Problem 1: The rhetorical harm of ‘never’

The first problem with never events lies in the continued use of ‘never’. Zaslow and colleagues3 observe that many of the patient safety events currently deemed never events are not entirely preventable, making the term a misnomer. The use of ‘never’ to describe these patient safety events states a normative preference, not an empirical frequency. ‘Never’ implies that something should not occur. When that something does happen, it engenders disbelief and outrage. The emotional reaction produced by the term is the point.2 7

Originally, the concept of the never event was a rhetorical strategy employed by a ‘moral entrepreneur’ in the early days of the patient safety movement in the USA. Moral entrepreneurs seek to influence the collective adoption of a norm through the attribution of a label (negative or positive) to a phenomenon. Then, they engage in efforts to spread the label through society with the hope it will change beliefs about the phenomenon.8 The first use of ‘never event’ accompanied the circulation of the 2002 report from the NQF Serious Reportable Events Steering Committee. A firsthand account of the deliberations reveals that Dr Kizer wanted to use the term ‘never events’ in the report but was rebuffed by the Committee.9 Dr Kizer, a well-known healthcare executive who was credited with reforming the performance of Veterans Affairs hospitals, had an intuition that the term never event ‘carried an extra psychological charge’ and believed the key to his past successes came from ‘attention to language’s psychological power’.7

It is notable that in official reports, NQF uses ‘serious reportable events’. The term ‘never event’ only appears on their website and in press releases, suggesting it was not intended to be a rigorous quality measure.

Patient safety advocates acknowledge that ‘never’ and ‘zero harm’ are aspirational, not accurate. As the 2013 Berwick Report states: ‘while ‘Zero Harm’ is a bold and worthy aspiration, the scientifically correct goal is ‘continual reduction’. All in the NHS (English National Health Service) should understand that safety is a continually emerging property, and that the battle for safety is never ‘won;’ rather, it is always in progress’.10 Nevertheless, a host of advocacy organisations and regulatory agencies, including the Leapfrog Group, the Centers for Medicare and Medicaid Services and the NHS have officially used the term ‘never event’. Other groups, like The Joint Commission, use alternative terms such as ‘sentinel event’.

What does it matter if we use ‘never event’ instead of ‘serious reportable event’ or ‘sentinel event’? The emotional resonance produced by ‘never’ does not come without a cost, especially for frontline clinicians.6 It is demoralising to be held to a standard you know is impossible to meet.11

While one could argue that ‘never event’ is another way of expressing the value of primum non nocere (first do no harm), values are not goals. Goals are targets; they must have definable conditions for success. ‘Never event’ is a target—one that is defined by external regulators not frontline clinicians. Even if ‘never event’ was intended as an aspirational concept at first use, the term has undergone a social shift in meaning because of regulatory policies that undercut whatever utility it had to inspire improvement. As a ‘never event’ became a quantifiable metric used to deny payment and be forcibly publicly reported, its purpose transformed from aspiration to judgement and discipline.5

Early on, the visceral reaction produced by the term may have served a useful purpose in driving attention to the patient safety movement. However, as the term has come to reflect a ‘measure and manage’ orthodoxy,12 the ‘never’ now obstructs the goal of promoting safety. Striving for absolute safety has unintended consequences, especially when this impossible goal is tightly coupled with incentives.11 Systems of measurement shape how attention is distributed, the interpretive scripts used to make sense of a problem and interactions between people.13 The measurement, reporting and punitive consequences surrounding ‘never events’ can lead to unintended consequences such as a reluctance to report errors, minimising risk by cherry-picking patients, lack of attention to non-reported events that may be just as harmful and a focus on preventing outcomes rather than improving processes.11 12

Problem 2: Commensuration and the decontextualisation of knowledge about safety

The second problem with never events lies in what the concept obscures. Zaslow and colleagues call for a more precise specification or a ‘unifying thread’ in the definition of never events. This effort is characteristic of commensuration or the social transformation of qualities into quantities that share a metric.13 The process of commensuration that creates a ‘never event’ unites incidents ranging from surgery performed on the wrong body part, pressure ulcers, vascular catheter-associated bloodstream infections, patient falls and suicide under a common metric. Attention is drawn away from forms of heterogeneity that exist among these incidents and redirected to a common relationship that has justified their grouping.

