MedSun: Newsletter #19, December 2007
This section contains a sample of reports from all the MedSun reports received during a particular period. The reports were submitted by MedSun Representatives. In some instances the reports have been summarized and/or edited for clarity. The entries that follow represent a cross section of device-related events submitted by MedSun reporters during the period September 1 through September 31. All other reports can be searched under the ‘MedSun reports’ menu pane. Note: the two month delay is due to quality control and follow-up.
Device(s): Ventilator, High Frequency, 3100B Peds/Adult
Manufacturer(s): SensorMedics Corporation
Problem: The respiratory therapist reports that the ventilator settings were drifting. On each ventilator check, the knobs were being turned a little to obtain the original settings. The settings were set as follows: mean airway pressure (MAP) 32, amplitude (Delta P) 75, frequency (Freq) 4, and O2 100%. The ventilator was checked per protocol and the ventilator settings were as follows: MAP 21, Delta P 92, Freq 0.6, and O2 60%. The arterial blood gases (ABG’s) were poor to begin with; they had become slightly worse but returned to baseline once the ventilator was changed out.
Device(s): Humidifier, Heated, Respiratory, Comfort Flo Humidification System Catalog# 2411-03; Respiratory Circuit, Heated Wire ConchaTherm IV Plus Catalog# 400-50; Nasal Cannula Comfort Flo Catalog# 2411-03
Manufacturer(s): Teleflex Medical (Hudson Respiratory Care, Inc.)
Problem: Baby was on Hudson RCI Humidification System ConchaTherm IV Plus ref# 400-50. System included Comfort Flo Humidification System including Concha Column and tubing set ref# 2410, Comfort Flo Nasal Cannula (NC) ref # 2411-03 and Soln Water Concha Pak 1650ml ref# 381-50. Baby was on High Flow Nasal Cannula (HFNC) 60%@4 liters per minute (LPM). The Concha suddenly overflowed and shot water down the tube into the pt’s nares. Nurse disconnected the NC from the Concha. Tubing continued to flow safely onto the floor. RT’s changed out the concha. RN assisted in blowby oxygen as the saturation was 37. There was no injury to the baby other than desaturation to the 30’s when water first entered the baby’s nares. Device setting at time of malfunction was Adult +/- 3 degrees. Respiratory therapist thinks that due to the soft and very pliable nasal prongs, it bends and kinks off causing back pressure in the Concha until it explodes out.
Device(s): Tubing, Resuscitation Bag, CPR bag with Universal O2 Connector
Manufacturer(s): Mercury Medical
Problem: The patient was noted to be non-responsive so the basic life support protocol was initiated and the EMS called. Staff members attempted to connect stock CPR bags to the O2 tanks; however they were unable to utilize the CPR bags so an O2 nasal cannula was used in the interim. After this incident, it was discovered that the O2 tanks and CPR bags were incompatible since the O2 tanks had a “standard” connector, which had been phased in about one year ago. All sites were notified to check their stock to ensure they had only the standard connectors on all CPR bags. The purchasing department was aware that only standard connectors were allowed, however this incident resulted from the use of old stock that was not found and purged. The off-site rental center still had old style of connector, or “universal” connector on their CPR bags. This incident occurred at a remote location that does not frequently use O2, and apparently did not have to “bag” a patient until recently. To address this issue internally, it was first verified that the old connectors were purged by sending a flyer out to all department managers, followed by a written e-mail or a verbal confirmation of receipt from each of them. Additionally, it was confirmed that all new CPR bags will have the standard (new) connectors and that the purchasing department does not have the old style in their system. (For additional information about responding to cardiac arrest in outlying areas of the hospital see the Case & Commentary article featured in the October 2007 AHRQ’s WebM&M http://webmm.ahrq.gov/case.aspx?caseID=162)
Device(s): Catheter, Thermal Regulation System, Coolgard Start -Up Kit
Manufacturer(s): Alsius Corporation
Problem: The patient presented with cardiac arrest and Code Freeze was initiated. The cooling process was started. It was noted that the tubing that connects the machine to the catheter was not functioning properly due to a leak in the tubing. The tubing was exchanged without any problems.
