There is an increasing demand for anaesthesia care in MRI suites. Soft-tissue lesions in critically ill patients can be depicted more precisely by high-speed MRI scanners. During scanning, the interaction between the strong magnetic field of the MRI scanner and the changing electrical currents in its gradient coils turns the scanner into a giant loudspeaker producing loud tapping, knocking, or chirping sounds. High levels of acoustic noise are potentially harmful to patients and staff who stay in the MRI room, for long periods, without hearing protection (earplugs or headphones) [1–4].
The American Occupational Safety and Health Administration (OSHA) and the International Electrotechnical Commission (IEC) limit the peak MRI acoustic noise to 140 dB, the maximum permitted daily noise level to 90 dB (A) for 8 h (with hearing protection) and 5 s of exposure (without hearing protection) [5,6]. The European Union (EU) Directive 2004/40/EC, on electromagnetic compatibility, restricts the presence of healthcare providers in the scanning room . Consequently, patient monitoring is generally conducted outside the MRI scanner room with or without interactive screens. There are international (ISO/TC 121/SC 3/JWG 2) and European (EN 740 and EN ISO 11196) standards for acoustic and optical alarms of anaesthesia workstations .
Despite the abundant literature on MRI-related acoustic noise hazards, to our knowledge, its consequences on anaesthesia safety concerns have not yet been investigated.
The objectives of this study were to provide subjective and objective evaluation of anaesthesia alarm audibility during MRI and to review the literature to determine whether or not to be concerned about our results.
Materials and methods
The study was performed in a 1.5-T MRI suite (Fig. 1) using the Signa high-speed scanner (Philips Medical Systems, Nederland B.V., The Netherlands) and a phantom. Institutional Review Board's approval and informed volunteer consent were obtained prior to evaluation.
Screening for normal hearing volunteers
An audioscope 3 (Welch Allyn Inc, New York, USA) was used to screen for normal hearing volunteers [9,10]. All the tests were performed by the same physician who was not involved in any other aspect of the study. The tests were conducted in a quiet room with an ambient noise within 40–60 dB (A-weighted scale). After sitting comfortably on a chair, the participant was instructed that he or she would hear faint tones of different pitches and should raise a finger each time a tone was heard. An appropriately fitting speculum was attached to the audioscope and gently inserted into the subject's ear canal. The tympanic membrane was visualized and participants without abnormalities continued the test. The audioscope was then used to present pure tones (500–4000 Hz; 25 dB threshold) at random intervals. The tests were repeated on the opposite ear. The participant was screened a second time if there was no response at one or more frequencies in either ear. Participants who failed to respond at one or more frequencies, on the second trial, were advised to consult an audiologist for further testing. Those who responded at all the frequencies were selected for the study.
Preselection of the anaesthesia alarm and MRI sequences
The sound level measurements were done by the same investigator (M.S.), and the MRI sequences were produced by the same technician throughout the study. All the tests were performed on the same MRI scanner (Fig. 1). Objective measurements of acoustic noise were obtained using a digital sound level meter (DVM322DI, Velleman Inc, Forth Worth, Texas, USA) connected to a computer for online data logging. The device met all requirements of the American National Standards Institute (ANSI) for type 2 instruments, with a measuring level of 30–130 dB, a resolution of 0.1 dB, an accuracy of 1.5 dB, and a frequency range of 31.5 Hz–8 kHz . Preliminary tests with the device placed at 15, 85 or 170 cm from the floor did not show any significant difference in sound levels. Initial measurements on the A and C scale showed that the latter was 8–10 dB higher than scale A (Appendix 1).
