Sage Journals: Discover world-class research

Abstract

Background

There are benefits to healthcare from reporting and learning from near misses in patient care. However, there have been longstanding issues with identifying near misses, with variation in definitions. Learning is being lost, unlike in other industries that have harnessed their learning potential. The features of a healthcare near miss have never been described nor modelled. This study aimed to identify those features to support near-miss identification, reporting and learning.

Methods

This study took a mixed-methods approach with participants from healthcare and four high-reliability industries – aviation, maritime, nuclear and rail. Qualitative exploration helped identify the features of a near miss, while quantitative supported assessment of agreement on features between participants through the use of a scenario.

Results

Participants from 17 healthcare and 35 industry organisations took part. Quantitative findings demonstrated variation in agreement of the features of a near miss using Fleiss Kappa. Qualitative findings identified the following themes in relation to the features of a near miss – context dependent, involve control, are complex and represent vulnerabilities. In particular, several industries have lists of specific situations that constitute near misses that support reporting and focus.

Conclusion

Without clear agreement of the features of a healthcare near miss, definitions will continue to vary. This study has, for the first time, provided exploration and clarification of the features of a near miss with the offer of a healthcare model for future validation. Without addressing the fundamentals, such as agreeing what a near miss is, healthcare cannot hope to learn from them.

Keywords

Near miss healthcare definition mixed methods

Background

For over 20 years, healthcare policy documents and the research literature have sought to encourage the reporting and analysis of near misses in patient care.^1–3 However, progress to harness their learning potential has been slow.^4,5 One fundamental barrier to the reporting of, and therefore learning from near misses is that their definition is not agreed nor understood.^5–7

At its simplest, a near miss may be seen as something that almost happened.⁸ However, when exploring definitions of a healthcare near miss, there is significant variation.^4–7 Variation relates to the features of a near miss, meaning one organisation's near miss may be different to another. In the National Health Service (NHS), for example, national definitions vary between bodies, and between national bodies and providers.⁵

Along with the statements of the need for healthcare to learn from near misses is a recognition that other industries, such as aviation, have harnessed their learning.^1,2,9 The process and transportation industries have long had clarification of the features of a near miss following their exploration in 1989 in Eindhoven (Netherlands).¹⁰ The result was the ‘Eindhoven Model’ (Figure 1).¹¹

Figure 1.

The Eindhoven Model.¹¹.

The Eindhoven Model was published some 30 years ago, but no similar model exists in healthcare despite some attempts to apply it.^12,13 The features of a healthcare near miss have never been formally described which is contributing to variation in definitions.⁵ The authors therefore set out to seek insights from healthcare organisations and non-healthcare high-reliability industries to inform description of the features of a healthcare near miss for an evidence-based model.

Methods

This study took a mixed-methods approach – qualitative exploration of the features of a near miss, and assessment of agreement on features between participants.

Qualitative component

Healthcare participants were approached from sectors representative of the English NHS – general practice, secondary care (acute hospital, mental health and ambulance) and regional/national bodies (regulatory and policy). For non-healthcare participants, representatives of ‘High Reliability Industries’ (HRIs) were approached. HRIs have rich reporting and learning cultures which include near misses.¹⁴ Due to the range of HRIs, it was pragmatically chosen to focus on aviation, maritime, rail and nuclear.

Participants were safety leads purposively approached via organisational websites to take part in semi-structured interviews. The interview protocol (Supplemental material) was developed with nine safety professionals. One researcher undertook interviews to support consistency. Each interview was audio recorded, transcribed and thematically analysed using a recognised approach.¹⁵ Two researchers undertook the analysis independent of each other. Themes were agreed amongst the researchers.

Quantitative component

To support exploration of the features of a near miss, a ‘scenario’ was developed, inspired by Henneman and Gawlinski.¹³ The scenario (Table 1) was tested during protocol design. It was shown to participants by asking whether they term each outcome as an incident, near miss, non-event or something else (depending on their industry). Discussion of responses contributed to the qualitative data. To quantify agreement between participants Fleiss’ Kappa was used.

Table 1.

The scenario and its associated outcomes (O1–O4).

