Office of Evidence Based Practice – Electromagnetic Interference (EMI) Specific · PDF file · 2013-09-12Office of Evidence Based Practice – Electromagnetic Interference (EMI)

Office of Evidence Based Practice – Electromagnetic Interference (EMI) Specific Care Question

If you have questions regarding this Specific Care Question – please contact [email protected] 1

Specific Care Question What is the effect of electromagnetic interference (EMI) on medical devices? Source of the Question: CJ Hutto, Senior Director Operations, Patient Care Services. Team Members: Evidence Based Scholars: Dusin, J., Gutierrez, C., Havlena, A., Menown, J., Radford, K., Thompson, L., & Tobin, T. Office of Evidence Based Practice: Allen, N., & Bartlett, J. Summary:

Based on moderate to high quality evidence a weak recommendation is made to allow the use of cellular phones within the hospital and clinics. The desirable effects of cell phone use are closely balanced with the undesirable effects EMI. Strong evidence from unbiased observational studies supports this recommendation. Evidence for no malfunction occurring was seen in 4 cohort studies and no malfunction occurring at distances ≥ 5cm in an additional 4 cohort studies support the recommendation,

The best action may differ, depending on circumstances or patients or societal values. Since the undesirable effects include the malfunction of medical equipment, the following caveats are made. From the included studies, the median distance for most inference is 10 cm (range: 0-125 cm) or 2 in. (range 0-50 in.). Cohen, et al. (2005) introduces idea of specifying a “sphere of risk” in specific locations medical equipment malfunction has the greatest impact on patient well being (i.e. critical care areas). In most instances the sphere would be ~ 10 cm (4 in.) around medical devices

Policy makers should be aware that EMI comes from various sources (tablet computers, alphanumeric pagers, radiofrequency tags and readers (RFID), walkie-talkies, computers on wheels, wireless monitors etc), not just cellular telephones. See Table 1. The specific effects of the devices included in this review can be found on Table 3.

The FDA regulates the shielding requirements of medical devices. The pre-market shielding requirements have been strengthened. However, the FDA recommends that when a medical device is received for service (or repair) and no problem is found, EMI should be investigated as a possible reason for the malfunction. Significance and importance of the question: Children’s Mercy Hospitals and Clinics is updating the policies related to use of devices that use radiofrequency (RF) wireless transmission. RF devices emit electromagnetic waves that may interfere with the function of medical devices. Current polices include:

o Provision of Wireless Communication Devices and Related Service Plans

o

- this policy includes alphanumeric pagers, Vocera devices, cellular telephones, wireless air cards, and personal digital assistants, such as Blackberry or Treo” Communication Equipment Use and Monitoring

o - this policy includes Vocera devices

Cellular Telephones (Wireless Devices), 2-Way-Radios, Pagers, and Personal Digital Assistants cellular telephones (wireless devices)

– This policy addresses cellular telephones but does not specifically address the other devices in the title of the policy.

Assuring the safety of patients cared for at our hospitals and clinics is the primary goal. This review serves to summarize the available research on this topic.

mailto:[email protected]�



Other wireless technologies that are likely to be in use within the CMH healthcare system are: Table 1. Sources Electromagnetic Interference

Adapted from: U. S. Department of Health and Human Services, Food and Drug Administration (FDA). (2007). Radio-frequency wireless technology in medical devices: Draft guidance. (Document number 1618). Retrieved from http://www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm077272.pdf

In the hospitals and clinics In patients’ homes Wireless operating room controllers Wireless monitors Wireless PDAs High frequency surgical devices Diathermy Wireless local area networks (WLAN) Wireless monitors Cellular phones Radio-frequency identification devices

(RFID) High RF power vehicle and portable

transmitter radios Radars RF toll systems (e.g., EZ Pass)

Cellular (mobile) phones Wireless PDAs Appliances Electronic products Two-way radios Amateur radio

Search Strategy and Results: ("Cellular Phone"[Mesh] OR "Wireless Technology"[Mesh] OR "Computers, Handheld"[Mesh] OR iPad[All Fields]) AND (interference[All Fields] OR "Equipment and Supplies"[Mesh] OR "Equipment Failure"[Mesh] OR "Equipment Design"[Mesh] OR "Equipment Safety"[Mesh]) AND ("2002/06/18"[PDat] : "2012/06/14"[PDat] AND English[lang]) "electromagnetic interference"[TIAB] AND ("Cellular Phone"[Mesh] OR "Wireless Technology"[Mesh] OR "Computers, Handheld"[Mesh] OR iPad[All Fields]) AND ("2002/06/19"[PDat] : "2012/06/15"[PDat] AND English[lang]) Results of the Search of PubMed can be found at: http://www.ncbi.nlm.nih.gov/sites/myncbi/collections/public/1tsoMk-Ksup4z5cfq972wk_ku/ (131 articles). Jason Newland, MD (Director, Evidence Based Practice) selected 34 articles to be closely read. 5 of these articles are included in the Carranza 2011 systematic review and are not included as separate articles in this review. Eleven articles were excluded (see Table 2.) 18 articles are included. The following web sites were reviewed:


http://www.ncbi.nlm.nih.gov/sites/myncbi/collections/public/1tsoMk-Ksup4z5cfq972wk_ku/�



https://www.ecri.org/Pages/default.aspx http://www.sciencedaily.com/releases/2007/03/070308220442.htm http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/TipsandArticlesonDeviceSafety/ucm225359.htm http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm077210.htm#4 Method Used for Appraisal and Synthesis: Studies were appraised by two reviewers, using the Critical Appraisal Skills Programme (CASP) tools for Cohort Studies. (Public Health Service, 2004) and the Review Manager 5.1 (RevMan) Results:

1. Cellular technology changes rapidly. Cellular phones used by employees or families can be of any of the supported technologies. Although 1G telephones are no longer in use, 2G phones are still in circulation, but support of the technology is waning. 3G phones (smart phones) are most available, and 4G phones are emerging. Most research has been done on 2G and 3G phones, so as more people use 4G phones, these results may not be applicable.