Device(s): Defibrillator, Lifepak 20 Model# 3202487
Manufacturer(s): Medtronic, Inc.
Problem: Our facility has 36 Lifepak 20 defibrillators. Over the last few months we have had 5 units with a defective Power Module printed circuit board that enables the battery to be charged. One of the 5 units has had the circuit board replaced twice. When the units are unplugged to be used, the battery fails. We have noticed that the self-test is not totally reliable. If the Power Module PC board fails, and the unit remains plugged in to the AC mains power, the unit will pass the self-test and prints “Test Successful”, however, when the unit is unplugged, the battery is depleted and the unit fails. This was discovered when a nurse was ready to use this device and the battery failed. The unit then had to be plugged in to the nearest outlet. The five units which have had failures of the Power Module PC board are listed in this report. All defibrillators in the facility have had the recall issues involving the batteries, and the software upgrade installed as necessary.
Device(s): Filter, Dialysis Machine, Continuous Renal Replacement Therapy (CRRT) Model# 7106505
Manufacturer(s): B. Braun Medical, Inc.
Problem: CRRT Filter UF alarming “UF greater than expected.” UF variable amounts despite Calc UF. Taking off more than set amount.
Device(s): Dialysis Machine, Continuous Renal Replacement Therapy (CRRT), Prisma
Manufacturer(s): Gambro Renal Products, Inc.
Problem: RN was attempting a temporary disconnect procedure to pt on Continuous Venovenous Hemodialysis (CVVHD). He attached the NSS to the return line & pressed the return button. The pump started running backwards (the blood was infusing into the flush bag.)The RN stopped the pump and retraced the lines and had it double checked by another RN who confirmed correct set up. He switched the line since it was running backwards-to see if this would make a difference in flow. When he pushed the return button this time the machine started to shake and alarmed “Return Pressure Too High”. Pump stopped. Circuit discontinued without returning blood to patient. Patient’s hemoglobin and hematocrit stable. Pump tagged and removed from service. Evaluation by Biomed could not reproduce the problem-although original tubing/set up had been removed. For Biomed personnel, blood pump loaded and unloaded normally, all pressures and solenoids were verified along with all scales.
Device(s): Dialysis Machine, Multi-Therapy NxStage System One Model# CYC-D2E
Manufacturer(s): NxStage Medical, Inc.
Problem: Taken from staff report: “An IV tech tripped on the power cord - the machine was plugged back in and went through a power recovery. The blood and therapy pumps continued to function appropriately and the computer seemed fine until the next hour’s volume history was reviewed. The computer reported that -16.86 liters of therapy fluid was used but the machine was set for +2.4 liters. It also reported that -1262 mL of ultrafiltration were removed but the machine had been set for 168 mL. The next hour’s values were similarly incorrect. The RN returned the patient’s blood and switched to a new machine. Soon after, the patient’s blood pressure was declining and the RN had to restart the neosynephrine drip. The machine was red tagged and sent to Clinical Engineering.” Manufacturer’s representative later evaluated the machine and verified the machine was working properly. According to the manufacturer, the machine uses a separate computer for the history display. The history computer data can apparently be corrupted by an unplanned shutdown. Since staff rely on the history to monitor the patient, any obviously erroneous data is cause to discontinue the machine and replace it. We have actually had the same problem with several machines, mostly following a “controlled” shutdown and re-boot.
Device(s): Syringe, Insulin
Manufacturer(s): Becton Dickinson
Problem: The nurse correctly calculated the patient’s next dose of insulin as 8.5 units. The nurse went to the drawer containing the insulin syringes, each of which was in a separate package (0.3 cc (30U), 0.5 cc (50U), and 1cc (100U) syringes). The nurse mistakenly thought she had taken out a 10 unit syringe, but in fact, she took out a 100 unit syringe. The nurse drew up what she thought was 8.5 units of insulin, verified this with the charge nurse, and administered it to the patient. As she was withdrawing the needle, she realized there was an additional “0” next to what she thought was a “10” marker on the syringe. The nurse immediately ordered D10 (dextrose 10%) from the pharmacy, called the physician, and the lowest glucose reading obtained was 61. In reviewing the insulin syringes that were available, all three sizes have orange colored needle covers. Although the sizes are different, the differences are not extreme to a user who only occasionally administers insulin.
