Introduction

Patient safety incidents (PSIs) occur across healthcare and evaluating these events presents an opportunity to learn, improve work systems and prevent future harm. Patient harm related to the care of patients due to undergo endoscopy, during endoscopy or during post-endoscopy care are devastating events calling for careful systems-based analyses to explore contributory factors to improve safety. The Systems Engineering Initiative for Patient Safety (SEIPS) model provides a framework for analysing PSIs to identify contributory factors (C.Fs) across the wider healthcare system, and has been used to analyse incidents and improve patient safety across a range of settings, although it has not been used previously for incidents related to endoscopy care (1,2). A large cohort of endoscopy PSIs across England and Wales were analysed using SEIPS, focussing on cases with patient fatality or severe harm outcomes.

Aims & Methods

This study aims to identify C.Fs to endoscopy-related PSIs involving death and severe harm outcomes, reported to a national incident database across England and Wales. 
PSI reports involving endoscopy-related care between January 2017 and January 2022 were obtained from the National Reporting and Learning System (NRLS) and analysed using content analysis, enriched by the SEIPS framework. Descriptive data was also reported.

Results

Of 6,971 reported endoscopy PSIs, 125 involved death and 305 involved severe harm. After exclusion of unrelated PSIs,368 (5.3%) were analysed, spanning pre-procedure (126;34%), peri-procedure (104;28%) and post-procedure (138;38%) care. Follow-up Issues (115;31%) and known complications (109;30%) were the most frequent incident types. Content analysis was performed on 176 PSIs (48% of included PSIs with severe harm and death outcomes) with sufficient narrative data. Using SEIPS, C.Fs related to “People and Teams” were the most common (see table 1), identified 106 times (36%) across 77 (44%) incidents, with the most common subtype being ‘Communication’, identified 70 times across 57 incidents (32%). Organisational factors were identified 88 times (29%) across 78 (44%) incidents, with the most common being ‘Resource Availability’. Task factors were identified 61 times (20%) across 52 (30%) incidents, with the most common being ‘Documentation’. Technology and Tools factors were identified 23 times (8%) across 23 (13%) incidents, with the most common being ‘Referral System Failures’. External environment factors were identified 13 times across 13 (4%) incidents, with the most common being ‘Pandemic Disruption’. Physical Environment factors were identified 8 times (3%) across 7 (4%) incidents, with the most common being ‘Out-of-hours Environment’.

Conclusion

This study highlights the importance of applying a systems approach to understand the complex interplay of C.Fs in endoscopy-related PSIs resulting in severe harm or death. The SEIPS framework revealed that factors related to people and teams, organization, and tasks were the most common contributors to these incidents. Addressing these factors through targeted interventions could potentially mitigate future incidents and improve patient safety in endoscopy services.

References

1. Carayon P, Schoofs Hundt A, Karsh B, Gurses AP, Alvarado CJ, Smith M, et al. Work system design for patient safety: the SEIPS model. Quality & safety in health care 2006 Dec;15(suppl 1):i50-i58.
2. Frith KH. Medication errors in the intensive care unit: literature review using the SEIPS model. AACN advanced critical care 2013 Oct 1,;24(4):389-404.

Disclosure

This project is funded by grants from the Medical Protection Society Foundation and the Midlands Gastroenterology Society.

Introduction

Patient Safety Incidents (PSIs) are not uncommon in gastrointestinal endoscopy, arising from complex interactions between factors at all levels of healthcare. Understanding the contributory factors (CF) to PSIs is crucial to improve safety, yet are poorly characterised in endoscopy care. This study presents an analysis of CFs associated with PSIs across the UK over 5 years, using a validated Human Factors Framework.

Aims & Methods

This was a retrospective mixed-methods study of anonymised PSI reports associated with at least moderate harm in GI endoscopy, submitted to the National Reporting and Learning System between 2017-2022. A descriptive analysis of cases was undertaken to identify incident type, phase of care and urgency. The Systems Engineering for Patient Safety (SEIPS 2.0) framework was used to identify CFs from reports, based on principles of content analysis.