Following the preliminary tests, isolated sound level measurements of four MRI-compatible anaesthesia equipment alarms and three standard MRI sequences were done. The anaesthesia equipment included the Maglife C anaesthesia monitor (Schiller, Doral, Florida, USA), the Continuum infusion pump (Medrad Inc., Warrendale, Pennsylvania, USA), the Parapac ventilator (Smiths Medical System, Watford, UK) and the MRidium infusion pump (IRadimed Corp, Winter Park, Florida, USA). The volume of each device was set at its maximum level during the evaluation period. The MRI sequences included spin echo, turbo spin echo (TSE) and echo planar imaging (EPI). All measurements were weighted on the A-frequency and fast time (125 ms) response scales. The MRI sequences were performed on a 3-l fluid-filled Plexiglas phantom. The sound levels of the different anaesthesia equipment alarms were similar (Fig. 2). Therefore, we arbitrarily selected the MRidium for the study. The least noisy (i.e. spin echo) MRI sequence was selected.
Sound level measurements during MR imaging
A 3-l fluid-filled Plexiglas phantom was placed at the isocentre of the MRI. The imaging parameters were chosen according to the method described by Price et al.. The pulse sequences were maintained within a bandwidth of 62 and 125 kHz. For safety reasons, we limited its intensity to a maximum of 100 dB by selecting 40 ms repetition time, 17 ms echo time, 27 mm slice width and 250 mm field of view. All the sequences were performed in the transverse plane.
The study protocol was conducted as shown in Table 1. First, the anaesthesia alarm (MRidium: IRadimed Corp) was set at its maximum acoustic level. Twenty normal hearing volunteers evaluated their audibility of the alarm using a 10-point numerical scale (0 = not audible to 10 = maximum audibility), which was adapted from the listening test recommended by international standards (ISO 7731) . The anaesthesia alarms were triggered, without prior notice, via a remote controlled system. No ear protection device was used to permit adequate comparison of the audibility scores between inside and outside the MRI scanner room. Exposure to MRI noise during each evaluation session was less than 30 s. Each volunteer noted his or her score on a piece of paper that was placed in an unidentified envelop. The latter was sealed until data analysis. Second, a calibrated digital sound level meter (DVM322DI, Velleman Inc.) connected to a computer was used to measure the different sound sources. The device was placed on a desk (85 cm) and measurements were made at 3.5 m (point X) and 3.7 m (point Y) beyond the isocentre of the MRI scanner (Fig. 1). The sound level meter was unaffected by electromagnetic radiation. It was adjusted to be sensitive to the frequency range of human hearing, that is calibrated on the A scale with an accuracy of 0.1 dB with measurements made at 125 ms.
The Systat version 12 and Sigmaplot version 10 statistical packages (Systat Software, Point Richmond, California, USA) were used for statistical analysis. The distribution of each variable was checked for normality using the Kolmogorov–Smirov test. Comparison of data sets was achieved by a paired t-test when the data were normally distributed; otherwise, the Wilcoxon rank sum test was used. Categorical data were analysed by χ2 and Fisher's exact tests when appropriate. A P value of less than 0.05 was considered as statistically significant.
Twenty volunteers (five anaesthesiologists, five radiologists, five nurses and five technicians), mean age 35 ± 15 years, successfully completed the audioscopic tests. Five participants failed the screening tests and were advised to consult an audiologist for further testing.
Figure 2 shows the preselection sound level measurements. The four anaesthesia equipment alarms tested exhibited similar sound intensities. Peak sound levels were observed with the EPI. The final tests were performed using one anaesthesia machine (MRidium) and one MRI sequence (spin echo).
Ambient noise levels in the magnetic scanning and the control rooms were comparable, that is 52–55 dB (A-weighted scale). MRI table movement or normal conversation of staff produced sound levels between 57 and 65 dB (A). Sound levels ranged between 70 and 80 dB (A) when the MRI scanner door was abruptly closed.
The subjective audibility scores (Fig. 3) and the objective sound level measurements (Fig. 4) show that the anaesthesia alarm sound was significantly masked by the MRI spin-echo sequence particularly outside the scanning room.
This study provides, for the first time, objective data with an expected result, that is, MRI acoustic noise masks the audibility of anaesthesia equipment alarms. On the basis of the tests performed on one anaesthesia machine (the MRidium infusion pump) and one MRI sequence (spin echo), our data revealed that the audibility of anaesthesia alarms was significantly reduced when the MRI scanner was operating, the door closed and the observer was outside the scanner room. Should we be alarmed by these findings? An extensive review of the literature permitted us to revive the ongoing debate among anaesthesiologists [13,14].