2. Devices other than cellular phones emit EMI (See Table 1.). 3. There is great heterogeneity across the included studies. See Table 3. Specific devices included in each study are found in Table 4. Heterogeneity is also

found in the outcomes assessed. The seriousness of the interference ranged from medical device screens going blank to infusion pumps stopping without alarm. Neither the duration of exposure to the RF device nor the duration of the effect of the interference was discussed.

4. In studies performed with human subjects, sample sizes were small. 5. Findings:

a. By source of potential interference: i. Ten studies evaluated either 2G or 3G phones and disruption of various medical devices. The distance where EMI did not occur ranged

from 0-125 cm (0-50 in.) away from various medical devices. Median distance was 30 cm (12 in.) ii. One study evaluated an in hospital cordless alpha numeric pager and ECG recordings. No interference was found when the distance was

0 cm iii. One study evaluated a wireless local area network (WLAN) and multiple medical devices. The distance where EMI did not occur was >5

cm (2 in.) iv. One study evaluated the iPad and VP shunts. The distance where EMI did not occur was > 1 cm. v. One study evaluated iPod and generic MP3 players against defibrillators and ECG respectively. The distance where EMI did not occur

was > 5 cm (2 in.) and >15 cm (6 in.) b. By medical device:

i. Three studies evaluated the performance of ECG recorders against RF emitting devices. The distance where EMI did not occur ranged from 0-125 cm. (0-50 in.)

ii. Five studies evaluated the performance of infusion pumps (including syringe and enteral pumps) against RF emitting devices. The distance where EMI did not occur ranged from 0-80 cm (0-32 in.). Median distance was 5 cm (2 in.)

iii. Four studies evaluated respiratory equipment (ventilators and CPAP/BiPAP). The distance where EMI did not occur ranged from 0-100 cm (0-40 in.) The median distance was 18 cm (7 in.)

iv. Five studies evaluated the performance of internal and external cardio-defibrillators (ICDs and ECDs) or pacemakers against RF emitting devices. The distance where EMI did not occur ranged from 0 cm to 125 cm (0-50 in.). The median distance was 5 cm (2 in).

v. Five studies evaluated the performance bedside monitors against RF emitting devices. the distance where EMI did not occur ranged from 2- 125 cm (1-50 in.) The median distance was 30 cm (12 in.)

vi. One study evaluated the performance of VP shunts against the iPad tablet. The distance where EMI did not occur was 0 cm.


https://www.ecri.org/Pages/default.aspx�

http://www.sciencedaily.com/releases/2007/03/070308220442.htm�

http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/TipsandArticlesonDeviceSafety/ucm225359.htm�

http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm077210.htm#4�



Included Single Study Characteristics and Risk of Bias Tables:

Strahle 2012 Characteristic of included study:

Methods Cohort Participants Ten magnetically programmable shunt valves were tested (Strata Valve, Medtronic, Inc.) Interventions • Measured magnetic field strength (magnetic flux density) near 32-GB iPad 2 devices

• Magnetic field strength near the tablet was recorded at distances between 0 mm (contact of the device to the magnetometer 0 mm and 100 mm.

• Magnetic fields were recorded for the tablet with and without the cover in place. • Two valves were set to 5 different performance levels (0.5, 1.0, 1.5, 2.0, and 2.5). • Valves were exposed to the tablet device at distances of less than 1 cm, 1–2.5 cm, 2.5–5 cm, 5–10 cm, and greater than 10 cm. Each

exposure lasted 10 seconds. For each distance tested, the valves were exposed 100 times to a tablet with a cover, resulting in 500 total valve exposures. Following exposure, the valve setting was investigated and performance level was recorded.

• The tablet alone, without a cover, was also tested at distances less than 1 cm for 30 valve exposures. Outcomes To determine the effect of tablet computer on magnetically programmable shunt valves at different distances.

Notes

Risk of bias table Bias Scholar’s judgment Support for judgment

Random sequence generation (selection bias)

High risk Unable to Randomize d/t Cohort study.

Allocation concealment (selection bias) High risk Unable to conceal d/t Cohort study. Blinding of participants and personnel (performance bias)

Unclear risk No participants were used. Unable to blind personnel.

Blinding of outcome assessment (detection bias)

High risk Not discussed and unlikely d/t Cohort study.

Incomplete outcome data (attrition bias) Low risk Unlikely incomplete data was not reported. Although, not discussed. Selective reporting (reporting bias) Low risk Unlikely study used selective reporting. Although, not discussed. Other bias High risk In order to complete a comparison of data this reviewer is making the assumption that "no

exposure to the tablet computer" would result in no change in valve settings.




Forrest Plots os Single Study

Distance of tablet computer to magnetically programmable shunt (0 and 1 cm.) vs. no exposure to tablet computer. Outcome- Altered valve setting

Distance of tablet computer magnetically to programmable shunt (> 1cm to < 2.5 cm) vs. no exposure to tablet computer. Outcome- Altered valve setting.

Distance of tablet computer to magnetically programmable shunt (> 2.5cm to < 5 cm) vs. no exposure to tablet computer. Outcome- Altered valve setting.

Study or SubgroupStrahle 2012

Total (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Z = 3.93 (P < 0.0001)

Events58

58

Total100

100

Events0

0

Total100

100

Weight100.0%

100.0%

M-H, Fixed, 95% CI276.67 [16.71, 4580.03]

276.67 [16.71, 4580.03]

Exposure 0 and 1cm Control Odds Ratio Odds RatioM-H, Fixed, 95% CI

0.001 0.1 1 10 1000Favours Exposure at 0 & 1 Favours control


Total (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Z = 1.65 (P = 0.10)

Events5

5

Total100

100

Events0

0

Total100

100

Weight100.0%

100.0%

M-H, Fixed, 95% CI11.58 [0.63, 212.19]

11.58 [0.63, 212.19]

Tablet Exposure No Tablet Exposure Odds Ratio Odds RatioM-H, Fixed, 95% CI

0.01 0.1 1 10 100Favours tablet exposure Favours control


Total (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Z = 0.68 (P = 0.50)

Events1

1

Total100

100

Events0

0

Total100

100

Weight100.0%

100.0%

M-H, Fixed, 95% CI3.03 [0.12, 75.28]

3.03 [0.12, 75.28]






Distance of tablet computer to magnetically programmable shunt (>5 to < 10 cm) vs. no exposure. Outcome- Altered valve setting.

