GUIDE TO INSPECTIONS OF ELECTROMAGNETIC
COMPATIBILITY ASPECTS OF MEDICAL DEVICE
not impossible to pinpoint because it is not possible to determine the exact circumstances of an incident after it has occurred. It might even be impossible to detect during an event if the source is unseen, intermittent, and/or mobile.
Example: A monitored patient (and his nurse and doctor) may not be aware of the cellular telephone in the pocket of another patient's visitor.
However, it is sometimes possible for a firm to establish the credibility of an incident by duplicating the problem in the normal use environment or in the laboratory.
A firm should be aware that if it services or investigates a device outside its normal use environment, the device may experience a significantly different electromagnetic environment which may possibly mask an EMD source found during normal use.
Example: Interference from some electrosurgical units (ESUs) disrupted the communication link between an anesthetic gas monitor and a central mass spectrometer, causing the monitor to fail to display the concentration of anesthetic gas in the operating room during surgery. When testing to find the failure mode, a firm might not think of simulating an ESU with these devices in a laboratory.
If a device type is known to be susceptible to EMD and a cause cannot be attributed to something other than EMD, the firm should conduct "ad-hoc" susceptibility testing in an environment simulating the actual use environment.
If the firm determined that EMI is the cause of the device's nonconformance, the investigator should question whether any product lines using similar designs and subject to the same environment have been assessed or determined to be prone to EMI. Investigators should also review records maintained under the requirements of 21 CFR 820.90, to determine if data can further identify the problem to a manufacturing step, work station, or one or more employees.
The specific EMD source that caused the device malfunction may not be the only source which will affect the device performance. The firm should also attempt to determine any other EMDsources that can be expected in the use environment which may have emerged since the design inception.
Whatever action is proposed should consider the frequencies, power output, modulation characteristics, proximity, and shielding of the EMD sources that affect the device, as well as the user's ability to be aware of and control the proximity of EMD sources.
Example: Cellular telephones are not always seen and proximity to equipment cannot always be controlled.
If the design change is not verified, the device performance must be validated in the actual environment or an accurate simulation of it. Labeling changes may be validated by querying different user populations as to whether they believe the labeling provides adequate warning.
If verification or validation indicate that the solution does not solve the problem and prevent recurrence, this step will have to be performed again in its entirety once a new solution is identified. The firm should verify that the correction does not introduce any new EMI, electrical safety, or other problems.
- Identifying a Solution or Preventive Action When EMI is confirmed, a firm will often choose to shield the device (e.g. different selection of shielded cable, cabinet shielding, etc.) and/or redesign the vulnerable circuit(s). Labeling containing warnings or precautions may be necessary to supplement the design changes.
- Verifying or Validating the Solution The device should be verified or validated with new shielding and/or new circuitry and/or other physical adaptation in place. If verification is performed, the firm should test the device in a laboratory to a generally recognized EMC standard or a reasonable and well-defined test protocol, simulating conditions of use. Testing should be conducted on finished devices. If a device is to be offered sterile and reusable, the device should be tested after several cycles of sterilization as well.