Scenario
The patient has been admitted with a suspected heart attack. Unknown to the doctor, they are planning to prescribe a medication that the patient is allergic to. The four potential outcomes are:
Outcomes
O1	Doctor prescribes (paper chart), and the nurse administers the medication. No checks are done, and the patient has a harmful reaction.
O2	Doctor prescribes (paper chart), and the nurse administers the medication. No checks are done, and the patient has no reaction.
O3	Doctor prescribes (paper chart) and the nurse approaches to the patient to administer the medication. The nurse checks with the patient whether they have any allergies. The medication is not administered.
O4	Doctor attempts to prescribe (electronic prescribing system) the medication. The system checks the patient's records and prevents prescription as the allergy is identified.

Enhancing trustworthiness

Research approvals were granted prior to commencement. Relevant standards were followed to support rigour.¹⁶ Trustworthiness was created through the protocol, triangulation of perspectives and by reaching theme saturation. There was no predetermined sample size as this was predominantly qualitative. Draft themes/findings were shared with participants for verification.

Results

Interviews were undertaken between 2019 and 2022. Twenty-one healthcare participants took part across 17 organisations – acute hospitals (n = 3), general practice (n = 6), mental health (n = 2), ambulance services (n = 2), regional bodies (n = 2) and national bodies (n = 2). Thirty-six HRI participants took part across 35 organisations – aviation (n = 12), maritime (n = 7), nuclear (n = 7) and rail (n = 9). Further detail is available as Supplemental material.

Quantitative findings

Fourteen healthcare participants responded to the scenario. Fleiss Kappa showed perfect agreement that O1 was an incident (ĸ = 1.00, 95% CI [0.88, 1.00]); fair agreement for O2 as an incident (n = 9) or near miss (n = 5) (ĸ = 0.26, 95% CI [0.14 to 0.38]); fair agreement for O3 as a near miss (n = 10) (ĸ = 0.29, 95% CI [0.17, 0.41]); and slight agreement for O4 as a near miss (n = 7) or non-event (n = 6) (ĸ=0.09, 95% CI [0.00 to 0.21]).

Twenty-two HRI participants responded. Fleiss Kappa showed perfect agreement for O1 as an incident (ĸ = 1.00, 95% CI [0.93, 1.00]), fair agreement for O2 as an incident (n = 14) (ĸ = 0.27, 95% CI [0.20, 0.35]), fair agreement for O3 as a near miss (n = 14) (ĸ = 0.23, 95% CI [0.15, 0.31]) and very good agreement for O4 as a non-event (n = 21) (ĸ = 0.86, 95% CI [0.78, 0.94).

Qualitative findings

All participants contributed data to the qualitative analysis. Table 2 provides a summary of the themes with example quotes. Themes are explored in turn below.

Table 2.

Themes identified and example quotes.

Near misses…	Example quotes
… depend on context – Operational definitions – Difficulty with generic definitions	‘Operational near misses might well be at a level crossing…’ Rail
	‘… It is certainly confusing for staff… To me it is a woolly definition’. Mental Health
… involve control – Controls prevent progression – Control effectiveness varies	‘… something that is almost consequential, but it was realised that it was wrong and reacted to..’. Nuclear
	‘[scenario O4]… you have high reliability, and the system works, it captures everything you expect it to’. Aviation
… are vulnerabilities – Representing systemic issues – Reframed as positives	‘If someone said we had a near miss, which would be a very serious incident, complete loss of control, near accident..’. Aviation
	‘We would actually want to use that to look at events that went well’. Ambulance
… vary in complexity – Representing linear events versus complex events	‘We split our near misses… High potential ones would be where there is an uncontrolled event…’ Rail

Near misses depend on context

This theme represents how a near miss is dependent on the industry in which it occurs. HRI near misses were described in relation to operations, such as driving a train. Several participants shared lists of specific situations that constitute a near miss. These ‘operational’ near-miss definitions often include detail of the actors (e.g. ship), situation (e.g. close proximity) and potential outcome (e.g. collision). Some of these definitions had specific names, such as Airprox.¹⁷

Healthcare examples commonly related to situations involving medication. Participants described that these are easier to conceptualise. Definitions given, whether an individual's or organisation's, were often generic without reference to actors or situation. The generic nature of healthcare definitions made them difficult to apply.