What is the common relationship that unites never events? Ideally, a modifiable condition predictive of never events could be isolated.2 Unfortunately, investigations to determine if the incidence of never events is indicative of unsafe hospital systems or problems with an organisation’s safety culture are contradictory. In the aggregate, these events appear rare and random.14 Based on what we currently know, the common relationship that unites never events is more rhetorical than it is statistical.

In uniting a range of qualitatively different errors into one category, commensuration decontextualises knowledge about these events, stripping away important details relating to the specific circumstances surrounding their occurrence. What Zaslow and colleagues call for is an extensive approach to understanding safety—a focus on studying the properties of an aggregation of a large number of cases.15 While this makes healthcare delivery more legible to external actors who are interested in managing costs, promoting accountability or producing generalisable knowledge, it may have limited relevance to those working on the frontlines of care.12 Clinicians whose collective actions produce safety do not approach the accomplishment of their work with a view of the aggregate—a never event is not a clinically meaningful category in the day-to-day.

Accomplishing safety is a practical and experiential exercise based on judgement, effective teamwork and local knowledge about how things work in specific clinical settings.16 Ironically, efforts to reduce the occurrence of never events require revisualising the heterogeneity that is erased through the use of the term. Efforts to improve surgical safety are different from efforts to reduce healthcare-associated infections. These efforts involve distinct groups of professionals working in different contexts, with different priorities, drawing from a specialised body of knowledge about how to accomplish safe care. That surgery on the wrong body part and vascular catheter-associated bloodstream infection are both categorised as a never event is not meaningful to the actual work it takes to minimise their occurrence. So much in patient safety is contingent and context-dependent. The categories we use to understand, communicate about and learn from error need to meaningfully reflect this reality if they are to be useful.

Problem 3: Countervailing powers and the impossibility of a universal definition

Finally, never events resist a universal definition because the various actors that use them—as well as those that are subjected to their use—all have different and often competing interests. Zaslow and colleagues observe that groups engaged in work on never events ‘have very different structures, purposes, and roles within the healthcare system and their work on NEs (never events) is motivated by very different purposes’.3 But these differences are glossed over and the author suggests that entities such as regulatory organisations, insurance companies, patient safety groups, health systems and professional societies all share a goal of better healthcare so should be able to develop a shared understanding and common definition of never events. Unfortunately, this argument exhibits an assumption of a ‘harmony model of organisational life’ that effaces the nature of power at work in institutions.17 But in fact, the collective negotiation and alignment work involved in defining administrative categories is immense.5 This is especially true when it comes to rapidly changing institutional fields characterised by power asymmetries and resource constraints, such as healthcare.

To understand why, for more than 20 years, there has been no universal definition of never events, it is helpful to centre countervailing power relations between stakeholders within healthcare.18 Efforts to monitor safety have triggered conflicts between physicians and regulatory agencies as they can be seen as challenging professional jurisdiction over expert knowledge and autonomy in clinical practice.19 The original vision of the NQF report on serious reportable events was premised on a market-based consumerist logic—that making data on never events publicly reportable would motivate clinicians to improve care through benchmarks, facilitate competition on the basis of quality and promote consumer choice.1 Non-payment for never events stems from the goal of increasing the healthcare industry’s accountability for performance and controlling costs.7 Patient advocates want to ensure that safety remains a priority, infused with a sense of urgency—calling for transparency and a commitment to improvement and learning.20 While there is an overlap in the interests of these actors, there are many points of divergence that make reaching an agreement extremely challenging.

Conceptual confusion surrounding never events did not appear out of thin air—it is a result of the negotiation between actors with different interests in a complex and dynamic environment. Medical errors are essentially contested—while there is agreement that error is bad and should be minimised, there is often disagreement about what causes it, what should be done to fix it and how these efforts should be balanced against competing priorities.6