Beyond the numbers: Classifying contributory factors and potentially avoidable adverse events in the gynaecology service of National Women's Health at Auckland District Health Board.

From January 2019 to December 2022, 153 cases were reviewed in 37 meetings. Of the 153 patients, 135 (88.2%) were surgical patients and the remainder were related to medical management of early pregnancy (9.8%), sepsis (1.3%) and delayed diagnosis of ovarian cancer (0.7%). Table  1 compares the demographics for patients with AEs with women who had surgery in the general gynaecology service. There were 33 (21.2%) Asian patients in the study population, compared to 7.9% of the GSD population. Of the AEs affecting Asian patients, there were contributory factors in 17 (51.5%) cases, and 17 (51.5%) were potentially avoidable. Demography of patients Annual Clinical Report Gynaecology Surgical Database (GSD) 2020. The Asian ethnicity group comprises of 12 Indian patients, six Chinese patients, five South East Asian patients and eight in the Other Asian category. Ethnicity data from the GSD was simply categorised as ‘Asian’. Contributory factors were present in 77 (50.3%) cases (Table  2 ). Personnel factors were present in 54 (35.3%) cases, followed by organisational factors in 45 (29.4%) cases, and barriers to patients accessing and engaging with care were present in 11 (7.2%) cases. In 28 (18.3%) cases, both organisational and personnel factors were present, and in four (2.6%) cases, a factor from each group of contributory factors was present. Of the 77 cases where contributory factors were present, 27 (35.1%) had only one contributory factor, 41 (53.2%) cases had two to four contributory factors, five (6.5%) cases had five to six contributory factors, and four (5.2%) cases had eight to 11 contributory factors present. Of the 77 cases where contributory factors were present, 65 (84.4%) were classified to be potentially avoidable. Of these 65 cases, 38 (58.5%) had organisational factors, 48 (73.8%) had personnel factors and nine (13.9%) had barriers to care. Contributory factors and potentially avoidable morbidity of adverse events in gynaecology admissions at Auckland City Hospital, 2019–2022 Table  3 summarises the clinical diagnoses and type of AE. Forty‐four (28.8%) of the patients had AEs related to early pregnancy care. Nineteen cases (12.4%) were related to benign adnexal pathology, 28 (18.3%) cases were related to gynaecological cancer and a further 23 (15.0%) were related to surgery for abnormal uterine bleeding. Fifty‐four (35.3%) involved a return to theatre and 36 (23.5%) involved a delay in diagnosis or care, 20 (13.1%) involved a surgical injury recognised at the time of the index surgery, eight (5.2%) related to sepsis and a further nine (5.9%) were due to an admission to DCCM. Clinical diagnoses and type of adverse events More than one possible. Table  4 reports on the recommendations and actions that followed the AE reviews. Recommendations for quality improvement The following policies were raised as areas in which updates should be provided as a priority: emergency department care of women with abdominal pain emergency department care of bleeding in early pregnancy early pregnancy assessment unit protocol ectopic pregnancy protocols handover policy escalation to operating surgeon in the post‐operative period referral processes emergency department care of women with abdominal pain emergency department care of bleeding in early pregnancy early pregnancy assessment unit protocol ectopic pregnancy protocols handover policy escalation to operating surgeon in the post‐operative period referral processes Development of new policies for post‐operative care and laparoscopic surgery Discussion with the following healthcare services about failure to follow policy or best practice: intensive care department, emergency department, inpatient ward care, ambulance services, pathology service, other surgical services Referral to the hospital‐wide Adverse Event Review Committee for more detailed investigation of case Request hospital and community ultrasound services to include history of caesarean births in their reporting template Acquire point of care testing equipment Revision of discharge documentation to primary care doctors following medical abortion Presentation of cases at staff education sessions Of the 135 patients who had surgery, 95 (70.4%) were identified as having a post‐operative complication. Using the Clavien‑Dindo classification system 14 for grading severity of post‐operative complications, eight (8.4%) of the post‐operative complications were Grade 1 (a deviation from the normal post‐operative course), five (5.3%) were Grade 2 (requiring pharmacological treatment), two (2.1%) were Grade 3a (requiring surgical, endoscopic or radiological intervention under regional / local anaesthesia), 70 (73.7%) were Grade 3b (requiring surgical, endoscopic or radiological intervention under general anaesthesia), five (5.3%) were Grade 4a (intensive care required for single organ dysfunction), three (3.2%) were Grade 4b (intensive care required for multi‐organ dysfunction) and two (2.1%) were Grade 5 (demise).

There is no support to declare.