Near misses involve control

This theme demonstrates the role of controls in near-miss situations. A control represents something that prevents the progression of events. Controls were also referred to by other terms, such as barriers.

When exploring the scenario, participants who responded to O3 and O4 as near misses or non-events qualified their responses with reference to controls – the nurse (O3) and prescribing system (O4). O2 led to debate about ‘luck’ in a near miss and whether this is a control.

Several HRI participants felt that the type of control is relevant to the definition. The human control in O3 was thought to represent a greater risk of future incident, than the system control in O4. The rationale was that a prescribing system may have functionality to prevent incidents, while a human is fallible. O4 was therefore seen as a non-event by most HRI participants.

Near misses vary in complexity

This theme represents how near misses involve different types and numbers of factors and controls. Most participants described near misses as linear with a single control, much like the scenario. However, several HRI participants, particularly from aviation and nuclear, challenged the linear representation when multiple factors and controls are likely to be involved.

The presence of the patient in the scenario led to differing views as to whether a near miss could still occur if events reached them. This was also demonstrated by the terming of O2. Some thought events reaching a patient must mean an incident has occurred, without harm. However, others argued that the no-harm outcome was the near miss. No participant termed O1 a near miss, but some questioned whether this is a near-miss death.

Near misses are vulnerabilities

This theme recognises that near misses represent risk of future incidents and need reporting. Several healthcare participants described the need to turn near misses into ‘positives’ to change the culture around them and celebrate the staff involved. In contrast, HRI participants described how their industries saw near misses as ‘vulnerabilities’. These vulnerabilities were of greatest concern when a human needed to be involved or where multiple controls had been overcome.

Discussion

Learning from near misses has the potential to support improvements in patient safety. However, without a clear, agreed and usable definition, near misses will continue to go unrecognised.^5–7 The authors set out to explore the features of a healthcare near miss to inform an evidence-based model. Those features are explored below, aligned with the literature.

Feature 1 – context-specific situations

The generic nature of definitions leads to ambiguity and difficulty identifying situations as near misses. As a result, learning is lost that would otherwise provide insights into contributory factors to incidents and activated controls to prevent those incidents.^4,5

To address the challenge of ‘missing’ the near miss, several HRIs have listed specific situations that meet criteria for a near miss.^18–21 While some argue that lists limit the breadth of reporting,²² they can support focus on significant situations.^19,23 Specifying what should be reported may be beneficial in industries such as healthcare where underreporting is problematic.²⁴

The specific situations meeting HRI criteria also show that a near miss in one industry is different to another. While operating contexts (e.g. plane versus ship) will contribute to differences, so will the angle with which the situation is viewed. For example, ‘pilot incapacitation’¹⁸ was described as a near miss by several HRI participants, but could be argued an incident from the perspective of the pilot. One person's (or plane's) near miss may be someone else's incident. A near miss is a situation defined by the context in which it occurs and their safety focus.²⁵

Feature 2 – complex human-based control systems

The number, nature and effectiveness of controls were found to be important features of a near miss. Near misses occurring in complex systems are likely to be the result of multiple factors and therefore multiple controls may be needed. Very rarely is safety dependent on a single decision or action.²⁶ Viewing near misses as single cause and control may limit learning by being too simplistic.^27,28 There is value in understanding the range of controls in a near miss, including any that were unsuccessful. In HRIs such as nuclear, events that overcome some controls before being stopped represent higher-risk near misses.²⁹

Controls will differ in their effectiveness. For example, those that engineer out a risk are likely to be more effective than relying on a human to follow a safety rule.³⁰ If luck is considered a control, it will be weak control due to being unplanned.³¹ The HRI terming of O3 and O4 demonstrates the differing effectiveness of controls. The nurse following a procedure will not always be reliable in preventing an incident, while a designed system with forcing functionality is more likely to. The terming of O4 as a non-event by HRI participants also highlights an interesting difference in perceptions between them and healthcare participants. This may be representative of differences in safety understanding and maturity across the industries.¹⁴ As described by one participant, ‘healthcare always needs to find someone to blame’. It may also be because HRI participants perceived prescribing systems to be ‘error proof’ when in reality their control functionality varies depending on the system.