Results

Of 1,575 reports analysed, 133 were excluded as unrelated to GI endoscopy. Of those remaining, 100 (6.9%) resulted in death, 268 (18.6%) in severe harm and 1,074 (74.5%) in moderate harm.
The most frequently occurring incident type was ‘known complication’ (n=587, 40.7%), with ‘perforation’ (n=341, 58.1%), ‘bleeding’ (n=94, 16.0%) and ‘missed cancer’ (n= 59, 10.1%) being most common. The second most common incident was ‘follow-up related’ (n=268, 18.6%), largely characterized by ‘delayed follow-up or diagnosis’ (n=152, 53.1%) and ‘lost to follow-up’ (n=84, 29.4%). Issues with endoscopy ‘timing or decision’ was the third most coded incident (n=124, 8.6%). Most PSIs occurred peri-endoscopically (n=642, 44.5%), followed by pre-endoscopy (n=409,28.3%) and post-endoscopy phases (n=392, 27.2%). Elective procedures (n=647,44.8%) had more incidents than emergency (n=245,17%).
Of included reports, 1,057 (73.3%) contained enough information to yield 1,441 CF references captured within SEIPS. Table 1 provides an overview of the three most frequently referenced SEIPS domains, with a non-exhaustive presentation of leading themes. 'Person' CFs, influenced largely by communication issues, was the most significant driver. 'Organisation' constituted over a third of references, driven largely by resource availability and recruitment. Most CF references within 'Task' related to complex tasks, dominated by challenging anatomy or physiology. Of the remaining domains, 'Tools and technology' (n=106, 7.4%) was mostly driven by equipment malfunction (48.1%). The 'External environment' (n=55, 3.8%) was almost exclusively COVID-19 policy changes (96.4%), while 'Internal environment' formed a minority of CFs (n=9, 0.6%).

Conclusion

This novel systems-based analysis of CFs in endoscopy-related PSIs provides critical insight for improving safety. The predominance of issues relating to People and Teams, Organisation and Task highlights key areas for intervention. The next phase of this research includes a content analysis of PSIs associated with mild harm and development of simulations based on findings, aiming to improve safety in endoscopy.

Introduction

Safety in endoscopy is of paramount importance with increasingly therapeutic and complex procedures. Despite safety initiatives, incidents still occur. In order to learn from these, we need to be able to better detect issues. The patient voice has been crucial in understanding healthcare from a different perspective. Specifically, patient reporting has been able to detect safety-related issues, which may not have been identified through other means [1, 2]. This is an area not previously explored in endoscopy.

Aims & Methods

Our aims were to develop a novel tool for patient reporting of safety incidents in endoscopy (PRISE), co-designed with patients. Secondary aims were to identify safety incidents, factors and perceptions.
This study was split into two phases. Phase one involved a literature review of articles related to patient reported incidents, followed by patient focus groups to identify key topic areas to address and survey design ideas. These findings informed development of the PRISE survey tool. A patient panel was created and, through regular refinement meetings with the research team, question items were created.
Phase two was a prospective pilot of the PRISE tool in a single academic centre. The survey was disseminated over an 8 month period through physical and electronic media. Patients were asked to complete this within 2 weeks of their procedure. Outcome measures included the percentage of positive responses (PPR) across question item domains, safety perception scores and number of safety incidents detected. Descriptive and comparative analyses were performed.

Results

In phase one, over 1600 articles were identified from search criteria. Following title, abstract and full text screens, 17 full text articles were included in the literature review. A total of 24 incident topic areas were identified. Three focus groups were conducted with 22 participants. Outcomes were focussed around five themes: question style, content, question items, timing, and dissemination. The PRISE co-design process was conducted over ten sessions. The final survey consists of the following sections: demographics, safety incidents and factors – 24 items, grouped into domains (pre-procedure, post-procedure, unit and staff), safety concerns – a qualitative section, and safety perceptions – a quantitative score.
In phase two, 596 patients responded to the survey. Perceptions of safety were overwhelmingly high (mean rating 9.7/10). The PPR for all domains was > 90%, with the staff domain (items on staff characteristics) rated significantly higher than all others (96.7% PPR; p < 0.001). Four clear PSIs were identified - three related to equipment failure and one regarding wrong bowel preparation. Six patients (1.0%) declared safety concerns. Two patients felt they suffered from a mistake. The five highest negatively rated question items were related to information handover, management of patient concerns during procedure, physical space in the unit, equipment faults during procedure, and unnecessary delays in care respectively. Thematic analysis of comments identified negative themes related to bowel preparation, care process issues, communication, constraints on service, environment, equipment, procedures, and staff.

Conclusion

A patient co-designed pilot prospective survey identified high perceived safety in the endoscopy unit. Staff characteristics were rated highly. Safety incidents were detected as well as areas to focus on for safety improvement.

References

1. Armitage, G., et al., Patient-reported safety incidents as a new source of patient safety data: an exploratory comparative study in an acute hospital in England. Journal of Health Services Research & Policy, 2017. 23(1): p. 36-43.
2. Weingart, S.N., et al., What can hospitalized patients tell us about adverse events? Learning from patient-reported incidents. Journal of general internal medicine, 2005. 20(9): p. 830-6.

Disclosure

HA is Chief Scientific Officer of Pre-emptive Medicine and Health Security at Flagship Pioneering. AD is Executive Chair of Pre-emptive Medicine and Health Security at Flagship Pioneering