This study aimed to report on contributory factors and potential avoidability of AEs in gynaecology in order to inform quality improvement. Contributory factors were present in 50.3% of AEs and 42.5% of all the AEs were considered potentially avoidable. The majority of contributory factors were organisational factors and personnel factors. In the cases where contributory factors were present, 65% had more than one factor identified. The five most common factors overall were related to personnel; lack of recognition of complexity or seriousness of condition, failure to follow recommended best practice, a lack of knowledge in the skills of staff, delayed emergency response by staff and inadequate communication between staff. Fifty‐two percent of the cases where personnel factors were present also had organisational factors such as poor organisation of staff and lack of policies, protocols or guidelines. Barriers to patients accessing and engaging with care were present in only 7.2% of cases. These study findings are consistent with other research on AEs. In 2008, a systematic review of eight studies of AEs in in‐hospital admissions of 74 485 patients from developed countries including USA, UK, Australia, Canada, NZ, reported that 43.5% were preventable. 1 A systematic review of three studies in 2019 reporting gynaecology AEs found that 52.5% were preventable; no contributing factors were reported in the systematic review. 4 In a review of surgical AEs from five teaching hospitals in the UK, multiple factors were found to contribute to the AEs and many related to human factors. 10 Our study differs from all of these studies as we reported on a range of contributory factors encompassing organisation of care and management factors, personnel factors and barriers to patients accessing or engaging with care, which allows for greater understanding of the event. There are several strengths of the GMMR process. Firstly, it allows for a comprehensive review of an AE. The most recent edition of the London Protocol comments that the notion of a root cause in the root cause analysis of an AE is an ‘oversimplification’ and that many factors are usually present. 11 Rather than seeking to identify a root cause, the London Protocol recommends ‘thoughtful and comprehensive investigation and analysis of an incident’, 11 which the GMMR process endeavours to facilitate. A Dutch study looked at the feasibility of uniform reporting of AEs in obstetrics and gynaecology. 15 Their study identified the limitations of categorising every contributory factor. The GMMR review system caters for this limitation by having an ‘Other’ category for organisational, personnel and patient factors. Secondly, the use of a standardised checklist and categorisation offers a structured and systematic approach, which is recognised as being important for quality improvement and allows for recognition of patterns of these contributory factors. 1 , 7 , 16 These can then be used by the department to create actionable points. Thirdly, the process is designed to avoid unnecessary blame. Traditionally, healthcare professionals have feared individual blame when an AE review is undertaken. Shifting from a ‘blame culture’ to a ‘learning culture’ has been identified as a way to increase incident reporting, which facilitates quality improvement 15 , 16 , 17 The GMMR system focuses on identifying all contributory factors. Although AEs relating to human factors were common, in half of these cases organisational factors were also present. Finally, the use of the Clavien‑Dindo classification system allows us to grade the severity of any post‐operative complications reviewed by the GMMRC. 14 , 18  This provides the benefit of being able to compare rates and severity of complications over time, and to identify trends in factors that may be contributing toward more severe complications. There are also limitations to our study. We did not report incidence of AEs as we did not have a valid denominator. We compared the study population to the general gynaecology population which did not include patients from the gynaecology oncology database or patients from the abortion service. We considered that this denominator was justified as only 18.3% of the cases in the study population were gynae‐oncology patients and only 6.5% were related to abortion services. A limitation of our review of four years of data was that it was retrospective, although most cases were reviewed within two months. The intention was to have multidisciplinary meetings that allowed the perspective and insights of healthcare professionals with different backgrounds and areas of expertise to be taken into consideration. 7 Unfortunately, this was not always possible due to additional pressures over the time period of the study. A qualitative study exploring the perspective of patients and family affected by AEs concluded that learning from the event was one of four important aspects in AE review. 19 Learning is described as the importance of demonstrating change and improving the healthcare system and patient safety. 17 We agree with the findings of this study and acknowledge that the overall purpose of AE review is to improve delivery of health care. The actions that were undertaken following our AE reviews included developing and updating clinical protocols, developing specific operational plans to mitigate risks around delayed diagnoses with ectopic pregnancies and ovarian torsion, and a continued emphasis on the use of early warning scores and escalation pathways in the post‐operative clinical areas. Communication with staff about the learnings from the reviews occurred either by directorate emails to the relevant staff, through operational managers or presentations to the clinicians at educational sessions on key learning points. Our hypothesis was to ask if AE review using contributory factors and reporting avoidability could lead to quality improvement. We consider that the GMMR system provides a structured, systematic approach to review AEs which will help to identify all contributory factors and ultimately, identify areas in patient care that we can address to reduce adverse outcomes by informing quality improvement and the development of recommendations. It will be important to continue to monitor AEs to truly know if this approach of identifying and classifying contributory factors and reporting avoidability will lead to improved care and fewer AEs.