Device(s): Tubing, IV Lifeshield Lot# 53072 Catalog# 11965-68; Other Device# Any 53 series lot number
Manufacturer(s): Hospira Global Medical Affairs
Problem: Hospira recently made a change to their ‘Cair Clamps’. The Cair Clamp is the white plastic roller clamp on their gravity drip IV tubings. We are not certain yet in which lot number the change began (Their lot numbers are sequential). We do know that anything that is Lot number 52-113-4W or lower is the old style Cair Clamp. We were not notified of the change prior to discovering it ourselves - spurred by concerns from our hospital. The change in the roller clamp affects how the tubing is compressed and how the rate of flow of the IV is regulated.
Staff from our hospital related that:
1. “Roller clamp does not regulate IV drip properly. Runs all or nothing”.
2. “Runs too fast or not at all. Very Difficult to Run. Problem not with IV Site.”
The problems that were encountered at the hospital that lead to the request to pull lot numbers is directly related to the product change in the Cair Clamps. We were able to confirm this today when we visited the PACU and talked through the matters. The result of this is:
1. We can utilize lot number 52-112-4W that we previously requested be pulled from stock. It was intermixed with newer lot number stock at the hospital that had the actual problem.
2. The hospital is working on sequestering lot numbers of the product that are 52-133-4W or lower.
3. We will work with Hospira and Nursing on long term solutions. To date, Hospira has not provided the hospital with the list of lot numbers this change started affecting. We believe this new clamp is also going to be used on all their other sets, but Hospira has not given us dates on the roll out. (Note from FDA: Similar roller clamp problems have been seen with other IV tubing manufacturers.)
Device(s): pump, PCA II
Manufacturer(s): Baxter Healthcare Corporation
Problem: Biomed received several reports of Baxter PCAII pumps “not delivering meds”. End result of the extensive investigation by Biomed discovered a very dangerous scenario - potential patient injury or death - that may have already been played out (perhaps numerous times, including other possibly pediatric uses of this device), that went undetected until now. The scenario goes like this: A pediatric patient is put on a Baxter PCAII pump for pain relief. Certain medications (such as fentanyl and hydromorphone) must be used at high concentrations to be safe and effective. Dose is set for 0.2mL, with a 10 minute lockout period (standard), with possibly an initial bolus of 0.5mL programmed in as well. PCA button is pushed and meds are “delivered”. Unknown to staff, the line is occluded (perhaps kinked or clamped off), so no med is actually delivered to the patient. The patient continues to request meds, as they are in pain and not receiving meds. Pump “delivers” a max of 0.2mL every 10 minutes, without it actually getting to the patient. Over an hour or so, 1.9mL of fentanyl or hydromorphone are “delivered” before the pump recognizes occlusion and alarms. RN responds to alarm, sees patient in a lot of pain, and perhaps either knowingly (unclamps) or unknowingly (line becomes unkinked), fixes the occlusion problem. At this point the RN has just released a backpressure of 1.9mL in concentrated meds, which is immediately delivered to the patient. If the RN does not recognize that this just happened, she may also manually deliver another 0.5mL bolus to relieve the patients obvious pain. Total “bolus” of med delivered = 2.4mL of fentanyl or hydromorphone. This is 5x the normal bolus amount, and 12x the normal dose request amount. On some pediatric patients, this could easily be potentially fatal! (Note: manufacturer specs for occlusion testing say to set rate to 99mL, device should alarm occlusion within 3 minutes. Quick math reveals up to 4.95mL could be “delivered” before it recognized the occlusion and alarms. This would be potentially even more deadly, and the device would still be within manufacturer’s specifications). Due to the easily disguised nature of this problem, this scenario has probably been played out before when used in this manner (pediatrics). Perhaps the pumps were even marked as “defective” and sent out for repair. Testing by Biomed’s and manufacturer’s would reveal the device is working “perfectly” within specifications, and would therefore probably be closed out as “no problem found” or perhaps “operator error suspected”. We feel this is a significant patient safety issue that other pediatric users of this device should be made aware of ASAP. As a remedy, we have sent out a housewide e-mail, and have a special nursing skills day station set up. We are exploring other (newer) devices that may be more sensitive to occlusion alarms.
