Portal de Periódicos da CAPES

Serial failure of Diprifuser infusion pumps

2001; Wiley; Volume: 56; Issue: 6 Linguagem: Inglês

10.1111/j.1365-2044.2001.2094-19.x

1365-2044

Sarah Laurent, R Fry, Chris Nixon,

Healthcare Technology and Patient Monitoring

We read with interest the letter ‘Failure of a Diprivan 1% prefilled propofol syringe’ (Breslin. Anaesthesia 2000; 55: 1030–1). Recently, we have experienced three SIMS Graseby 3500 TCI pump failures resulting in one case of patient awareness, which we would like to report. The first case of failure of the SIMS Graseby 3500 TCI syringe pump (serial number 43902) occurred in a patient having a laparoscopic cholecystectomy. The patient was induced with propofol using an initial target of 3 µg.ml−1, followed by morphine, tenoxicam and rocuronium. The patient became tachycardic and hypertensive, the target propofol concentration was increased to 6 µg.ml−1. When the total infused reached 15 ml, the patient was intubated and surgery commenced soon after. The patient remained tachycardic; the pump totaliser and target concentrations were checked, as were the intravenous lines to the patient. Only after some time was it noticed that the syringe plunger had not advanced beyond the initial bolus despite the totaliser and target concentrations depicting normal operation. No alarm sounded at any time. The second case involved failure of another SIMS Graseby 3500 TCI syringe pump (serial number 51564) in a neurosurgical patient, anaesthetised with propofol and remifentanil infusions several hours into the procedure. Knowledge of the preceding event, in conjunction with a rising pulse and blood pressure requiring an escalating dose of remifentanil, aroused suspicion in this case. It was noted that although the totaliser continued to increase, indicating ongoing infusion, the syringe plunger on the propofol pump had ceased moving. It was possible after recognition of the problem to observe this phenomenon over a period of approximately 20 min, during which time anaesthesia was maintained. After confirming the problem was happening, the syringe clamp was opened and closed, after which the pump resumed infusing and continued to do so until completion of the surgery several hours later. Testing of the first pump by engineers revealed a slight step on the casting of the drive half nut and that the leading edges of the thread had been chamfered. This resulted in failure of the drive half nut from being in full engagement with the leadscrew causing the plunger to stop pushing. The OPTO circuit board was incorrectly adjusted. When the OPTO board was reset the pump worked correctly. SIMS Australia recommends that pumps up to serial number 56385 have the existing half nut replaced with the latest issue known as the super nut. By upgrading to the ‘super nut’ adjustment of the OPTO board will not be necessary during routine maintenance. No specific mechanism for the failure of the second pump could be found on testing by the engineers but the fault was well demonstrated intra-operatively A suggestion to withdraw the entire fleet was deemed unnecessary but a general New Zealand wide warning was issued. The pumps were to have been quickly upgraded, the half nuts being replaced by super nuts, which apparently should prevent the problem reoccurring. The upgrade unfortunately could not be instituted immediately but a future date was set by the suppliers to commence the process. A departmental policy was instituted to record in the patient's anaesthetic chart, not only target infusion concentrations, but also the actual syringe volume at regular short time intervals to ensure early identification of a similar problem. Additionally, where possible, monitoring of the depth of anaesthesia by bispectral index should be performed. This policy did not assist in the early identification of a third SIMS Graseby 3500 TCI syringe pump failure (serial number 51566). This patient was again anaesthetised with propofol and remifentanil for a frontal neurosurgical procedure. One hour into the operation, after the third syringe change, the patient moved her arm. The syringe had delivered 10 ml and the pump appeared to be functioning normally. The start LED was flashing correctly and the control panel indicated active ongoing infusion. The syringe plunger was, however, motionless and propofol infusion had ceased. This resumed only after the syringe clamp was released and then re-closed. No alarms sounded and no patient harm resulted. Interestingly, the syringe pump alarm indicated an empty syringe when 7 ml remained in the syringe. This presumably represented the propofol not infused during the failure but recorded in the pump algorithm. All three pumps had been serviced earlier this year and were mains powered. The infusion tubing and non-return valve used were felt to be consistent with those recommended by the manufacturers. Although we do not have an engineer's report for the third pump, it was felt the failures should be published quickly due to the number of incidents within a very short period of time. Until we have an independent reliable feedback loop that confirms patients are indeed receiving intravenous anaesthesia, critical incidents such as these will always be a potential problem. The monitoring of drug effect during intravenous techniques remains the best option for continued use of this method, although in the two neurosurgical cases described here, not always practicable.