This study found no theme around the proximity of controls to the point of incident for a situation to be defined a near miss. Participants varied from those who thought a near miss is where a control acts just prior to the point of incident (as per O3), to others who thought any control no matter when/where makes a situation a near miss. This debate is also had in the literature with no evident consensus.^32,33 The lack of agreement may again highlight how near misses are dependent on the context in which they occur. More mature industries may seek reporting of near misses with more proximal control, while less mature, such as healthcare, may focus on near misses with controls just prior to the point of incident.

Feature 3 – include the patient

The patient is an important feature of a healthcare near miss. However, there was no consensus about whether events could reach a patient or not and be considered a near miss. The literature highlights a similar debate,⁶ with examples of situations termed near misses that have harmed patients.³⁴ Several participants agreed that events reaching a patient means harm is possible, and so an incident must have occurred. From a patient-centred perspective this seems reasonable, and is also the interpretation held by almost 4000 participants of a survey undertaken by the Institute or Safe Medication Practices.³⁵

The debate about the position of the patient in relation to events may be academic, as long as all situations are reported. However, there is known underreporting of safety events in healthcare, and even when reported learning is superficial.^4,5 If near misses are not recognised as such, there is a risk that efforts to learn will only focus on contributory factors, without attention to the controls involved.

Feature 4 – represent vulnerability

This study found differences in how near misses are viewed between HRIs and healthcare. Healthcare participants were keen to reframe near misses as positive events to encourage their reporting. Their rationale was that this reduces the fear of blame, aligns with efforts to learn from ‘what went well,’ and celebrates staff for providing resilience. The literature shows examples of near misses being reframed as ‘good catches’.³⁶

However, HRI participants and literature suggest risk with the positive reframing of near misses.^37,38 Research cautions viewing near misses as evidence of resilience because it can inhibit future actions, resulting in riskier decisions.^39,40 Celebrating the human may place an over reliance on their role in catching events, when the situation represents system vulnerability by the very fact a human was required. It is therefore important for near misses to be seen as representing vulnerability rather than resilience.³⁹ Rather than healthcare reframing near misses, it may be more appropriate to address the underlying reasons for why this is thought to be needed, such as punitive cultures.

Updating the Eindhoven Model

The findings offer an opportunity to translate and update the Eindhoven Model for healthcare, with acknowledgement of how safety science has evolved over the past 30 years. The original model is linear with two controls described – ‘defences’ and ‘recovery’. Defences were defined as the safety rules, training and equipment that bring systems back to normal states, while recoveries were final human controls.^10,11

Much of the original Model is still relevant, however, the findings of this study challenge the original's inclusion of human involvement as a defence (whether by training or following safety rules).¹¹ This study suggests that any involvement of a human to prevent progression of events represents a system vulnerability and therefore a near miss. As systems have become more complicated and human fallibility understood, so too has our understanding of effective of controls. The proposed model therefore refers to barriers which are robust and reliable controls that can be relied upon.⁴¹

The authors propose an updated Eindhoven Model for healthcare (Figure 2). This provides a representation of a healthcare near miss for future research. It includes contextualisation to healthcare with inclusion of the patient; recognition of complexity and their representation of vulnerability; and recognises the differences in effectiveness of controls.

Figure 2.

The updated Eindhoven Model for healthcare.

Strength and limitations

This study has the potential to support improvements in patient safety through learning from near misses. The features were developed from a comprehensive exploration of primary and secondary data. While not all HRIs and healthcare sectors were included, theme saturation was reached. It is acknowledged that defining a near miss is only one part of a complex process for effective learning, and that a model can never represent the complexity of healthcare and the fact that some safety events are emergent.⁴² Further research is required to validate the model and to better understand how learning can best be achieved. A future study may also seek to quantify the reliability of different controls to prevent incidents to help truly distinguish the distinction between a near miss and non-event.