Device(s): Transanal Endoscopic Microsurgery System Model# 2232
Manufacturer(s): Richard Wolf Medical Instruments Corp.
Problem: After completion of a transanal endoscopic microsurgery procedure it was noted that a black nitrogen tank had been connected to the insufflator when a gray carbon dioxide cylinder should have been used. When the E-cylinder connecting yoke was examined it was noted that it could be attached to the cylinder valve without the safety index pins coming in contact with the gas cylinder valve body. Further examination revealed that the yoke had been assembled with a dual threaded fitting that was too long for the yoke body. This resulted in the plastic sealing washer making contact with the cylinder prior to safety pin insertion. This defect would have allowed connection to any gas cylinder with any safety pin key configuration.
Device(s): Attachment, Drill, Surgical, Dura Guard Model# 5400-10-58
Manufacturer(s): Stryker Instruments
Problem: There are four attachments for the Stryker power drills that are used when raising the bone flap for craniotomies. Four of these attachments have demonstrated poor performance. After further investigation, it was discovered there was rust or debris deposits in a supposedly sealed area of the device that cannot be cleaned. Presence of this foreign material could contaminate the sterile field. There have been no patient injuries at this time. The manufacturer is aware of the issue and informed us that they are redesigning the attachments.
Device(s): Software, Stereotactic Radiosurgery XSight Spine Software, Model# XSight Spine Version 7.1.2
Manufacturer(s): Accuray, Inc.
Problem: The patient has metastatic lung cancer to the spine and other areas. The XSight software is designed to detect, correct and help the Cyberknife track the target lesion. Specifically, the software is supposed to detect orientation and anterior/posterior, superior/inferior, left/right and 3 rotations that include pitch, left/right roll and yaw. The left/right roll would get locked in on a certain number. One of those numbers would allow you to treat. You can reposition the patient and roll the table 5 degrees in either direction. The left/right roll number remains the same and provides a “start” button that allows you to treat. It was then noted that this existed for her prior treatment and she was treated 5 times with this error. No known sequelae to patient, but any sequelae is not likely to show up for months to years.
Device(s): Attachment, Shoulder Positioner, T-Max Manufacturer(s): Tenet Medical Engineering, Inc.; Smith and Nephew, Inc. Endoscopy Division
Problem: A patient was being positioned in the OR for a right shoulder arthroscopy using a T-Max Shoulder Positioner supplied for trial use by a representative from Tenet Medical Engineering Company. Prior to elevating the shoulder positioner portion of the table, the stability of the table attachment was checked. When the extension was raised, part of the table collapsed causing the intubated patient to slip. The Anesthesia and OR team caught the patient as well as supported the table to prevent it from falling. The surgery had to be postponed and the patient was extubated successfully. The patient had minor skin abrasion as a result of the incident. One staff member who helped support the patient and table from falling completely ironically sustained a shoulder injury. Analysis of the event revealed that the clamp on the extension failed when it was raised. Prior to the failure, the proper procedures were followed. This device will not be used for future procedures.
Device(s): Software, Laboratory Information SOFTMic(GUI), Model# SOFTLabMic 4.0.1 Build 14:5
Manufacturer(s): SOFT Computer Consultants, Inc.
Problem: Soft Computer Consultant Laboratory Information System loaded comprehensive patch 4.0.1 build 14:5 this past spring. Since then SoftMic result entry lost an important feature that alerts technologist when someone else is in the same patient database. Previous to the comprehensive patch, there was a properly functioning warning message that would alert the technologist when someone else was using the same patient database. Without this warning message, the following patient result entry could potentially not be saved. This is a user definable function that could lead to the loss of patient data.