central nervous system tumors: Evaluation with PET.  Radiology  1993; 186:37.

Griswold et al., 1988. Griswold JD, Vacanti FX, Goudsouzian NG: Twenty-three sequential out-of-hospital halothane anesthetics in an infant.  Anesth Analg  1988; 67:779.

Gulch and Lutz, 1986. Gulch R, Lutz O: Influence of strong static magnetic fields on heart muscle contraction.  Phys Med Biol  1986; 31:763.

Hansen et al., 1995. Hansen T, Joensen H, Henneberg S: Laryngeal mask airway guided tracheal intubation in a neonate with the Pierre Robin syndrome.  Acta Anaesthesiol Scand  1995;129.S39

Hassall, 1993. Hassall E: Should pediatric gastroenterologists be i.v. drug users?.  J Pediatr Gastronterol Nutr  1993; 16:370.

Hassall, 1994. Hassall E: Who should perform pediatric endoscopic sedation?.  J Pediatr Gastroenterol Nutr  1994; 18:114.

Haxby and Liban, 1995. Haxby E, Liban J: Fiberoptic intubation via a laryngeal mask in an infant with Goldenhar syndrome.  Anaesth Intensive Care  1995; 23:753.

Health, 1989. Health C: MR product reporting program and medical device report program,  Washington, DC, U.S. Food and Drug Administration, 1989.

Henderson et al., 1991. Henderson KH, Lu JK, Rockoff MA, et al: Radiation exposure of anesthesiologists.  Anesthesiology  1991; 75:A896.

Henneberg et al., 1992. Henneberg S, Hok B, Wiklund L, et al: Remote auscultatory patient monitoring during magnetic resonance imaging.  J Clin Monit  1992; 8:37.

Holzman et al., 1988. Holzman R, Nargozian C, Florence F: Lightwand intubation in children with abnormal upper airways.  Anesthesiology  1988; 69:784.

Hostetler and Davis, 2002. Hostetler M, Davis C: Prospective age-based comparison of behavioral reactions occurring after ketamine sedation in the ED.  Am J Emerg Med  2002; 20:463.

House of Delegates, 1994.   House of Delegates: Guidelines for nonoperating room anesthetizing locations. Annual meeting of the American Society of Anesthesiologists, San Francisco, CA, 1994, American Society of Anesthesiologists.

Hubbard et al., 1992. Hubbard AM, Markowitz RI, Kimmel B, et al: Sedation for pediatric patients undergoing CT and MRI.  J Comput Assist Tomogr  1992; 16:3.

International Non-Ionizing Radiation Committee of the International Radiation Protection Association, 1991. International Non-Ionizing Radiation Committee of the International Radiation Protection Association : Protection of the patient undergoing a magnetic resonance examination.  Health Phys  1991; 61:923.

Jackson et al., 1993. Jackson I, Carreno R, Potparic Z, et al: Hemangiomas, vascular malformations, and lymphovenous malformations: Classification and methods of treatment.  Plast Reconstr Surg  1993; 91:1216.

Jeffries, 1988. Jeffries G: Radiotherapy and children's anaesthesia.  Anaesthesia  1988; 43:416.

Jehenson et al., 1988. Jehenson P, Duboc D, Lavergne T, et al: Change in human cardiac rhythm induced by a 2-T static magnetic field.  Radiology  1988; 166(1 Pt 1):227.

Jehenson, 1991. Jehenson PM: Reducing doses of glucagon used in radiologic examinations.  Radiology  1991; 179:286.

Jones et al., 1991. Jones RM, Diamond JG, Power SJ, et al: A prospective study of liver function in infants and children exposed to daily isoflurane for several weeks.  Anaesthesia  1991; 46:1087.

Kain et al., 2001. Kain Z, Mayes L, Caramico L, et al: Parental presence during induction of anesthesia. A randomized controlled trial.  Anesthesiology  2001; 84:1060.