Why audible anaesthesia alarms are important in MRI suites
Alarms warn anaesthesiologists of mishaps and professional standards mandate that they be used [15–17]. Ineffective coverage and failure to hear alarms have been implicated in patients' death and countless ‘near-miss’ incidents . In the operating room, audible alarms alert anaesthesiologists to problems when they are busy attending to other tasks (such as positioning the patient). In the ICU, audible alarms are important because continuous bedside surveillance of patients is impossible. MRI suites are considered as remote locations for anaesthesiologists as they have specific difficulties such as limited accessibility to the patient and higher morbidity and mortality than in the operating room . Therefore, maximum safety measures should be applied in these remote sites. Our study demonstrates, objectively, that MRI acoustic noise is a hindrance to auditory anaesthesia alarm detection. Several factors have been identified. First, MRI-related acoustic noise is enhanced when section thickness, field of view, repetition time and echo time are decreased . For safety reasons, because the measurements were obtained without ear protection, we maintained the sound level to a maximum of 100 dB by adjusting the above-mentioned parameters. Second, acoustic noise is dependent on the MR system hardware, construction and the surrounding environment. Third, the presence and size of a solid object (phantom or patient) along the magnetic bore may affect the level of acoustic noise, as the sound waves are reflected and undergo an in-phase enhancement. Hedeen and Edelstein  have observed that, with a subject in place, the sound level can be 3 dB above that measured in an empty scanner bore.
Why we do not need audible anaesthesia alarms in MRI suites
Many anaesthesiologists consider audible alarms as nuisance alarms and simply turn them off [13,14]. Moreover, alarms are associated with a high number of false positives (providing an alarm in a no alarming condition) and false negatives (failing to provide an alarm in an alarming condition). A 1988 report demonstrated that 75% of alarms were totally spurious and that only 3% actually indicated patient risk . Another study, conducted in the ICU, found that as few as eight out of 1455 alarms (0.5%) indicated potentially life-threatening problems . This high incidence of false positives has led in a few situations to noncompliance with professional guidelines [23,24]. It should be noted, however, that the US Joint Commission on Accreditation of Healthcare Organizations addressed the use of all clinical alarms in its Patient Safety Goal 6 for 2003 and takes the view that all clinical alarms should be responded to, irrespective of the frequency of false positives .
Our data support previous reports that the current alarm capabilities, be it the operating room, ICU or MRI suite, do not meet the needs of anaesthesiologists and effective solutions are warranted [25–27]. Guidelines of international (ANSI ISO/TC 121/SC 3/JWG 2) and European workgroups (EN 740 and EN ISO 11196) which deal with acoustic and optical anaesthesia alarms are the grounds for various companies providing MRI-compatible anaesthesia machines . Specifically, these are equipped with additional optical alarm displays, which can be observed from outside the scanning room. Further research is required, particularly in two areas: the design of alarms and measures to combat MRI noise. A previous survey has indicated that anaesthesiologists would welcome the use of more advanced technology in instrument design and that they preferred context-specific alarms . According to Edworthy and Heillier , the ideal alarm sound should be easy to localize, resistant to masking by other sounds, would not interfere with communication, would be easy to distinguish from other alarms and easy to learn and retain. In our study, the alarm sound was overshadowed by the background (MRI) acoustic noise. Subjective and objective measures were done because our perception of loudness does not reflect the sound pressure level (Appendix 1).
Current methods used to combat MRI noise include mounting the gradient coils to the floor, lining the bore with a vacuum, the design of quiet gradient coils or the development of silent MRI-pulse sequences [29–32]. Simply using a spin-echo sequence rather than a gradient-echo sequence and running the sequence with reduced gradient parameters (rise time and amplitude) can significantly reduce the levels of acoustic noise.
Because audibility is significantly reduced outside the MRI scanning room, methods permitting the transmission of alarm sounds from inside to outside the MRI scanner should be encouraged.
