[5442357]
This spontaneous medical device incident report, from a health care professional in (b)(6) involves a cd medical oxygen cylinder fitted with a gce combilite valve. The anaesthetist was called to a patient on the surgical ward. The patient was extremely sick and deteriorating, and needed transfer to intensive care for stabilisation/resuscitation. He was being given maximal flow oxygen via a facemask with a reservoir from the wall-piped oxygen. For the transfer the patient was swapped onto oxygen delivered by one cylinder. The physician met the nursing team in the corridor, rushing back to the ward as the patient had deteriorated en route. Once back on the ward, it became apparent that the patient had arrested. Resuscitation attempt was unsuccessful and the patient subsequently died. It was noticed the oxygen was not flowing at 15 1/min as it should have been. On further interrogation of the cylinder, it was not empty as expected, but faulty. Gas would flow when the gauge was between numbers, but when exactly on a number, e/g/ 15, there was no flow. No outcome of any post-mortem examination on this patient is available, so no cause of death provided. The anaesthetist reported that the failure of the cylinder will not have helped the situation and may well have worsened the state of the patient, but he doubted very much it was the direct cause of the cardiac arrest. Analysis of the cd medical oxygen cylinder including flow measurement from the fir tree, pressure, leak and purity testing was carried out. The analytical report ((b)(4)) dated (b)(6) 2015 revealed that there were no leaks detected. Further, no flow was detected at any of the correct click stop settings. By moving the position of the click stop to approximately half way between the marked settings a flow could be obtained. The flow values measured in these half way positions varied depending on whether measurements were made moving up through the scale or back down. No other defects were identified. This incident was considered as a classic worn flow disc, an issue that was identified in 2008/2009. The preventative action of a new flow disc design was approved in (b)(4) 2009. Any cylinders then identified with a faulty flow head (incorrect flow) was fitted with a new flow head. In (b)(6) 2010, a decision was made to routinely change the flow head of any cylinder presented to the test shop under the valve re-life process. It was later in 2010, that the routine flow head change at re-life came into effect following a period of testing, procedure update and spare part stocking. Details of the cylinder involved in the incident: barcode:(b)(4). Valve number: (b)(4) (manufacture date jan 2005). Cylinder number: (b)(4). The valve was fitted on cylinder on (b)(4) 2010 from the records in inlabel, but it may have actually gone in (b)(4) 2010. The inlabel record show accessory (valve) assembled on (b)(4) 2010 at (b)(6) and this will have been the initial associating the valve with the package in icc. A valve repair was performed on (b)(4) 2010 (recorded by admin on location 9999), but no details available. The accessory (valve) was changed on (b)(4) 2012 at (b)(6) and on (b)(4) 2012, by admin at 999 and these were just record changes. The actual valve was not changed on these dates and there is no record of any repair or test shop processing.
Patient Sequence No: 1, Text Type: D, B5