Distance of tablet computer to magnetically programmable shunt (>10 cm) vs., no exposure. Outcome – Altered valve setting.


Total (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Not applicable

Events0

0

Total100

100

Events0

0

Total100

100

Weight M-H, Fixed, 95% CINot estimable

Not estimable




Total (95% CI)Total eventsHeterogeneity: Not applicableTest for overall effect: Not applicable

Events0

0

Total100

100

Events0

0

Total100

100

Weight M-H, Fixed, 95% CINot estimable

Not estimable






Synthesis of relevant studies:

Author, date, country, and industry of

funding

Devices (See Table 4

for list of devices per

study)

Level of Evidence (Oxford)

Research design Significant results Limitations

Baranchuk 2009

Canada

Communication devices tested:

3 cell phones 3G (CDMA)

1 in hospital phone-cordless

1 alpha-numeric pager

Three ECG Instruments

1b Validating cohort study Devices were tested

on 3 different ECG instruments at 4 distances -(2 m, 1 m, 0.5 m and 0.25 m. and 0 cm in both the active and deactivated mode

No interference was detected when any of the devices were at 2 m, 1 m, 0.50 m, or 0.25 m in either the active or deactivated mode.

Are the ECG instruments similar to those used in our hospital?

Reporting bias- They report EMI when a phone is placed on the ECG instrument- it is not a study question at the outset of the study

Reporting bias- also occurred when they stated the differences in ECG interpretation among different levels of practitioners (RN, med student, cardiologist). This was not question at the outset, and there could be other reasons for misinterpretation than EMI.

Calcagnini 2004

Italy

Three mobile phones 2G (GSM) were tested against seven infusion pumps and four syringe pumps

1b Validating cohort study

Cohort pump study Outcomes: Interference

Five out of seven infusion pumps and 1 out of 4 syringe pumps were affected by the GSM phones either at 900 MHz or 1800 MHz.

The distance varied, did not get better or worse with various MHz or distances.

Emitted power (W) of each phone has an effect on EMI

Suggest reducing the emitted power (W) will reduce the risk of EMI significantly.

GSM phone are designed to reduce W to battery saving if adequate signal is present

Install in building amplifiers Install hospital base-stations

It is an old study, cell phone technology has changed since 2004.

Calcagnini 2007

Italy

3a Systematic Review of heterogeneous cohort

Systematic Review The SR includes 6 studies of GSM mobile phones and infusion pumps. The percentage of interference reported is greater in the 1997 included study than in the 2006 included

Search strategy is not specific. Method to select include articles is not clear.

Did not rate the quality of the




studies study, suggesting as cell phone and blocking technology becomes more sophisticated, interference becomes less likely.

Suggests the probability of EMI would be reduced if the field coverage was increased by installing in-building repeaters amplifier or dedicated mini base amplifiers

included studies. Did not group the studies in any way, due to heterogeneity of the included studies.

Calcagnini 2008

Italy

Communication devices tested

3 mobile phones 2 G (GSM)

Seventeen infusions pumpsd


Cohort pump study Outcomes: detect EMI

related abnormal functioning

Minimum distance at which malfunction occurred.

Both Digital Enhanced Cordless Technology (DECT) and WiFi was tested

There were 8 syringe pumps, 7 volumetric pumps, and 2 nutrition pumps

Five out of 17 pumps were affected by the GSM phones either at 900 or 1800 mHz. 3/5 of the malfunctions occurred with pump stoppage and alarm sounding. These EMI devices were within 30 cm of the infusion pump.

At distances > 30 cm neither WiFi nor cellular phones caused pump malfunction.

Uncertain if technology of today’s mobile devices and medical devices are comparable to the devices in this report.

Calcagnini 2011

Italy

Four types of wireless local area networks (WLAN) and

45 medical devices


45 medical devices were tested by ANSI C63.18.1997* guideline Interferences were further defined as either a basic safety risk (may cause and unacceptable) risk or a performance risk (compromise the use for which the device was purchased)

*ANSI-C63.18.1997 is

a technical guide to aid clinical and biomedical engineers in assessing the immunity of medical devices to RF fields from portable RF

Each medical device was tested again four combinations of WiFi adapters/antennas for a total of 180 tests.

No malfunction was seen in 42 of 45 devices, even when the devices were in contact with the transmitters/antennas.

Three medical devices malfunctioned 1. Enteral pump- stopped when the

LogiLink adapter with a 6 cm antenna was operated in contact with the device, and for one device up to 2 cm away from the pump. The pump was returned to normal after being turned off and back on. This may impact the basic safety of the pump. The other WiFi adapters did not affect a change in the pump.

2. External defibrillator- malfunctioned when the LogiLink, with 6 cm dipole antenna. The synchronized cardioversion function was affected. It added trace noise that provoked the defibrillator to lose synchronization with the QRS complexes and deliver

Products are continually being upgraded. This data may not reflect products that are available in today’s market.




transmitters. the shock with random timing. The malfunction was reversed when the WiFi adapter was at a farther distance from the defibrillator. The malfunction may impact the basic safety of the defibrillator

3. Monitor- emitted sounds when the three of the four adapters were within 5 cm. of the medical device. The malfunction impacted the performance, but not the safety.

Carranza 2011 Spain

47 studies on EMI

Literature search from 1999-2009

3a Systematic Review of heterogeneous cohort studies

Systematic Review Inclusive search.

Defined inclusion criteria, 2 reviewers, using a checklist.

Selected 47 studies out of 820 potentially relevant studies

Great heterogeneity among the studies. Five of the selected studies looked at medical

devices interfering with the LAN. The audio component of the wireless devices was most disrupted.

Three included studies measured the electromagnetic fields in healthcare environment. All articles conclude that hospitals should take protective measures against incoming radio waves.

