Anaesthesia Critical Incident Data in Ireland

Following a request by the College, the State Claims Agency (SCA) undertook to ascertain the types of clinical incidents being reported on the National Incident Management System (NIMS) that are potentially of relevance to the practice of anaesthesiology and its sub-specialities. By way of background, the SCA manages personal injury and third-party property damage claims against the State and State authorities, as delegated to it, and in providing related risk management services. The SCA has a statutory risk management function. The Clinical Risk Unit (CRU) implements the Agency’s clinical risk management role and aims to reduce the risk of litigation from adverse clinical events through advancing patient safety. The SCA hosts the National Incident Management System (NIMS) a confidential end-to-end risk management tool developed by the SCA that allows the SCA and State authorities to manage incidents throughout the incident lifecycle. State authorities are required to use NIMS to fulfil their statutory requirement to report incidents to the SCA, and may also use the system for their own risk management purposes. NIMS provides State authorities' risk managers and the SCA's own risk teams with complex adverse incident data analysis and reporting capabilities. This enables risk management and mitigation responses that help to improve the safety of patients and service users, and minimise the cost of claims against the State in the future. The accurate reporting of incidents on NIMS is critical to the SCA's risk management function and the SCA works actively with State authorities on an on-going basis to improve the level and quality of reporting. Methodology: In the normal course of events, the SCA carries out a quarterly incident surveillance process of certain categories of new service user incidents reported on NIMS. Inclusion criteria vary by quarter and not all service user incidents are included in each quarter. For the first 6 months of 2020, incidents of potential relevance to anaesthesia were identified by manual review, and then subjected to more extensive analysis by a single clinical risk adviser (CRA), who has clinical experience in anaesthesia. Results: The SCA’s Clinical Risk Unit reviewed 14,215 service user incidents in Q1 2020, and 12,054 service user incidents in Q2 2020, yielding a total of 26,269. The initial high-level analysis of these incidents was based on the information available within the ‘summary of the incident’ field on NIMS. • 163 of the 26,269 clinical incidents reviewed (68 in Q1, and 95 in Q2) were initially identified as being of potential relevance to anaesthesia. Following further review, 12 of these were excluded. • A total of 151 events (0.6% of the total) were identified as pertinent to anaesthesia, and were subjected to further assessment using all of the additional data fields available on NIMS. • Two of the 151 had a fatal outcome, but further information about cause of death or extent of anaesthesia involvement is unavailable to us at the present time. • Of the remaining 149 non-fatal incidents, almost half related to maternity services. It is notable that only 2 incidents were specifically logged as “anaesthesia”, the remainder coming under a variety of service headings (fig 1).

The sub-analysis that follows will focus only on significant anaesthesia-related incidents. Some of these might be considered to be normal, rather than adverse, events but have been reported nonetheless. 1. Anaesthesia for Maternity Services (n=72). a. Several of the logged incidents related to GA for CS. i. One of these was a documented aspiration. ii. 15 involved the unplanned conversion from RA to GA for caesarean section. o Two of these related to high spinal blocks, and o One to an “acute loss of consciousness” while awaiting CS under epidural. iii. Three medication events necessitated conversion from spinal anaesthetic to GA for CS: o 1 x thiopentone administered in error for antibiotic. o 1 x suxamethonium administered in error for anti-emetic. o 1 x potential allergic reaction to antibiotic. b. The majority of maternity-anaesthesia events related to regional anaesthesia. Of note amongst these: i. 1 x epidural medication connected to IV cannula. This was quickly noted and rectified and no harm resulted. It is still a significant “near miss” incident. ii. 1 x mix up in syringes, so that the drug designated for intrathecal injection was used for skin infiltration. iii. 1 x out of date epidural infusion was connected. iv. 1 x complaint of inadequate psychological preparation for operation under RA. c. Others i. 1 x maternal collapse post emergency LSCS, on patient controlled analgesia (PCA), with a multi-disciplinary response. No further information currently available. ii. 1 x potential aspiration or allergic reaction requiring intensive care in a postnatal patent. iii. 1 x clinically contraindicated IV site used, requiring to be changed. iv. 1 x anaesthetic machine failure and closure of an operating theatre resulting in an out of hours elective caesarean section. d. Outcomes are categorised on NIMS according to severity as being negligible, minor, moderate (significant injury requiring medical treatment), major (injury with long term incapacity or disability) or extreme (incident leading to death or major permanent incapacity) (fig. 2). i. Of the 72 of the anaesthesia related incidents within the maternity service logged on NIMS, none were recorded as major or extreme in severity of outcome. ii. 53% (n=38) were recorded as being negligible (n=36) or minor (n=2) in severity; iii. 47% (n=34) were of moderate severity. 2. Events in other services (n=77)). a. 11 cases of dental or oral trauma were logged. b. 1 each of: i. Wrong site (side) block. This was a nerve block carried out on the wrong side. This was 1 of the 2 events which were logged as “anaesthesia”. ii. Potential awareness. iii. Anaphylaxis. iv. Soft tissue injury during to positioning under SAB. v. Bronchospasm during insertion of hysteroscope. vi. There were 2 incidents of extravasation from peripheral cannulae. c. There were a number of incidents relating to the lack of availability, or failure, of ‘essential’ equipment. None of these appear to have resulted in patient harm. d. A substantial amount of logged incidents in this group related to communication issues in some form. Of particular note were i. Patients who had protracted or unnecessary anaesthesia because of the unavailability (temporary or complete) of surgeons, equipment, tests, relevant radiological images, or high-dependency beds. ii. Patients had made the trip to hospital, and were admitted, but where there was no anaesthesia service available, or provided, for that list. e. Outcomes: i. Of the 77 anaesthesia related incidents logged on NIMS in services other than maternity, no events were reported with a severity outcome of major or extreme (fig 2). o 86% (n=66) of these incidents were recorded as being negligible in severity. o 6% (n=5) as minor. o 8% (n=6) of moderate severity (‘injury requiring medical treatment’).

