Target controlled infusion pump failure due to worn drive nut.
The patient, 110 kg, 183 cm, with poorly controlled hypertension, asthma and a previous deep venous thrombosis, was in a prone position, receiving inhalational anaesthesia (sevoflurane) and remifentanil (Minto model) when the syringe was changed at 1 hour 56 minutes. This alarm had been observed by the same anaesthetist on two previous occasions. The pump had been sent to the Biomedical Engineering Department where software examinations revealed only that the drive had experienced 'excess deviation', which the manufacturer notes (personal communication, CareFusion Australia) occurs when there is a greater than 2.4 mm mismatch between the expected syringe plunger position (derived from motor encoder counts) and its measured position (measured by linear potentiometer). The biomedical engineers did not recommend any further action and the pump was returned to service. Communication with the manufacturer's representative had previously questioned the possibility of operator misuse, with the syringe being removed prior to placing the infusion on hold.
As a result of the two previous problems, attention to ensuring the pump was on hold prior to syringe removal and regular recordings of the infusion rate were kept on all of the reporters' TCI anaesthetics, and when the pump would not recommence TCI, a continuous infusion was commenced at 18.6 ml/hour, representing the most recent infusion rate. The anaesthetic was completed without further incident.
On this occasion the problem was reported using the TGA Medical Device Incident Reporting scheme, and the manufacturer returned the pump to the UK for investigation. Examination of the pump revealed the half nut to be heavily worn (Figure 1) and data recording revealed that nine such errors (identified as 'PL3' on PK pumps) had occurred in the previous seven months. The event log also revealed a 'drive disengage' error at 58 minutes into the case, for which the reporter was not present, being the result of the plunger being released prior to placing the pump on hold, which is different from a 'PL3' error and is recoverable.
[FIGURE 1 OMITTED]
The 'PL3 Excess Deviation' error is not recoverable and the pump cannot return to TCI, although after powering down the pump on two occasions a new TCI infusion can be commenced, as illustrated by the presence of nine such errors. The manufacturer recommends that when such an error occurs the pump be removed from use and serviced. A general 'drive disengaged' fault will allow recommencement of the infusion and retention of data.
While this particular failure was a relatively minor inconvenience, as the infusion had reached a stable state and little change in the target concentration was required, the same would not apply earlier in a case or when multiple changes are necessary, as all of the TCI pump benefits of bolus and rate change adjustments to increase target concentration and delay infusion and rate adjustments to decrease target concentration would be lost.
Problems arising when a half nut is worn, and resultant failure to engage fully with a leadscrew, have been reported on the Graseby 3500 TCI pump (1). However, in those reported cases the pump did not alarm to indicate a mismatch between expected and measured plunger positions, and one patient experienced awareness. (Subsequently half nuts in earlier versions of the Graseby pump were replaced with 'super nuts' designed to reduce the potential for wear.)
The Alaris manufacturer has suggested that an incorrect declutch technique, in which the finger grips are not fully squeezed when sliding the plunger, may result in wear of the half nut, which would suggest that we should be paying careful attention to this action. The manufacturer also suggests that it is likely that a rasping sound would be heard during this incorrect operation, as the half nut is being dragged across the lead screw thread.
It may also be the case that there was miscommunication in describing the alarms, and that 'drive disengaged' and 'PL3 Excessive Deviation' were not clearly distinguished, including during the communication with the company representative. This would suggest that we all need to be well versed in the various alarms and faults that may be displayed on all of the devices we use.
I. R. COX
Sydney, New South Wales
(1.) Laurent S, Fry R, Nixon C. Serial failure of Diprifuser infusion pumps. Anaesthesia 2001; 56:596-597.
|Printer friendly Cite/link Email Feedback|
|Publication:||Anaesthesia and Intensive Care|
|Article Type:||Letter to the editor|
|Date:||Jan 1, 2012|
|Previous Article:||Correction: adherent transversus abdominis plane catheter--reply.|
|Next Article:||Three strikes and you're out: unanticipated hyperkalaemic cardiac arrest following rapid sequence intubation.|