The great heterogeneity among studies of interferences on medical instruments with wireless networks makes global conclusions difficult. Interference between GSM technology and pacemakers is the most prolific. Interference varies between 1.5% and 5%. And interferences decrease as distance between the devices is increased. In terms of defibrillators, no evidence of serious interferences in the function of the implanted defibrillator has been observed. For other papers on this topic with other medical instruments, the seriousness of the interference in not routinely reported. For example, when cell phones were tested against apnea monitors, 3 of 4 monitors had serious interference. When interference was tested against ventilators, the interference was minor like bad screen readings, to serious interference such as increasing the frequency or stopping the machine. Real damage to patients is rarely

No restriction on study design because studies dealing with patient security and possible harmful effects of EMI on medical devices are scarce.

Great heterogeneity among the studies.

Studies do not answer the questions asked by the SR

Sample sizes are small




reported.

Dang 2007 Canada

Tested 2 mobile phones 3 G (TDMA- available in Asia) in three modes (idle, conversation searching) each and one two way radio against 3 adult ventilators (one portable) and 3 pediatric invasive and non-invasive ventilator and one CPAP machine


Each piece of medical equipment was in the on position at previous determined settings, and attached to a plastic lung. Each mobile device was tested at distances starting at 0 m and subsequently at distances that increased by 0.5 m. Mobile devices were tested in the three modes. They were tested for 5 minutes. Mobile devices were moved in 360˚ around the ventilator.

Bottom line: Cellular phones and 2-way radios are safe and cause no interference unless operated at very close distances of less than 1 meter.

The 2-way radio (MRK Ericsson GE 810-815

mHz) caused significant interference on almost all of the devices:

Puritan-Bennet 7200 (adult) at 1.0m Pulmonetics LTV 1000 (adult) at 0m Draeger Babylog 8000 (ped) at 0m Bird VIP GOLD (Ped) at 0m Respironics BiPAP Synchrony at 0m The Motorola v300 2G (GSM) cellular phone in

conversation mode caused significant interference on the Puritan-Bennett 7200 (adult) at 0m

The study is limited to ventilators. The researchers assumed five minutes is a sufficient time period for EMI to cause changes in a ventilator device.

Guertin 2007 USA

Tested implantable cardioverter-defibrillator devices (ICD) against electrocautery devices used in gastrointestinal endoscopy (EGD)


41 patients with ICD who were undergoing EGD with potential for cautery were followed. Post procedure, the ICD was analyzed for detection of EMI or tachy-arrhythmias

Thirteen patients had single chamber devices, 25 had dual changer devices and three had biventricular devices o Guidant pulse generators and leans (St.

Paul, MN) o Medtronic pulse generators and leads

(Minneapolis, MN)

There were no cases of esophageal interventions

Small sample size, imprecise findings.

Hans 2008 2G GSM and 1b Validating Mobile phones were Syringe pump- was studied for a 13.5 hour




India 3G CMDA mobile phones and syringe pump, mechanical ventilator, and bed side monitor

cohort study placed at a distance of 1 foot from the medical devise in off (30 min), standby (30 min) and talking (10 min) modes

GSM was also tested linked to a Bluetooth device.

Medical devices were observed for EMI

All studies were repeated for a total of 3 replications.

period. No interference was seen during the off or standby mode. During the talk mode and the syringe pump set at low rate (40 ml/hr) volume delivered were similar. When the syringe pump was set at the fast rate (200 ml/hr) the variation between calculated and actual delivered volume:

• Off mode- mean variation =0.45ml, maximum variation =1.54 ml

• Standby mode- mean variation= 0.2 ml, maximum variation =0.7 ml.

• Talk mode - mean variation = 2.66 ml range 0.7-6.5 ml. Calls lasted form < 2 minutes to <15 minutes,

Mechanical ventilator- was studied for 7.5 hours (off mode 3 hours, standby mode 3 hours and talk mode for 1.5 hours) No effect was seen on set respiratory rate of FIO2. The following variations were observed in tidal volume

• Off mode- mean variation =38.33 ml, maximum variation =56ml

• Standby mode- mean variation 32.39 ml maximum variation38 ml

• Talk mode- mean variation =44.22 ml maximum variation = 78 ml

Bedside monitor- was studied for 15 hours (off mode 6 hours, standby mode 6 hours and talk mode for 3 hours. No unexplained malfunctions or shut downs of the monitor occurred. Conclude mobile phones should be at least one foot away from the diameter of the syringe pump.

Helhel 2011 Turkey

2G (GSM900, GSM 1800) and 3G (CDMA) mobile devices and medical


Mobile devices were placed at pre-specified distances from the medical equipment in the units where the equipment was

The following medical devices were affected by mobile devices

Mobile device

Hospital equipment

Interfering distance

2G/3/G Cardiofax 1.25m/1m

Imprecise in the description of the mobile devices and hospital equipment tested.

Specific results of what incidences occurred when interference was detected were not given.




equipment in various hospital departments

used. (Effort-1 Unit)

2G/3G ECG 0.5m/0.4m

2G/3G Cardiofax (Surgical Unit)

1m/0.35m

2G Neurofax (EEG)

1.2m

No information was provided on the observers for each test, or if the same observers were used for all tests.(potential for detection bias)

Length of time medical equipment was exposed to mobile device is not specified.

Houliston, 2009

New Zealand

The affect the Tracient Padl-R passive UHF RFID reader or the SkyeTek M9 UHF reader has on the Graseby 3500 infusion pump

4 Factorial Experimental

Design

Tracient Padl-R passive UHF RFID reader placed directly on the pump to 10 cm, 20 cm, 30 cm, 50 cm, and 1 m away from the pump with no interference noted. One Tracient reader placed directly on the pump and 10 cm away from the pump (the pump was tested with and without a RFID tag) with no interference noted. Four Tracient readers placed side-by-side directly on a pump (the pump was tested with and without a RFID tag) with no interference noted. SkyeTek M9 UHF reader placed 10 cm away from the pump (the pump was tested with and without a RFID tag) interference was noted with the RFID tag. Interference was also noted when one SkyeTek M9 UHF reader and four Tracient readers were placed side by side 10 cm away from the pump without an RFID tag used and interference was noted. Interference = the pump stopped working, sounded an alarm, and displayed the message ‘FAULT CODE 10’. The pump could not be reset, and had to be switched off and back on. Each pump failure, in both cased, could not be replicated. Although, once the pump had failed the first time, the same failure occurred three times between tests, when no RFID readers were active.