Limitations of the study
The present study design focused on the audibility of normal volunteers. Studies based on the general population should provide more insights into the problems associated with anaesthesia alarm detection in MRI suites. Reduced hearing acuity is one of the problems that have been documented in the operating room and ICU [33,34].
We used only those MRI-compatible anaesthesia devices that are available in our institution. Therefore, our results could not be generalized to the variety of anaesthesia equipment in the market.
The present study demonstrates a significant reduction in anaesthesia alarm audibility during MRI scanning, particularly when the anaesthesiologist is outside the scanner room. Consequently, our results suggest that, for anaesthesia safety reasons, there must be optical alarms (visible outside the scanning room) and interactive screens available without any exception. Further studies are warranted to improve the design of anaesthesia alarms and to combat the noise generated by MRI.
1 Price DL, De Wilde JP, Papadaki AM, et al
. Investigation of acoustic noise on 15 MRI scanners from 0.2 T to 3 T. J Magn Reson Imaging 2001; 13:288–293.
2 Brummett RE, Talbot JM, Charuhas P. Potential hearing loss resulting from MR imaging. Radiology 1988; 169:539–540.
3 Shellock FG, Morisoli SM, Ziarati M. Measurement of acoustic noise during MR imaging: evaluation of six ‘worst-case’ pulse sequences. Radiology 1994; 191:91–93.
4 Radomskij P, Schmidt MA, Heron CW, Prasher D. Effects of MRI noise on cochlea function. Lancet 2002; 359:1485–1486.
5 Occupational Health and Safety Act 1970: OSHA Hearing Conservation Program Amendment. Fed Reg
6 IEC 60601-2-33. Medical electrical equipment. Part 2: Particular requirements for the safety of magnetic resonance equipment for medical diagnosis. Geneva: International Electrotechnical Commission; 1995.
7 Riches SF, Collins DJ, Charles-Edwards GD, et al
. Measurements of occupational exposure to switched gradient and spatially-varying magnetic fields in areas adjacent to 1.5 T clinical MRI systems. J Magn Reson Imaging 2007; 26:1346–1352.
8 Dain SL. Anesthesia standards for facilities and equipment. Can J Anesth 2001; 48:41–47.
9 Bienvenue GR, Michael PL, Chaffinch JC, Zeigler J. The audioscope: a clinical tool for otoscopic and audiometric examination. Eur Hear 1985; 6:251–254.
10 Bagai A, Thavendiranathan P, Detsky AS. Does this patient have hearing impairment? JAMA 2006; 295:416–428.
11 Boyle Y, Bentley DE, Watson A, Jones AK. Acoustic noise in functional magnetic resonance imaging reduces pain unpleasantness ratings. Neuroimage 2006; 31:1278–1283.
12 International Standard ISO 7731. Ergonomics, danger signals for public and work areas, auditory danger signals. 2nd ed. Geneva: International Electrotechnical Commission; 2003.
13 Hagenouw RRPM. Should we be alarmed by our alarms? Curr Opin Anaesthesiol 2007; 20:590–594.
14 McIntyre JWR. Alarms in the operating room. Can J Anaesth 1991; 38:951–953.
15 American Society of Anesthesiologists. Standards for basic anesthetic monitoring. Park Ridge, IL: American Society of Anesthesiologists; 2005.
16 Catalano K. Clinical alarms management and the JCAHO standards. Biomed Instrum Techol
17 International Standard IEC 60601. Medical Electrical Equipment-Parts 1–8: general requirements for safety-collateral standard. General requirements, tests and guidance for alarm systems in medical electrical equipment and medical electrical systems. Geneva: International Electrotechnical Commission; 2005.
18 Critical alarms and patient safety. ECRI's guide to developing effective alarm strategies and responding to JCAHO'S alarm-safety goal. Health Services
19 Robbertze R, Posner KL, Domino KB. Closed claims review of anaesthesia for procedures outside the operating room. Curr Opin Anaesthesiol 2006; 19:436–442.
