Summary of MedSun Reports Describing Events With Physiological Monitoring Systems with Software Issues
MedSun: Newsletter #47, April 2010

The four primary groups of devices that are in this report are described as:

*Physiological monitors intended for multiparameter patient monitoring. The devices will produce visual and audible alarms if any of the physiological parameters monitored vary beyond preset limits; timed or alarm recordings will be produced (1).

*Physiologic/hemodynamic monitoring and information gathering systems that facilitate clinical data acquisition and analytical assessment for cardiac catheterization, invasive radiology and electrophysiology laboratories (1).

*Central monitoring of multiple adult, pediatric, and neonatal patients; and where the clinician decides to monitor cardiac arrhythmia of adult, pediatric, and neonatal patients and to gain information for treatment, to monitor adequacy of treatment, or to exclude causes of symptoms (1).

*A module is a modular removable unit that, when used in conjunction with a mating physiological monitor provides the capability to acquire various common physiologic data in a clinical setting (1).

Device descriptions may be referenced at the FDA website for 510K Premarket Notifications

Over the past 2 years, MedSun has received 17 adverse event reports associated with devices manufactured by Draeger Medical Systems, Philips Medical/WITT Biomedical, GE Healthcare, Philips Medical, and Spacelabs Healthcare. The reports were submitted by 12 hospitals between February, 2008 and January, 2010.

The reported device problems were:

5 - Device Rebooted/reset on its own (when it should not have)
4 - Clinical data loss/miscalculation
3 – Communication problems between systems
3 - Screen/display blanking
1 - Software updates (sites did not know upgrade was available)
1 - Human factors

No reports involved a patient death. The potential for patient harm was listed in 16 of these reports. The potential patient outcomes:

9 - Prolonged patient care
4 - Possibility of delayed patient treatment
3 - Possibility of misdiagnosis

Of the reports that listed patient age, 1 had a patient age listed as less than 21 years, and 5 had a patient age listed as greater than 21 years. Of the reports that listed patient gender, a total of 5 reports involved female patients, and a total of 1 report involved male patients.

These MedSun reports contributed to FDA awareness of the device problems. FDA follow up with the manufacturer resulted in software updates to systems and devices.

The following table lists the MedSun reports that are described in the device problem summary above.