Experimental Design though rigorous could have been improved through the use of a cohort study in which both types of technology were trialed using the same increments.

Uncertain if CMHC uses this technology (pump & RFID tags, readers).




Ismail, 2010 Germany

Communication devices tested:

3rd generation mobile phones

(Nokia 6650, Motorola

A835) with a peak power of 0.25 W

2b Individual cohort study

Devices were tested on 100 patients with permanent pacemakers in standby, dialing and operating mode with the phone positioned directly above the pacemaker pocket.

No interference was detected with any of the 31 pacemakers tested in the three modes.

As the pacemakers tested the same models used in the CMHC population?

4 G phones were not studied. Phones that are unable to modify the power output > 0.26 W were not tested.

Kruk, 2003 Australia

Guideline from New South Wales

4 Guideline, not stated how it was developed

States GSM 1800 and CDMA phones may operate as close as 0.5 meter (20 inches) from electronic equipment and GSM 900 phones restricted to 2 meters (6.5 ft) distance.

Strahle 2012 USA

Two 32GB iPad tablet computers against 10 magnetic programmable shunt valves (Strata Valve, Medtronics, Inc)

Level 2b Individual RCT with wide confidence intervals.

First they measured the iPads for the area of highest magnetic field strength, which was the upper left corner.

Shunt valves were set to five different performance levels. (0.5,1.0.1.5.2.0 and 2.5)

The valves were positioned on the testing surface and exposed to the tablet device the following distances: less than 1 cm, 1-2.5 cm, 2.5-5 cm, 5-10 cm and greater than 10 cm

Each exposure lasted 10 seconds

See RevMan tables and graphs. Interference occurred when the iPad was < 5 cm from the valve.




Valves were exposed to a tablet with the cover closed.100 times for each distance n= 500 exposures.

The tablet without the cover was tested 30 times at the distance of < 1 cm.

The tablet was “on” and in the wireless internet mode.

Thaker, 2008 USA

iPods and cardiac

pacemakers 100 patients for

800 tests. Telemetry was

used to monitor pacemaker


iPods were placed 2” anterior to the pacemaker. The iPod had headphones attached, turned up to maximum volume and the equalizer was turned off. The iPod was turned on, played, fast-forwarded and turned off. Tested for 10 seconds

Sixty-three Type 1 interference events [7.88%] were seen in 19 patients and 98 Type II interference events (11.75%) were seen in 32 patients.

iPods were only tested for 10 seconds

Only iPods were tested No range of variation is presented

Results are reliable, but not useful. The authors conclude that if someone with a pacemaker holds an IPod within 2 inches of their pacemaker, it can cause interference.

Tri 2005 USA

6 cellular phone technologies

(CDMA, 2 models, GSM, iDEN TMDA, and analog)

16 medical devices


In an area with week incoming cellular tower signal, medical devices were tested in normal operating mode. A 22- step procedure was followed; increasing the distance the cellular device was placed from the medical device.

The following was measured

EMI during ringing,

Interference occurred in 7/16 (44%) of medical devices tested.

Interference occurred in 108/510 (21.2%) of the specific tests performed.

Clinically important interference (defined as interference that may hinder interpretation of data or cause equipment to malfunction) was 1.2%

Most interference occurred with devices that displayed ECG waveform.

The Respironics Esprit 2581 ventilator and the TBird Legacy 15812 portable ventilator were affected.

Type of interference on ECG and EEG devices was noise on waveform.

Did not test all devices Only had mobile devices in conversation transmitting functions, not data or pictures.




rotation of the phone, distance,

Interference was defined as: noise on ECG waveform, baseline movement, equipment reset, display of wrong information,

The farthest distance interference occurred was 32 inches.

The older modules on the ECG/Resp devices were more susceptible to interference. There was no interference with the newer modules.

The ventilators showed interference when the analog phone or the TDMA (2G) phone was held within 2 inches from the top of the ventilator in ringing mode.

Tri 2007 USA

2 cellular devices from two different cellular 2G and 3G (GMS and CDMA) carriers versus medical equipment in 11 patient care areas


Followed an 8 step procedure involving carrying the mobile device in 10 separate patient care areas. The investigator walks around the patient room with the cellular device both taking a call and making a call.

75 patient care rooms, in 11 patient care areas (medical cardiology intensive care unit, medical cardiology unit, echocardiography laboratory, neuroepilepsy monitoring unit, transplant critical care unit, cardiovascular surgery step-down unit, cardiovascular surgery intensive care unit, neurosurgery intensive care unit, medical intensive care unit, vascular surgery postoperative care unit and pulmonary ventilator rehabilitation unit)

300 tests were performed (75 each for each of the 2 telephones with a call connected and 75 for each of the 2 telephones without a call connected, the ringing test.)

Interference occurred in 0 of 300 tests..

Tests were not performed in the surgical suites or the pediatric intensive care unit.

The study did not test pediatric environments, but performance of equipment should be similar.

Trigano 2006 France

2 mobile phone devices 2G (GSM) versus automated external defibrillators


Mobile devices were place in three positions, on the AED, on the patient’s chest between the electrodes and otn the connector between the electrodes and the AED cable. Calls were place to the phone after they were place during the AED analyzing phase.

Patients with the implanted pacemaker had no interference when the phone rang,

There was no signal distortion or artifacts on real time ECG waveform in any test.

Cellular phone ringing did not interfere with cardiac rhythm analysis or with the emission of markers for pacemaker spikes,

A disturbance, noise emission from the AED speaker, occurred with 13 of the 26 tests when the phone receiver was placed over the AED, The minimum distance where this could be prevented was 15 cm.