Conclusion

Healthcare definitions for a near miss vary, with underreporting and loss of learning. Until now the features of a healthcare near miss have not been defined or modelled, unlike in other industries. With better recognition and reporting of near misses, these situations have the potential to provide insights into the factors that contribute to incidents, and the availability and effectiveness of controls. It may be beneficial for future research to validate the provided model and to look to develop specific definitions for healthcare situations that represent near misses, using the features found in this study.

Supplemental Material

sj-docx-1-cri-10.1177_25160435241247096 - Supplemental material for Updating Eindhoven: Clarifying the features of a patient safety near miss

Supplemental material, sj-docx-1-cri-10.1177_25160435241247096 for Updating Eindhoven: Clarifying the features of a patient safety near miss by Nick Woodier, Charlotte Burnett and Paul Sampson, Iain Moppett in Journal of Patient Safety and Risk Management

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Ethical approval

The Health Research Authority's Integrated Research Application System provided approval (262065) on 7 May 2019. An amendment was approved for additional healthcare sites on 10 December 2019. The University of Nottingham's Research Ethics Committee provided approval (264-1903) on 15 March 2019.

Funding

The authors received no financial support for the research,authorship,and/or publication of this article.

ORCID iD

Nick Woodier

Supplemental material

Supplemental material for this article is available online.

References

Kohn

Corrigan

Donaldson

(eds). To err is human: Building a safer health system. Washington: National Academy Press, 2000.

Department of Health. An organisation with a memory: report of an expert group on learning from adverse events in the NHS chaired by the Chief Medical Office. UK: Department of Health, 2002.

Leape

. Error in medicine. JAMA 1994; 272: 1851.

Feng

T-T

Zhang

Tan

L-L

, et al. Near miss research in the healthcare system: a scoping review. J Nurs Adm 2022; 52: 160–166.

Woodier

Burnett

Sampson

, et al. Patient safety near misses – Still missing opportunities to learn. J Patient Saf Risk Manage 2023; 29: 47–53.

Nation

Dooley

. Multiplicity of medication safety terms, definitions and functional meanings: when is enough? Qual Saf Health Care 2005; 14: 358–363.

Marks

Kasda

Paine

, et al. “That was a close call”: endorsing a broad definition of near misses in health care. Jt Comm J Qual Patient Saf 2013; 39: 475–479.

Capucho

. Near miss: almost error or potential adverse event? Rev Latino-Am. Enfermagem 2011; 19: 1272–1273.

Barach

Small

. Reporting and preventing medical mishaps: lessons from non-medical near miss reporting systems. Br Med J 2000; 320: 759–763.

10.

van der Schaaf

(ed). Near Miss Reporting as a Safety Tool. Burlington: Elsevier Science, 1991.

11.

van der Schaaf

. Near miss reporting in the chemical process industry. Eindhoven: Technische Universiteit Eindhoven, 1992.

12.

Kanse

van der Schaaf

Vrijland

, et al. Error recovery in a hospital pharmacy. Ergonomics 2006; 49: 503–516.

13.

Henneman

Gawlinski

. A “near-miss” model for describing the nurse’s role in the recovery of medical errors. J Prof Nurs 2004; 20: 196–201.

14.

The Health Foundation. High Reliability Organisations: Evidence scan 2011, https://www.health.org.uk/sites/default/files/HighReliabilityOrganisations.pdf (2011, accessed 23 January 2024).

15.

Braun

Clarke

. Using thematic analysis in psychology. Qual Res Psychol 2006; 3: 77–101.

16.

O’Brien

Harris

Beckman

, et al. Standards for reporting qualitative research: a synthesis of recommendations. Acad Med 2014; 89: 1245–1251.

17.

SkyBrary. AIRPROX, https://skybrary.aero/articles/airprox (2021, accessed 25 January 2024).

18.

International Civil Aviation Organisation. Annex 13 to the Convention on International Civil Aviation: Aircraft Accident and Incident Investigation. Montréal, Quebec, Canada: International Civil Aviation Organisation, 2016.