20 Hedeen RA, Edelstein WA. Characterization and prediction of gradient acoustic noise in MR imagers. Magn Reson Med 1997; 37:7–10.
21 Kestin IG, Miller BR, Lockhart CH. Auditory alarms during anesthesia monitoring. Anesthesiology 1988; 69:106–109.
22 O'Carroll TM. Survey of alarms in an intensive care unit. Anaesthesia 1986; 41:742–744.
23 Smith AF, Mort M, Goodwin D, Pope C. Making monitoring ‘work’: human-machine interaction and patient safety in anaesthesia. Anaesthesia 2003; 58:1070–1078.
24 Beatty PCW, Beatty SF. Anaesthetists' intentions to violate safety guidelines. Anaesthesia 2004; 59:528–540.
25 Edworthy J, Heillier E. Alarms and human behaviour: implications for medical alarms. Br J Anaesth 2006; 97:12–17.
26 Edworthy J, Heillier E. Fewer but better auditory alarms will improve patient safety. Qual Saf Healthcare 2005; 14:212–215.
27 Mondor TA, Finley GA. The perceived urgency of auditory warning alarms used in the hospital operating room is inappropriate. Can J Anesth 2003; 50:221–228.
28 Nazir T, Beatty PCW. Anaesthetists' attitudes to monitoring instrument design options. Br J Anaesth 2000; 85:781–784.
29 McJury M, Blug A, Joerger C, et al
. Short communication: acoustic noise levels during magnetic resonance imaging scanning at 1.5 T. Br J Radiol 1994; 67:413–415.
30 Katsunuma A, Takamori H, Sakakura Y, et al
. Quiet MRI with novel acoustic noise reduction. MAGMA 2002; 13:139–144.
31 Mansfield P, Haywood B, Coxon R. Active acoustic control in gradient coils for MRI. Magn Reson Med 2001; 46:807–818.
32 Hennel F. Fast spin echo and fast gradient echo MRI with low acoustic noise. J Magn Reson Imaging 2001; 13:960–966.
33 Wallace MS, Ashman MN, Matjasko MJ. Hearing acuity of anesthesiologists and alarm detection. Anesthesiology 1994; 81:13–28.
34 Momtahan K, Hetu R, Tansley B. Audibility
and identification of auditory alarms in the operating room and intensive care unit. Ergonomics 1993; 36:1159–1176.
35 Harris CM. Handbook of acoustical measurements and noise control. New York: The Acoustical Society of America; 1998. pp. 20–44.
Appendix 1: Definitions of basic audiologic terms used in this article
Sound originates from any source that vibrates and causes the air around it to vibrate, leading to pressure variation within our eardrums. It is characterized by the intensity (measured in decibels), the frequency (measured in Hertz) and the duration. The human ear does not tend to judge sound powers in absolute terms but assesses how much greater one power is than the other [35,36].
Decibel (dB), the logarithmic scale of loudness, represents the ratio between two sound levels. The decibel scale ranges from 0 dB (threshold of hearing) to 130 dB (threshold of pain). A difference of 1 dB is the minimum perceptible change in volume, 3 dB is a moderate change, and 10 dB is a doubling of volume.
‘A’ and ‘C’ weighting scales refer to different sensitivity scales for noise measurement. They are used to mimic the perception of the human ear. The latter does not perceive very low (below 100 Hz) and high frequencies (above 10 000 Hz). The frequencies that people hear more easily are between 500 and 4000 Hz. The A-weighting scale follows the frequency sensitivity of the human ear at low levels. Sound level meters set to the A-weighting scale will filter out much of the low-frequency noise they measure, similar to the response of the human ear. Sound measurements made on the A-weighting scale are designated dBA. The C-weighting scale (dBC) follows the frequency sensitivity of the human ear at very high sound levels. Low frequencies are more easily filtered by the human ear than high frequencies. The A-weighting scale is more commonly used than the C-weighting scale because of the higher capacity of the human ear to filter low frequencies.
36 Gray L. Properties of sound. J Perinatol 2000; 20:S6–S11.