Commentary: This is the first Irish data in respect of safety events relating to anaesthesia. It provides a 6 month snapshot of adverse incidents that have been logged. In many ways the data is in keeping with the findings of the 12 year review of litigation related to anaesthesia in the NHS where the majority of cases related to regional or obstetric anaesthesia, and dental injury featured highly. In that series, anaesthesia-related claims accounted for 2.5% of all claims and 2.4% of the value of all claims. A simple analysis of the Irish data would seem to indicate that anaesthesia-related incidents account for 0.6% of logged events. However, it should be borne in mind that the UK data relates to claims, and the Irish figures to logged incidents, and even then not all service user incidents were included in the denominator. It is important to highlight the reported medication errors, particularly “syringe swap” events. These included the administration of thiopentone in error for an antibiotic, and the administration of suxamethonium instead of an anti-emetic. It is worth recalling that the NAP 5 study on accidental awareness also included cases of similar medication errors. Such medication complications appear perennial, and make a strong case for the use of prefilled syringes. The accidental connection of epidural infusion to a venous cannula is what might be considered “a never event”. However, we know that humans, no matter how motivated to do otherwise, will inevitably find a way to make particular errors unless a technological solution has been imposed that renders this impossible. In this instance, it is regrettable that Ireland has been slow to introduce a dedicated central axial delivery system, which has been mandatory in the NHS for some time. The CAI Quality and Safety Committee continue to advise that all members and fellows should follow The European Board of Anaesthesiology recommendations for safe medication practice. Recently, the College has undertaken an initiative to roll out "bootcamp" simulation training for all new entrants onto the scheme, which includes a medication safety module. It is also notable that communication difficulties featured strongly in the list of recorded incidents. Modern healthcare is delivered by multidisciplinary healthcare teams who rely on effective teamwork and communication to ensure successful and safe patient care. There is a substantial body of literature that attributes unintended patient harm to failures in communication between health professionals. And, accumulated evidence indicates that improvements in teamwork can lead to significant gains in patient safety, including efficiency of care and reductions in complication rate and mortality. Tribalism, or us and them teams, should have no role in contemporary healthcare. And, while some multidisciplinary simulation training does take place in Colleges, little to none occurs in our hospitals. There remains considerable scope for improvement in interdisciplinary communication. Finally, very few of the incident forms were completed by anaesthesiologists. In keeping with the national trend, the vast majority were submitted by nursing or other HSC professionals. Therefore, while some anaesthesiologists might express frustration with incidents of negligible severity being logged, the reality is that this situation reflects the failure of many of us to engage effectively, or at all, with incident reporting or monitoring systems. Acknowledgements: The College is grateful for the efforts of the Clinical Risk Unit within the State Claims Agency for compiling and categorising the above data.