19.

Nesmith

Keating

Zacharias

. Investigating process safety near misses to improve performance. Proc Safety Prog 2013; 32: 170–174.

20.

Jobush

. Identifying near miss reporting criteria for the Gilbert Fire Department. Gilbert, Arizona: National Fire Academy Executive Fire Officer Program, 2005.

21.

International Atomic Energy Agency. Nuclear Power Plant Operating Experience from the Iaea/Nea International Reporting System for Operating Experience 2015–2017. Vienna, Austria: International Atomic Energy Agency, 2020.

22.

Gnoni

Saleh

. Near-miss management systems and observability-in-depth: handling safety incidents and accident precursors in light of safety principles. Saf Sci 2017; 91: 154–167.

23.

Hamilton

Pham

Minzenmayer

, et al. Are we missing the near misses in the OR?-underreporting of safety incidents in pediatric surgery. J Surg Res 2018; 221: 336–342.

24.

Mandal

Adams

Fraser

. “Near misses” in a cataract theatre: how do we improve understanding and documentation? Br J Ophthalmol 2005; 89: 1565–1568.

25.

Fieder

Wartenburger

Abdel Rahman

. A close call: interference from semantic neighbourhood density and similarity in language production. Mem Cognit 2019; 47: 145–168.

26.

Rasmussen

. Risk management in a dynamic society: a modelling problem. Safe Sci 1997; 27: 183–213.

27.

Thoroman

Goode

Salmon

. System thinking applied to near misses: a review of industry-wide near miss reporting systems. Theor Issues Ergon Sci 2018; 19: 712–737.

28.

Thoroman

Goode

Salmon

, et al. What went right? An analysis of the protective factors in aviation near misses. Ergonomics 2019; 62: 192–203.

29.

International Atomic Energy Agency. Defence in Depth in Nuclear Safety: A report by the International Nuclear Safety Advisor Group. Vienna, Austria: International Atomic Energy Agency, 2015.

30.

National Institute for Occupational Safety & Health. Hierarchy of Controls, https://www.cdc.gov/niosh/topics/hierarchy/default.html (2023, accessed 25 January 2024).

31.

Sheikhtaheri’

. Near misses and their importance for improving patient safety. Iran J Public Health 2014; 43: 853–854.

32.

Sheridan

Cardosi

Hannon

. Rating the severity of close-call events. Proc Hum Factors Ergon Soc Annual Meeting 2004; 48: 2050–2054.

33.

Ginsburg

Chuang

Y-T

Richardson

, et al. Categorizing errors and adverse events for learning: a provider perspective. Healthc q 2009; 12: 154–160.

34.

Moreno

Zuberi

. Characterization of medication-related near miss safety events. Can J Hosp Pharm 2019; 72: 64–98.

35.

Institute For Safe Medication Practices. ISMP survey helps define near miss and close call, https://www.ismp.org/resources/ismp-survey-helps-define-near-miss-and-close-call (2009, accessed 22 January 2024).

36.

Wallace

Mamrol

Finley

. Promote a culture of safety with good catch reports. Pa Patient Saf Advis 2017; 14.

37.

Barnes

Gruntfest

Hayden

, et al. False alarms and close calls: a conceptual model of warning accuracy. Wea. Forecasting 2007; 22: 1140–1147.

38.

Tinsley

Dillon

Cronin

. How near-miss events amplify or attenuate risky decision making. Manage Sci 2012; 58: 1596–1613.

39.

Dillon

Tinsley

Burns

. Near-misses and future disaster preparedness. Risk Anal 2014; 34: 1907–1922.

40.

Dillon

Tinsley

Cronin

. Why near-miss events can decrease an individual’s protective response to hurricanes. Risk Anal 2011; 31: 440–449.

41.

Chartered Institute of Ergonomics & Human Factors. Barrier Management. https://archived.ciehf.org/learn/barrier-management/ (2016, accessed 25 January 2024).

42.

Woods

Hollnagel

Woods

, et al. Resilience engineering: concepts and precepts. 1st ed. Boca Raton: CRC Press, 2017.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.03 MB

0.00 MB