Kanal, 1992. Kanal E: An overview of electromagnetic safety considerations associated with magnetic resonance imaging.  Ann N Y Acad Sci  1992; 649:204.

Kanal et al., 2002. Kanal E, Borgstede J, Barkovich A, et al: American College of Radiology White Paper on MR Safety.  AJR Am J Roentgenol  2002; 178:1335.

Kanal et al., 1990. Kanal E, Shellock F, Talagala L: Safety considerations in MR imaging.  Radiology  1990; 176:593.

Karian et al., 2002. Karian V, Burrows P, Zurakowski D, et al: The development of a pediatric radiology sedation program.  Pediatr Radiol  2002; 32:348.

Karlik et al., 1988. Karlik SJ, Heatherley T, Pavan F, et al: Patient anesthesia and monitoring at a 1.5 T MRI installation.  Magn Reson Med  1988; 7:210.

Keeter et al., 1990. Keeter S, Benator RM, Weinberg SM, et al: Sedation in pediatric CT: National survey of current practice.  Radiology  1990; 175:745.

Keller et al., 1997. Keller C, Sparr H, Brimacombe J: Laryngeal mask lubrication. A comparative study of saline versus 2% lignocaine gel with cuff pressure control.  Anaesthesia  1997; 52:592.

Kelly et al., 1978. Kelly J, Patterson R, Lieberman P: Radiographic contrast media studies in high risk patients.  J Allerg Clin Immunol  1978; 62:181.

Kenna et al., 1987. Kenna JG, Neuberger J, Mieli-Vergani G, et al: Halothane hepatitis in children.  Br Med J  1987; 294:1209.

Koh et al., 2001. Koh J, Black D, Leatherman I: Experience with an anesthesiologist interventional model for endoscopy in a pediatric hospital.  J Pediatr Gastroenterol Nutr  2001; 33:314.

Koskinen, 1985. Koskinen MF: Magnetic resonance imaging (MRI) site planning and economics.  Med Instrum  1985; 19:244.

Kross and Drummond, 1991. Kross J, Drummond JC: Successful use of a Fortec II vaporizer in the MRI suite: A case report with observations regarding magnetic field-induced vaporizer aberrancy.  Can J Anaesth  1991; 38:1065.

Liang et al., 1989. Liang M, Narayanan K, Kanal E: Magnetic ports in tissue expanders—a caution for MRI.  Magn Reson Imaging  1989; 7:541.

Lind and Mushlin, 1987. Lind LJ, Mushlin PS: Sedation, analgesia and anesthesia for radiologic procedures.  Cardiovasc Intervent Radiol  1987; 10:247.

Loeffler et al., 1990. Loeffler J, Rossitch EJ, Siddon R, et al: Role of stereotactic radiosurgery with a linear accelerator in treatment of intracranial arteriovenous malformations and tumors in children.  Pediatrics  1990; 85:774.

Lunsford and Linskey, 1992. Lunsford LD, Linskey ME: Stereotactic radiosurgery in the treatment of patients with acoustic tumors.  Otolaryngol Clin North Am  1992; 25:471.

Manuli and Davies, 1993. Manuli MA, Davies L: Rectal methohexital for sedation of children during imaging procedures.  AJR Am J Roentgenol  1993; 160:577.

Mason et al., 2002. Mason K, Michna E, DiNardo J, et al: Evolution of a protocol for ketamine-induced sedation as an alternative to general anesthesia for interventional radiologic procedures in pediatric patients.  Radiology  2002; 225:457.

Mason et al., 2000. Mason K, Michna E, Zurakowski D, et al: Serum ethanol levels in children and adults after ethanol embolization or sclerotherapy for vascular anomalies.  Radiology  2000; 217:127.

Mason et al., 2001. Mason K, Neufeld E, Karian V, et al: Coagulation abnormalities in pediatric and adult patients after sclerotherapy or embolization of vascular anomalies.  AJR Am J Roentgenol  2001; 177:1359.