Devices were tested with their original electrodes and Quik-Combo electrodes for two AEDs and Adult Pluss multifunction electrodes in the other.

Webster 2008 USA

4 digital music players (DMP) versus pacemakers, implantable cardioverter defibrillators (ICDs)


Tested children (> 4 years) and young adults who had an active pacemaker.ICD.

DPMs were tested in in the on mode, with music playing and headphones plugged into the headphone jack.

The device programmer was placed over the patient’s pacemaker or ICD. ECG was monitored for disturbances. Each DMP was tested on each subject in random order, If interference was noted the DMP was moved 6 inches away and re tested.

Definition of interference: any aberration that caused a change in the cardiac rhythm as seen on the surface ECG, or any clinical symptom- discomfort, pain,

51 subjects (34 with a pacemaker and 17 with an ICD) who participated in five tests each. (baseline plus each DMP).

Three types of events were recorded o Effect on the pacemaker or ICD o Effect on the device programmer o Effect on the communication between the

pacemaker/ICD and the device programmer.

There was no change in the surface ECG during any of the baseline or DMP tests (0/255)

There was no oversensing, undersensing, mode reversion, asynchronous pacing or abnormal functioning.

There was no inappropriate ventricular tachycardia sensing events, nor inappropriate shocks

Interaction between the DMP and routine function of the pacemaker or ICD was not observed.

There was high incidence of interference with telemetry programmer communication when the DMPs were placed directly against the subject. (29 / 204 total DMP tests or 51/ 51 patients)

Larger DMPs (30GB) showed more interference compared to the ≤4GB DMPs)

Only one subject complained of symptoms. The subject complained to tingling and burning at the site of the pacemaker. Recordings did not reflect interference and the symptom remained when the pacemaker was taken away.

Small number of subjects Small number of DMPs Only looked for acute interferences

Two sources of interference present in the study: 1. The implanted device and 2. The device programmer head.




dizziness, pre-syncope, syncope, headache or palpitations the correlated with an ECG abnormality.

Wh n the DMPs ere move six inches from the subject, the interference was not detectable.

Updated September 18 2012

Table 2 Studies excluded from the review

Study Reason

Aliyev 2010 Case report

Censi 2007 Narrative review

Censi 2010 Narrative review

Ettelt 2006 Narrative review

Hahn 2005 Letter to the editor

Pearce 2009 Abstract

Phunchongharn 2010 Narrative review

Ramesh, 2008 Does not address the question asked

Rogan 2005 Letter to the editor

Ruskin 2006 Narrative review

Simon 2009 Letter References

Baranchuk, A., Kang, J., Shaw, C., Campbell, D., Ribas, S., Hopman, W., Alanazi, H., Redfearn, D., & Simpson, C. (2009). Electromagnetic interference of communication devices on ECG machines. Clin Cardiol. 32, 588-592.

Calcaginini, G., Bartolini, P., Floris, M., Triventi, M., Cianfanelli, P., Scavino, G., Proietti, L., & Barbaro, V. (2004). Electromagnetic interference to infusion

pumps from GSM mobile phones. IEEE. 3515-3518.




Calcagnini, G., Censi, F., Triventi, M., Mattei, E., & Bartolini, (2007). Electromagnetic immunity of infusion pumps to GSM mobile phones: A systematic review. Ann Ist Super Sanita, 43, 3, 225-228.

Calcagnini, G., Censi, F., Triventi, M., Mattei, E., LoSterzo, R., Marchetta, E., & Bartolini, P. (2008, August). Electromagnetic interference to infusion pumps.

Update2008 from GSM mobile phones. The 30th Annual International IEEE EMBS Conference, Vancouver, British Columbia, Canada. Calcagnini G, Mattei E, Censi F, Triventi M, LoSterzo R, Marchetta E, & Bartolini P. (2011). Electromagnetic compatibility of WLAN adapters with life-

supporting medical devices. Health Physics, 100. 5, 497-501.

Carranza,N., Febles, V., Hernandex, J. A., Bardasano, J. L., Manteagudo, J. L., Fernandez de Aldecoa, J. C., & Ramos, V., (2011). Patient safety and electromagnetic protection: A review. Health Phys, 100, 5, 530-541. doi: 10.1097/HP.0b013e3181f0cad5

Censi, F., Calcagnini, G., Triventi, M., Mattei, E., & Bartolini, P. (2007). Interference between mobile phones and pacemakers: A look inside. Ann Ist Super Sanità, 43, 254-259.

Cohen, T., Ellis, W. S., Morrissey, J. J., Bakuzonis, C., David, Y., & Paperman, W. D. (2005). Safe use of cellular telephones in hospitals: fundamental principles and case studies. Journal of Healthcare Information Management 19, 4, 38–48.

Dang, B. P., Nel, P. R., & Gjevre, J. A. (2007). Mobile communication devices causing interference in invasive and noninvasive ventilators. Journal of Critical Care, 22, 137-141. doi: 10.1016/j.jcrc.2006.09.00

ECRI (Emergency Care Research Institute). (2009, January) Healthcare risk control: Cell phones and electromagnetic interference. (Executive Summary, Medical Technology 12) Retrieved from https://www.ecri.org/Documents/Patient-Safety/HRC_TOC/MedTech12ES.pdf

Guertin, D., Faheem, O., Ling, T., Pelletier, G., Mcconmas, D., Yarlagadda, R. K., Clyne, C., & Kluger, J. (2007). Electromagnetic interference (EMI) and arrhythmic events in ICD patients undergoing gastrointestinal procedures. PACE, 30, 734-739.

Hans, N., & Kapadia, F. (2008). Effects of mobile phone use on specific intensive care unit devices. Indian J Crit Care Med, 12, 4, 170-173.

Helhel, S., Colak, Z., & Ozen, S. (2011). Distance and location of both mobile phones and health care units: Determines the interference level. American Journal of Biomedical Engineering, 1, 2, 78-82. doi: 10.5923/j.ajbe.20110102.14

Houliston, B., Parry, D., Webster, C. S., & Merry, A. F. (2009). Interference with the operation of medical devices resulting from the use of radio frequency identification technology. The New Zealand Medical Journal, 122, 1297.