Events With Physiological Monitoring Systems with Software Issues
ManufacturerDevice Identifiers Description
Draeger Medical Systems, Inc. Infinity Kappa XLT (Report 1)
After the running 15-30 minutes, the monitor powered off and restarted automatically after 30-60 seconds without intervention by the anesthesiologist. Manufacturer response (as per reporter) The manufacturer replaced four of the monitors and reportedly confirmed one as a hardware failure, two units as inconclusive and one report remains outstanding. The stability of the software was in question.
Draeger Medical Systems, Inc. Infinity Kappa XLT (Report 2)
After the running 15-30 minutes, the monitor powered off and restarted automatically after 30-60 seconds without intervention by the anesthesiologist. Manufacturer response (as per reporter) The manufacturer replaced four of the monitors and reportedly confirmed that one as a hardware failure, two units as inconclusive and one report remains outstanding. The stability of the software was in question.
Draeger Medical Systems, Inc. Infinity Kappa XLT (Report 3)
After the running 15-30 minutes, the monitor powered off and restarted automatically after 30-60 seconds without intervention by the anesthesiologist. Manufacturer response (as per reporter) The manufacturer replaced four of the monitors and reportedly confirmed one as hardware failure, two units as inconclusive and one report remains outstanding. The stability of the software was in question.
Draeger Medical Systems, Inc. Infinity Kappa XLT (Report 4)
After the running 15-30 minutes, the monitor powered off and restarted automatically after 30-60 seconds without intervention by the anesthesiologist. Manufacturer response (as per reporter) The manufacturer replaced six of the monitors and reportedly was unable to identify the problem. The stability of the software was in question.
WITT Biomedical Subsidiary of Philips Medical Systems Xper Xper Hemodynamic Cath System, Information Management (IM) upgraded to 1.3.1 on November 3. On November 4, the Xper crashed in the morning during cases/ couldn't chart. Message was "Net Failure".
GE Healthcare Solar 8000M
Tram T451M
I was called to the CICU yesterday afternoon regarding a concern related to an entry on a patient's full disclosure report obtained from a patient's cardiac monitor. It was noted on the disclosure report @ 22:20:59 there was a second tracing of asystole. However, the monitor did not note or record an alarm at that time. The patient's record notes she was transferred to the CICU earlier @ 2100, alert and oriented with a heart rate of 30's; treated with a Dopamine gtt and fluids. Her medical record does not note this 10 second episode, nor does it demonstrate any change in her status (no injury) with this occurrence. On assessment, the patient remains alert and the monitor functioning without incident. Monitor tech (nurse) put in place yesterday to facilitate close visual watch over central monitoring; nurse monitor tech) was released this morning after both patient's bedside monitor and Tram module were replaced for further review.
Philips Medical Systems IntelliVue IntelliVue During patient telemetry EKG/Arrhythmia monitoring, the transmitter battery got weak and died (stopped functioning). The central station was configured to alert the nursing staff through a secondary alarm messaging system to a RF wireless phone system when the battery became weak, and again when the battery needed to be replaced. The "battery weak" alert was transmitted to the nurse by way of the RF wireless phone, the "replace battery" alert was not sent to the nurse. Manufacturer response (as per reporter) A manufacturer's representative has examined the system and observed the problem. It was indicated that further tests needed to be performed at the factory to determine the cause of the problem. It was determined that the weak battery message was sent to the secondary alarm system, but the replace battery alarm was not sent.
GE Healthcare Solar 9500 Patient was induced and shortly after the sternotomy a "lead fail message" was displayed on the monitor as one of the ECG leads (lead V) disappeared. The monitor was adjusted to display two of the limb leads (lead II & lead III) instead. All of the ECG leads and electrodes had been carefully placed and covered, so it was considered unlikely that a pad (electrode) had come loose. The cable connection near the patient was also checked. The CVP pressure tracing at the same time was noted to exhibit significant noise artifact and this did not improve after replacing the pressure cable. A few minutes later, it was observed that the arterial line tracing would disappear suddenly from the screen without anyone disturbing or changing anything. Then the arterial line tracing would reappear. This did not improve after changing the pressure cable several times and changing the transducer, so it was assumed that the problem was with the physiologic monitor. In view of the fact that the monitor was critical to the surgical procedure, it was opted to use a transport monitor to display the patient's physiologic information. It was determined that the memory in the monitor needed to be rebooted to clear this issue.
GE Healthcare S/5 Extension Module Intraoperatively, some of the remote monitors were noted to be blank in the OR. The monitors came back online spontaneously and then MD was able to reproduce problem by flexing cable on the Extension Rack Module with BioMed in room. BioMed Assessment: Replaced and tested with a new module. Removed the defective module from inventory. The module was suspect to having a software issue.
GE Healthcare Carescape, PDM Five previous Medsun reports were filed last year. Resolution from GE is still pending. Over fifty PDM, Carescape modules are sitting in a box in storage, which have not been deployed to clinical use since receiving them. Unable to obtain the update or definite time line to when the problems will be resolved. The previous Medsun reports filed were for: 1) NIBP Hardware failure with PDM 2) NIBP Issues with PDM and Tram 3) Rebooting issues in PDM 4) SpO2 issues with PDM 5) Temp Issues with PDM. Manufacturer response: Released software upgrade for Solar 8000i. Upgraded monitors at the hospital. Put PDM's (Carescape module) back into clinical use into a small pod of the OR's as a pilot. PDM's experienced the same failures as mentioned in the MedSun reports filed. Pulled devices out of use. Alerted GE. Solution still the hospital is pending a response from GE, with no update or timeline since. The PDM's have been out of clinical use for ~ 1 year at this time.