Mason et al., 2001. Mason K, Zurakowski D, Karian V, et al: Sedatives used in pediatric imaging: Comparison of IV pentobarbital with IV pentobarbital with midazolam added.  AJR Am J Roentgenol  2001; 177:427.

McLoughlin et al., 1981. McLoughlin MJ, Langer B, Wilson DR: Life-threatening reaction to glucagon in a patient with pheochromocytoma.  Radiology  1981; 140:841.

McRobbie and Foster, 1985. McRobbie D, Foster M: Cardiac response to pulsed magnetic fields with regard to safety in NMR imaging.  Phys Med Biol  1985; 30:695.

Melendez and McCrank, 1993. Melendez JC, McCrank E: Anxiety-related reactions associated with magnetic resonance imaging examination.  JAMA  1993; 270:745.

Menache et al., 1990. Menache L, Eifel PJ, Kennamer DL, et al: Twice-daily anesthesia in infants receiving hyperfractionated irradiation.  Int J Radiat Oncol Biol Phys  1990; 18:625.

Menon et al., 1992. Menon DK, Peden CJ, Hall AS, et al: Magnetic resonance for the anaesthetist. Part I: Physical principles, applications, safety aspects.  Anaesthesia  1992; 47:240.

Morrison and Friesen, 1990. Morrison JE, Friesen RH: Elevated serum bromide concentrations following repeated halothane anaesthesia in a child.  Can J Anaesth  1990; 37:801.

Mulliken and Young, 1988. Mulliken J, Young A: Vascular birthmarks, hemangiomas, and malformations,  Philadelphia, PA, WB Saunders, 1988.

Neumark et al., 1989. Neumark J, Petricek W, Schramm W: Patientenferne beatung und Narkose bei der Kernspintomographie.  Anaesthetist  1989; 38:273.

New York Academy of Sciences, 1992. New York Academy of Sciences : Principles for the protection of patients and volunteers during clinical magnetic resonance diagnostic procedures.  Ann N Y Acad Sci  1992; 31:372.

Nias and Perry, 1989. Nias AHW, Perry PM: Variation of tumour radiosensitivity with time after anaesthetic.  Br J Radiol  1989; 62:932.

Ohlms et al., 1996. Ohlms L, Forsen J, Burrows P: Venous malformation of the pediatric airway.  Int J Pediatr Otorhinolaryngol  1996; 37:99.

O'Tuama and Treves, 1993. O'Tuama L, Treves S: Brain single-photon emission computed tomography for behavior disorders in children.  Semin Nucl Med  1993; 23:255.

Palmer et al., 2001. Palmer G, Davidson A, Sethna N, et al: Use of the Bispectral Index monitor to aid titration of propofol during a drug-assisted interview.  Paediatr Anaesth  2001; 11:245.

Parke et al., 1992. Parke TJ, Stevens JE, Rice ASC, et al: Metabolic acidosis and fatal myocardial failure after propofol infusion in children: Five case reports.  Br Med J  1992; 305:613.

Patteson and Chesney, 1992. Patteson S, Chesney JT: Anesthetic management for magnetic resonance imaging: Problems and solutions.  Anesth Analg  1992; 74:121.

Pavlicek et al., 1983. Pavlicek W, Geisinger M, Castle L, et al: The effects of nuclear magnetic resonance on patients with cardiac pacemakers.  Radiology  1983; 147:149.

Peden et al., 1992. Peden CJ, Menon DK, Hall AS, et al: Magnetic resonance for the anaesthetist, Part II: Anaesthesia and monitoring in MR units.  Anaesthesia  1992; 47:508.

Persson and Stahlberg, 1985. Persson B, Stahlberg F: Safety aspects of magnetic resonance examinations.  Int J Technol Assess Health Care  1985; 1:647.