Ismail, M. M., Badreldin, A.,M. A., Heldwein, M., & Hekmat, K. (2010). Third-generation mobile phones (UMTS) do not interfere with permanent implanted pacemakers. Pacing Clin Electrophysiol, 33, 7, 860-864. doi: 10.1111/j.1540-8159.2010.02707.x

Kruk, R. (2003), Mobile phones and wireless communication devices – interference with medical equipment – use of. New South Wales Health Department Guideline, Retrieved from: http://www.health.nsw.gov.au/policies/GL/2005/pdf/GL2005_045.pdf

Mayo Clinic (2007, March 11). Hospital Equipment Unaffected By Cell Phone Use. ScienceDaily. Retrieved September 14, 2012




Public Health Service, Critical Appraisal Skills Programme (CASP). (2004) Cohort 12 Questions. Retrieved from http://www.sph.nhs.uk/search?SearchableText=CASP

Ramesh, J., Carter, A., Campbell, M., Gibbons, N., Powlett, C., Moseley, H., . . . Carter, T. (2008). Use of mobile phones by medical staff at Queen Elizabeth Hospital, Barbados: Evidence for both benefit and harm. Journal of Hospital Infection, 40, 160-165. doi:10.1016/j.jhin.2008.06.007

Review Manager (RevMan) [Computer program]. Version 5.1. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011. Strahle, J., Selzer, B.J., Muraszko, K.B.,Garton, H.J.L., Maher, C.O. (2012). Programmable shunt valve affected by exposure to a tablet computer. J

Neurosurg 10,118-120.

Thaker, J. P., Patel, M. B., Jongnarangsin, K., Liepa, V. V., & Thakur, R. K. (2008). Electromagnetic interference with pacemakers caused by portable media players. Heart Rhythm, 5, 538-544.

Tri, J. L., Severson, R. P , Firl, A. R., Hayes, D. L. & Abenstein, J. P. (2005) Cellular telephone interference with medical equipment. Mayo Clin Proc, 80, 10,

1286-1290. Tri, J. L., Severson, R. P., Hyberger, L. K., & Hayes, D. L. (2007). Use of cellular telephones in the hospital environment. Mayo Clin Proc, 82, 3, 282-285. Trigano, A., Blandeau, O., Dale, C., Wong, M-F., 7 Wiart, J. (2006), Clinical testing of cellular phone ringing interference with automated external defibrillators.

Resuscitation, 71, 391-394. doi:10.1016/j.resuscitation.2006.04.013 U. S. Department of Health and Human Services, Food and Drug Administration (FDA). (2007). Radio-frequency wireless technology in medical devices: Draft

guidance. (Document No. 1618). Retrieved from: http://www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm077272.pdf

Webster, G., Jordao, L., Martuscello, M., Mahajan, T., Alexander, M. E., Cecchin, F., …Berul, C., I. (2008). Digital music players cause interference with

interrogation telemetry for pacemakers and implantable cardioverter-defibrillators without affecting device function. Heart Rhythm, 5, 545-550.




Table 3. Study Summary, by RF and Medical Device

Author, date,

country, and

industry of funding

RF device

Medical device- This is not inclusive of all devices. Many studies looked at many devices, not included here

Greatest distance where interference did not occur

Cellular phone GSM CMDA

In hospital cordless Alphanumeric pager

WLAN

iPad tablet

MP3 players

ECG Infusion pump (incl syringe pumps)

Resp equip Vents & Cpap/Bipap

Defibrillators ICDs ECDs and pacemakers

Bedside monitors

VP Shunts

Baranchuk 2009

Canada

x x x 0 cm

Calcagnini 2004

Italy

x x 0 cm

Calcagnini 2008

Italy

x x 30 cm

Calcagnini 2011

Italy

x x x x x 5 cm

Dang 2007

Canada

x x 1 m (40 in)

Hans 2008 India

x x x 30 cm (12 in)

Ismail 2010

Germany

x x 0 cm

Helhel 2011

Turkey

x x x x 1.25 m (50 in)

Strahle 2012

USA

x X 2 cm

Thaker, X x 5 cm (2 in)




2008 USA

(iPod)

Tri 2005 USA

x x x x 80 cm (32 in)

Tri 2007 USA

x x x x 0

Trigano 2006

France

x x 2 cm

Webster 2008

USA

x x 15 cm (6 in)

Total 10 1 1 1 2 3 5 4 5 5 1 Calcagnini

2007 Italy

Systematic review

States interference between 1.5-5% if the time. There was in inverse relationship between the distance from the medical device and the interferences. Major recommendation is to decrease power emitted by increasing coverage

Restrict use of mobile phones, 1 m, 1.2 m, 0 for PHS phones

Carranza 2011

Spain

Systematic review -EMI of GSM (2G) phones on infusion pumps.

Effect s: none, hazardous, % with possible real damage, interferences on the screen, cardiac rhythm changes and malfunction display

200 cm, 8.6 cm; 10 cm; 0 cm (PHS phones, Asia);

Guertin 2007

USA

GI procedure room electrocautery vs. defibrillators

Houliston 2009

New Zealand

RFID tags placed on medical devices evaluated interference when in proximity to two RFID

readers, one high power and one low power vs. infusion pump

The high power RFID reader interfered with the infusion pump when an RFID tag was on the infusion pump and the high power

reader was within 10 cm of the pump

Kruk, 2003 Australia

Guideline from New South Wales




Table 4. Type of Electromagnetic interference device and medical device in the included studies

Study EMI Device Medical Instrument Baranchuk 2009 GSM- (Motorola V220; 900 MHz, 1800 MHz,

1900 MHz) CDMA- (Sanyo SCP 2300; 800 MHz, 1900 MHz) Analog phone (Nokia 6275i; 800 MHz) In-hospital cordless phone- (Nortel WLAN

Handset 2211; 2 400–2483.5 MHz) Alpha-numeric pager -(Suntelecom ST800 Flex;