Spacelabs Healthcare Inc Command Module, 90496 Spacelabs solution was to add an additional server but this still did not work. There are two fundamental things deficient with this system; 1. We have no active monitoring system that detects when the 72-hour disclosure interface is not sending info to the database. The only way staff knows the system is down is when they need to review cardiac rhythms or vital signs and the screen is blank. 2. The current system configuration saves information in a Flash Type memory, when the interface/server is restarted all the information is purged and gone forever. The Spacelabs system needs to write to a permanent archive (hard-drive) that can be accessed after the system is rebooted and restored; the flash memory does not save the tracings when it is restarted. I don't know how this patient critical system can be deficient in both these areas and still be FDA approved? Manufacturer response: It was our IT dept's responsibility to come up with a solution to monitor the interface.
WITT Biomedical Subsidiary of Philips Medical Systems Xper (Report 1)
This is one of three systems having the same issue. The hemodynamic monitoring system has a known software problem that causes the machine to return to pre-edited measurements after a few minutes. Staff need to make notes on print outs and dictate correct results. Philips updated the software in September to: 1.1.1.1291 and again in January 2009 to version 1.2.1474 to resolve the latest issue.
WITT Biomedical Subsidiary of Philips Medical Systems Xper (Report 2)
This is one of three systems having the same issue. The system samples invasive hemodynamic pressures, upon review of a sampled pressure measurement, the technologist edited the measurements. After a few minutes, the system changed the edited pressures back to the pre-edited measurements. The system doesn't print out accurate information. Staff has to make notes and dictate the actual results. Company has stated that it is a known software problem. Philips updated the software in September to: 1.1.1.1291 and again in January 2009 to version 1.2.1474 to resolve the latest issue. Manufacturer response: This system performance issue is a potential patient safety issue. Philips has escalated our issues through their system and has agreed to an on sight meeting on in January to develop a plan of action to address these problems.
WITT Biomedical Subsidiary of Philips Medical Systems Xper (Report 3)
This is one of three systems having the same issue. The system samples invasive hemodynamic pressures, upon review of a sampled pressure measurement, the technologist edited the measurements. After a few minutes, the system changed the edited pressures back to the pre-edited measurements. The system doesn't print out accurate information. Staff has to make notes and dictate the actual results. Company has stated that it is a known software problem. Philips updated the software in September to: 1.1.1.1291 and again in January 2009 to version 1.2.1474 to resolve the latest issue. Manufacturer response: This system performance issue is a potential patient safety issue. Philips has escalated our issues through their system and has agreed to an on sight meeting on in January to develop a plan of action to address these problems.
GE Medical Systems Information Technologies and Cerner Corporation Dash, Solar, Millennium Nurses are currently manually entering vital signs data into the electronic medical record (EMR). Nurses have noticed that the mean pressure value that is calculated by the EMR is sometimes different from the mean value displayed on the patient monitor. A few sample values were obtained: (systolic/diastolic mean) 129/65 81 EMR mean 86; 146/78 95 EMR 101; 134/54 81 EMR 84; 150/98 116 EMR 115. We were unable to verify if this occurs with a different patient monitoring vendor (such as in the ER) because in that case the EMR does not display the mean value. Nurses reporting this problem say that they have seen much larger variations but I do not have specific numbers to report yet. Manufacturer response: Per clinical support with GE, the mean pressure is measured in the oscillometric method and is the most accurate pressure value since the systolic and diastolic are then calculated from the mean. Most EMR's take the systolic and diastolic numbers and then calculate the mean using a formula intended for auscultatory readings. Therefore there will be variations.
Philips Medical Systems IntelliVue Telemetry Unit reported that there were issues associated with a Philips monitor displaying a discharged patient's rhythm and other Protected Health Information (PHI) showing up on a newly admitted patient's rhythm and PHI. Risk Management spoke with a Philips Customer Service Rep. who had already been contacted by our facility Biomedical Dept. about it, and the monitor was sent to their corporate office for analysis and for a report to be provided.
A Philips Field Service Engineer (FSE) came on-site to the facility to download the monitor equipment's history for evaluation and complete a performance testing on the device. Two months after the event a Philips Healthcare Informatics/Pt Monitoring Quality & Regulatory Specialist provided a report specifying that on the date of the report a newly admitted patient had an alarm occur and staff noted an alarm strip with a similar rhythm was in this new patient's alarm history, but the strip had an earlier date on it which was prior to the admission date of the second patient. The report said they had investigated the issue and determined there was insufficient information provided to confirm that any device malfunction occurred. The FSE verified the device was performing according to specifications and was alarming and storing/recording alarm strips appropriately. The FSE collected and analyzed log files but a complete analysis could not be performed due to the prior patient's admission and discharge had not been provided. The Logs did show the last discharge/remove data action performed for a patient prior to the incident two days prior to the incident @ 16:22. In some cases, a patient may be discharged from the central but may remain on their monitoring device for short time and may continue to generate alarms which will then be stored for that bed. If another discharge is not performed before a new patient is admitted, this prior alarm data will be appended to any new patient admitted to the bed. Philips indicated information about admitting/discharging could be found in the IntelliVue Information Center Instructions for Use, Part # M3150-9001J, in the Section on Patient Management. Philips HC indicated they would continue to review and monitor internal quality records associated with the involved product and would take appropriate actions as necessary.
GE Medical Systems Information Technologies Dash 4000 Monitors would beep then screen would go blank, a few seconds later screen would come back but data would be changed and clock time is off. This happened with several monitors so it affected several babies.


Additional Information:

1. "510(k) Premarket Notification". U.S. Department of Heath and Human Services, Food and Drug Administration.
http://www.accessdata.fda.gov/scrIpts/cdrh/cfdocs/cfPMN/pmn.cfm


MedSun Newsletters are available at www.fda.gov/cdrh/medsun