Pohost et al., 1992. Pohost GM, Blackwell GG, Shellock FG: Safety of patients with medical devices during application of magnetic resonance methods.  Ann N Y Acad Sci  1992; 649:302.

Prasad et al., 1990. Prasad N, Wright D, Ford J, et al: Safety of 4-T MR imaging: Study of effects on developing frog embryos.  Radiology  1990; 174:251.

Prasil et al., 2001. Prasil P, Laberge J, Barkun A, et al: Endoscopic retrograde cholangiopancreatography in children: A surgeon's perspective.  J Ped Surg  2001; 36:733.

Reves et al., 1985. Reves JG, Fragen RJ, Vinik HR, et al: Midazolam: Pharmacology and uses.  Anesthesiology  1985; 62:310.

Riles et al., 1993. Riles T, Berenstein A, Fisher F, et al: Reconstruction of the ligated external carotid artery for embolization of cervicofacial arteriovenous malformations.  J Vasc Surg  1993; 17:491.

Rockoff and Goudsouzian, 1981. Rockoff MA, Goudsouzian NG: Seizures induced by methohexital.  Anesthesiology  1981; 54:333.

Rooks et al., 2003. Rooks V, Chung T, Connor L, et al: Comparison of oral pentobarbital sodium (Nembutal) and oral chloral hydrate for infant sedation during radiologic imaging.  AJR Am J Roentgenol  2003; 180:1125.

Ross and Williams, 1970. Ross DM, Williams DO: Combined axillary plexus block and basal sedation for cardiac catheterization in young children.  Br Heart J  1970; 32:195.

Saint-Maurice, 1992. Saint-Maurice C: Sedation techniques in pediatric anesthesia.  Curr Opin Anesthesiol  1992; 5:1.

Schenck et al., 1992. Schenck JF, Dumoulin CL, Redington RW, et al: Human exposure to 4.0 Tesla magnetic fields in a whole-body scanner.  Med Physics  1992; 19:1089.

Scherzinger and Hendee, 1985. Scherzinger A, Hendee W: Basic principles of magnetic resonance imaging—an update. [review].  West J Med  1985; 143:782.

Shellock, 1989. Shellock FG: Biological effects and safety aspects of magnetic resonance imaging.  Magn Reson Q  1989; 5:243.

Shellock and Crues, 1988. Shellock FG, Crues JV: MRI: Safety considerations in magnetic resonance imaging.  Magn Reson Imaging Decis  1988; 2:25.

Shellock and Curtis, 1991. Shellock FG, Curtis JS: MR imaging and biomedical implants, materials and devices: An updated review.  Radiology  1991; 180:541.

Shellock et al., 1986. Shellock FG, Schaefer DJ, Gordon CJ: Effect of a 1.5 T static magnetic field on body temperature of man.  Magn Reson Med  1986; 3:644.

Shellock and Slimp, 1989. Shellock FG, Slimp GL: Severe burn of the finger caused by using a pulse oximeter during MR imaging.  AJR Am J Roentgenol  1989; 153:1105.

Shuman et al., 1988. Shuman W, Haynor D, Guy A, et al: Superficial- and deep-tissue temperature increases in anesthetized dogs during exposure to high specific absorption rates in a 1.5-T MR imager.  Radiology  1988; 167:551.

Smith et al., 1986. Smith DS, Askey P, Young ML, et al: Anesthetic management of acutely ill patients during magnetic resonance imaging.  Anesthesiology  1986; 65:6.

Snowdon, 1989. Snowdon S: Defibrillator failure in a magnetic resonance unit [letter].  Anaesthesia  1989; 44:359.

Soriano et al., 1993. Soriano S, Sethna N, Scott R: Anesthetic management of children with Moyamoya syndrome.  Anesth Analg  1993; 77:1066.

Spear et al., 1993. Spear RM, Waldman JY, Canada ED, et al: Intravenous thiopentone for CT and MRI in children.  Paediatr Anaesth  1993; 3:29.