900 MHz).

ECG instruments MAC 5000 (General Electric, Chicago, IL, USA) MAC 1200 (General Electric, Chicago, IL, USA) ELI 100 (Mortara, Milwaukee, MN, USA)

Calcagnini 2004 Motorola V3688 Nokia 3510 Ericsson SH888

Infusion pumps (not specified) From the following manufacturers

Alairs Abbot Nutricia B|Braun

Calcagnini 2008 Nokia 6125- Type RM178 (max ear SAR 0.64 W/kg)

Nokia 6070-Type RM166 (max ear SAR 0.88) Siemens C72 (max SAR 0.70)

Infusion pumps (year of fabrication) Alaris - Asena PK-MK III (2005) Alaris - Asena CC-MK III (2001) Alaris - CC Guadrails (2007) Bbrown Perfusor Compact (-) Fresenius Pilot A2 (-) Fresenius Orchestra DPS (-) Alaris PK 2007 (2007) Alaris SE 7131 (2007) Alaris 7231 (2002) Alaris 7101 (1999) Abbott lifecare 5000 (2002) Bbraun infusomat FMS (-) MicroMacro XL (-) Orchestra Module MVP PT Tyco Kangaroo 624 2002 Nutricia Flocare 800 2000




Calcagnini 2011 Local area networks WLAN IEE 802.11 b/g 2.45 Ghz, 100 mW

Medical devices (not by brand name, but by type) Item Number of models

tested Syringe pumps 4 Volumetric pumps 8 Enteral pumps 2 Defibrillators 8 Monitors 11 Lung ventilators 5 Anesthesia machines 6 External pacemaker 1

Dang 2007 MRK Ericcson GE (radio) 810-815 mHz/ Samsung 680 (TDMA) idle mode Samsung 680 (TDMA) conversation mode Samsung 680 (TDMA) search mode Motorola v300 (GSM) idle mode Motorola v300 (GSM) conversation mode Motorola v300 (GSM) search mode

Puritan-Bennet 7200 (adult) at 1.0m Siemens Servo 300 Pulmonetics LTV 1000 (adult) at 0m Draeger Babylog 8000 (ped) at 0m Bird VIP GOLD (Ped) at 0m Respironics BiPAP Synchrony at 0m Siemens Servo 300 Pediatric CPAP Sullivan III

Guertin 2007 Unipolar electrocautery device: Endostat TM ‼ Bipolar/Monopolar Electrosurgical Generator (Boston Scientific Natik, MA, USA)

Placement of ICDs Transvenous left pectoral implants n=40 and one

left abdominal implant Hans 2008 GSM- (Motorola V3i, Nokia 6600, and Nokia

5310) CDMA- (LG 5130)

Syringe infusion pumps- B Braun Mechanical ventilator- Versa Med Bedside monitor- Philips-Intellivu MP40

Helhel 2011 GSM900 GMS1800 3G

Cardiofax (Effort-1 ECG Intensive Care Monitor Serum Equipment Cardiofax (Surgical Unit) Cardiofax (Surgical Unit) Delivery Unit Equipment Dialysis equipment Ultrasound equipment Non-Stres Test Equipment X Ray Equipment Neurofax (EEG) Injector Equipment Emergency Baby Care Unit Monitor Causally Department Monitor




Defibrillator EMG Equipment

Tri 2005 Cellular phone technologies Code Division Multiple Access (CDMA) Global System of Mobile communications (GSM) Integrated Digital Enhanced Network Time Division Multiple Access Analog

Philips Viridia 24C vital sign monitor o With Rev 1001A ECG/Resp module

(older module) o With Rev 1002B ECG/Resp module

(newer module Hewlett-Packard (Merlin) component monitoring system o With Rev 1001A ECG/Resp module

(older module o With Rev 1002B ECG/Resp module

(newer module)⁄⁄ None NA Xltek EEG desktop system With Mobee amplifier With Mobee amplifier and patient connected Philips IntelliVue MP 70 monitor Propaq 104 portable patient monitor Marquette/GE ECG cart Nellcor N-595 Pulse Oximeter Zoll M series defibrillator Baxter Colleague Volumetric Infusion Pump Datascope System 97 Intra-Aortic Balloon Pump Siemens ventilator Nellcor Puritan Bennett 840 Ventilator System Respironics Esprit 2581 ventilator TBird Legacy 15812 portable ventilator B Datex-Ohmeda Aestiva anesthesia system Philips 2600 telemetry pack b`

Tri 2007 Nokia 3587i CDMA Nokia 3120 GSM

Xltek EEG system with Mobee Amp Philips VS1 vital signs monitor Respironics CPAP machine Philips IntelliVue MP30 and MP70 patient monitors Baxter COLLEAGUE Volumetric Infusion Pump Siemens ACUSON Sequoia ultrasound system GE Vivid 7 cardiovascular ultrasound system Medtronic 5388 external pacemaker Puritan Bennett 7200 Ventilatory System Nellcor Ross Patrol enteral feeding pump Hospira, Bard CritiCore System urine output monitor




LifeCare PCA Plus 3 Infusion Pump Hospira, Philips Model M4841A Telemetry Pack Philips Viridia 1176 Patient Monitor Baxter blanket heater and water pump Abbott LifeCare PCA3 Infusion System Puritan Bennett 840 Ventilator System Philips IntelliVue with intracranial pressure monitoring capability Aircast VenaFlow System Nellcor OxiMax N-595 pulse oximeter GE DINAMAP PRO 100 noninvasive blood pressure monitor Datascope CS100 with IntelliSync counter pulsation balloon pump Total No. of medical devices 192 24

Trigano 2006 GSM receiver PCS receiver (Personal Communication Services)

LifePack 20 monitor/defibrillator LifePak 20P monitor/defibrillator/stimnulator HeartStart XL M4735A monitor/defibrillator

Webster 2008 Apple Nano Apple Video SanDisk Sansa Microsoft Zune

29 unique pacemaker/.ICD models Manufactures: Medtrionic, Inc (15 models, 27 devices Boston Scientific Corporation/Guidant (7 models,

13 devices St. Jude Medical Inc.( 7 models, 11 devices)