Adverse events (AEs) are defined as unintended injuries or complications resulting in prolonged hospital stays, disability at the time of discharge or death that is caused by health care rather than the patient's underlying condition. 1 AEs are common and can have a major impact on the overall wellbeing of patients. 2 , 3 , 4 Reducing harm from AEs is an important part of modern healthcare systems. 5 , 6 , 7 New South Wales and South Australia reported in 1995 that 16.6% of admissions were associated with an ‘adverse event’ which resulted in disability or a longer hospital stay for the patient, and 51% of the AEs were considered preventable. 2 A 2003 study of AEs in 13 New Zealand public hospitals reported similar results with 13% of admissions being associated with an AE. Nearly 40% were preventable. They concluded that system‐related issues were a prominent issue. 3 The Royal Australian College of Surgeons (RACS) guideline on Mortality and Morbidity (M&M) meetings recommends that robust M&M meetings should have a clear format, be conducted appropriately and effectively and result in outcomes or recommendations for improvements which are implemented and followed up. 8 Although some disciplines such as surgery and maternity care have reported if the AEs were preventable, 9 , 10 most M&M reviews do not report either contributing factors or avoidability. In 2016 the Gynaecology Morbidity and Mortality Review Committee (GMMRC) was established to review AEs in the gynaecology service at National Women's Health, Auckland, using a multidisciplinary process to identify contributory factors and avoidability with the overall aim of improving clinical care. The primary aim of this report was to determine the incidence of contributory factors and avoidability of AEs in the gynaecology service and to identify recommendations that could address these factors. Our hypothesis was that AE review with contributory factors and reporting potential avoidability would identify areas for quality improvement.

The gynaecology service at National Women's Health comprises the inpatient ward at Auckland City Hospital, the Women's Assessment Unit for acute gynaecology and outpatient clinics at Greenlane Clinical Centre. This includes general gynaecology, gynae‐oncology, urogynaecology, fertility treatment and abortion services. This study was a retrospective study of the clinical care of all patients who had an AE in the gynaecology service and were reviewed by the GMMRC from January 2019 – December 2022. Women over the age of 16 who were admitted under Women's Health Gynaecology Service at National Women's Hospital and had an AE were referred for review. An AE included the following: anaesthetic incident intraoperative complications involving additional support or unplanned treatments long stay in gynaecology ward, defined as more than four days for general gynaecology patients and more than six days for gynaecology oncology patients unplanned return to theatre within 30 days of first operative procedure unplanned admission to Department of Critical Care Medicine (DCCM) cardiopulmonary resuscitation event delay to treatment of gynaecological conditions, eg, missed diagnoses of cancer, ovarian torsions or ectopic pregnancies post‐operative sepsis leading to delay in discharge other cases identified by any staff members where concerns were raised. anaesthetic incident intraoperative complications involving additional support or unplanned treatments long stay in gynaecology ward, defined as more than four days for general gynaecology patients and more than six days for gynaecology oncology patients unplanned return to theatre within 30 days of first operative procedure unplanned admission to Department of Critical Care Medicine (DCCM) cardiopulmonary resuscitation event delay to treatment of gynaecological conditions, eg, missed diagnoses of cancer, ovarian torsions or ectopic pregnancies post‐operative sepsis leading to delay in discharge other cases identified by any staff members where concerns were raised. The primary outcomes were the number and category of contributory factors present in the AE reports and potential avoidability of AEs in the gynaecology service, National Women's Hospital. The list of contributory factors is presented in Table  2 . The secondary outcomes were demographic and clinical factors, including age, body mass indices and ethnicity for women with a gynaecological AE, and this was compared with women in the Gynaecological Surgical Database (GSD) at National Women's Health. The GSD does not include patients from the gynaecology oncology or abortion services. Clinical staff were asked to report all AEs in the gynaecology service to the Chair of the GMMRC. AEs at Women's Health are also reported using the GSD completed by the operating surgeon which informs the Annual Clinical Report. 9 Cases from the surgical database that had not been notified by clinical staff were also reviewed by the GMMRC. During 2019 to 2022, there was one case not reviewed by GMMRC as it involved both a coronial review and the hospital‐wide Adverse Event Review Committee (AERC). Two additional cases of patients with delayed diagnosis of ovarian torsion were reviewed by the AERC as they involved multiple services including primary care and the emergency department. The GMMRC protocol identifies contributory factors for AEs in gynaecology. A review committee of senior gynaecologists, gynae‐oncologists, the gynaecology ward charge nurse, senior nurses, and anaesthetists met monthly to review cases. The members of the clinical team for each AE were invited to take part in the review. The meetings were open to all clinical staff in the women's health department. For each case contributory factors were recorded using a classification system adapted from the London Protocol. 11 , 12 , 13 At each meeting the summary of the case was presented by a member of the panel who had also reviewed the medical records. Following the presentation, the review committee then sought to answer the following questions for each case, using a checklist of contributory factors. Have organisational and/or management factors been identified? Have factors relating to personnel been identified? Have barriers to patients accessing or engaging with care been identified? If contributory factors were identified, then the committee considered if the AE was potentially avoidable. An AE was considered potentially avoidable if the absence of the contributory factors would have prevented the event. For example, if a lack of adherence to protocols was considered a contributory factor in an AE, and the GMMRC considered that if there had been adherence to the protocol the AE could have been avoided, then the AE would be classified as potentially avoidable. All surgical cases also were classified using the Clavien–Dindo system for surgical complications. 14 Following each review, recommendations were made, including a communication plan. Ethics approval for the study was obtained from Auckland Health Research Ethics Committee (AHREC). AH25359 20/08/2023.