Spencer and Barnes, 1980. Spencer HT, Barnes PJ: Spinal anaesthesia for paediatric radiotherapy.  Anaesth Intensive Care  1980; 8:214.

Sperber et al., 1984. Sperber D, Oldenbourg R, Dransfeld K: Magnetic field induced temperature change in mice.  Naturwissenschaften  1984; 71:100.

Stokes et al., 1995. Stokes M, Soriano S, Tarbell N, et al: Anesthesia for stereotactic radiosurgery in children.  J Neurosurg Anesthesiol  1995; 7:100.

Strain et al., 1988. Strain JD, Campbell JB, Harvey LA, et al: IV Nembutal: Safe sedation for children undergoing CT.  AJR Am J Roentgenol  1988; 151:975.

Sussman, 1974. Sussman D: A comparative evaluation of ketamine anesthesia in children and adults.  Anesthesiology  1974; 40:459.

Tait et al., 1998. Tait A, Pandit U, Voepel-Lewis T, et al: Use of the laryngeal mask airway in children with upper respiratory tract infections: A comparison with endotracheal intubation.  Anesth Analg  1998; 86:706.

Tenforde et al., 1983. Tenforde T, Gaffey C, Moyer B, et al: Cardiovascular alterations in Macaca monkeys exposed to stationary magnetic fields: Experimental observations and theoretical analysis.  Bioelectromagnetics  1983; 4:1.

Teng et al., 2000. Teng R, Yokohata K, Utsunomiya N: Endoscopic retrograde cholangiopancreatography in infants and children.  J Gastroenterol  2000; 35:39.

Thomsen and Bush, 1998. Thomsen H, Bush WJ: Treatment of the adverse effects of contrast media.  Acta Radiol  1998; 39:212.

Tobin et al., 1992. Tobin JR, Spurrier EA, Wetzel RC: Anaesthesia for critically ill children during magnetic resonance imaging.  Br J Anaesth  1992; 69:482.

Towbin and Ball, 1988. Towbin RB, Ball WS: Pediatric interventional radiology.  Radiol Clin North Am  1988; 26:419.

Van Hemelrijck et al., 1991. Van Hemelrijck J, Smith I, White PF: Use of desflurane for outpatient anesthesia. A comparison with propofol and nitrous oxide.  Anesthesiology  1991; 75:197.

Van Wagoner and Worah, 1993. Van Wagoner M, Worah D: Gadodiamide injection. First human experience with the nonionic magnetic resonance imaging enhancement agent.  Invest Radiol  1993; 28:S44.

Vangerven et al., 1992. Vangerven M, Van Hemelrijck J, Wouters P, et al: Light anaesthesia with propofol for paediatric MRI.  Anaesthesia  1992; 47:706.

Weller et al., 1985. Weller EB, Weller RA, Fristad MA: Use of sodium amytal interviews in prepubertal children: Indications, procedure, and clinical utility.  J Am Acad Child Psychol  1985; 24:747.

Wells et al., 1991. Wells H, Hyrnchak M, Burbridge B: Direct effects of contrast media on rat lungs.  Can Assoc Radiol J  1991; 42:261.

Westbrook et al., 1987. Westbrook C, Glaholm J, Barrett A: Vomiting associated with whole body irradiation.  Clin Radiol  1987; 38:263.

Whitwam et al., 1978. Whitwam JG, Owen JR, Spiers ASD, et al: General anaesthesia for high-dose total-body irradiation.  Lancet  1978; 1(8056):128.

Winter, 1978. Winter J: Efficiency of utilization of a computed tomography scanner.  AJR Am J Roentgenol  1978; 131:89.

Wolfe and Rao, 1992. Wolfe T, Rao C: Anesthesia for selected procedure.  Semin Pediatr Surg  1992; 1:74.

Zonana, 1979. Zonana HV: Hypnosis, sodium amytal, and confessions.  Bull Am Acad Psychiatry Law  1979; 7:18.