The Bystander effect and diffusion of responsibility

When everyone is responsible, no one is responsible.9

Each high-profile inquiry cites a ‘lack of effective…leadership’,7 confusion over job responsibilities, and complex and unclear accountability arrangements (locally and nationally) as significant causal factors leading to breaches in patient care. Essentially, employees were unclear regarding who had responsibility for what. In most cases, that there were significant problems with the control of infection,5 ,8 inadequate standards of patient care,6 alarmingly high mortality rates,7 and concerns about governance and staffing,6 ,7 was already known, but why did nobody—Board members, clinicians or nurses—intervene to clarify responsibilities, or pursue their concerns with vigor? Empirical studies have shed light on the psychological mechanisms underlying human inaction in a critical situation.10 The bystander effect refers to the phenomenon that an individual's likelihood of intervening decreases when passive bystanders are present.11 ,12 Research shows the bystander effect to be a robust phenomenon, observed across many domains (eg, serious emergencies13). Significant moderator variables include the number of bystanders (higher numbers leads to less help), the ambiguity of the situation (high ambiguity leads to less help) and the similarity of the bystander to the victim (greater similarity leads to more help14).

While there are no studies examining the bystander effect in a clinical context, staff behaviour in such a setting may be readily explained by this psychological phenomenon. First, the large number of staff employed in a main hospital such as Stoke Mandeville, Ysbyty Glan Clwyd (YGC) in North Wales, or Mid-Staffordshire Hospital, made it less likely that any individual member would take action. Second, the Kennedy8 and Duerden5 reports state there were conflicting messages regarding healthcare-associated infection (HCAI) statistics in YGC and Mid-Staffordshire, creating a high-ambiguity situation, which according to bystander theory inhibits human intervention. Finally, the extent to which hospital employees identify with their patients, much as bystanders do (or do not) with victims, may contribute to the passivity of hospital staff. For example, the commodification of patients in modern healthcare settings,2 and the overemphasis on targets and bottom-line achievements is likely to breed depersonalisation. Patients become anonymous thereby cultivating an environment where staff are able to psychologically distance themselves from their patients. Under such circumstances, the process of becoming a bystander is likely facilitated.

A psychological process related to the bystander effect—diffusion of responsibility11—seems particularly relevant to the context outlined in a number of inquiries. Diffusion of responsibility refers to the tendency to (subjectively) divide the personal responsibility to intervene in a critical situation by the number of bystanders—the more bystanders, the less personal responsibility any individual bystander will feel. Similarly, the individual bystander will only feel responsible for a portion of the cost to the victim associated with non-intervention. Consequently, diffusion of responsibility has been used to explain empirical findings showing that members of a group feel less responsible for negative consequences11 ,13 compared with when acting alone (ie, high responsibility conditions). Using diffusion of responsibility may explain why nobody intervened in infection control matters in YGC or Stoke Mandeville; it was assumed that someone else would take responsibility. Additionally, due to the number of staff in a busy main hospital, even with the knowledge that the infection prevention and control (IPC) service was weak, staff were likely to feel little responsibility for this.

Aversive control and negative reinforcement

A person who has been punished is not…simply less inclined to behave in a given way; at best, he learns how to avoid punishment.16

Behavioural theory posits that all human behaviour is learnt through interaction with our environment. Decades of psychological experimentation have demonstrated that when placed in an environment characterised by the presence of an aversive stimulus (eg, an electric shock), avoidance behaviour is learnt (eg, an animal will learn to successfully avoid an electric shock by performing a specific behaviour, such as running from one side of a compartment to another). According to behavioural theory, avoidance learning occurs, and is maintained through a negative reinforcement contingency.ⁱ Because avoiding the aversive stimulus becomes the primary motivator for behaviour, any response that leads to the successful removal or termination of the aversive stimulus is more likely to occur in the future.

Akin to animals in Skinner boxes,ⁱⁱ hospital staff behaviour in these hospitals was governed by aversive contingencies—fear of blame, vilification and subsequent public shaming, all operated to suppress unwelcome news, including the true incidence of HCAI or mortality rates. As such, professionally questionable behaviours, such as misreporting key statistics, become negatively reinforced; such behaviours served to avoid the negative consequences associated with the truth. Thus, the lack of clinical transparency so detrimental to patient safety in these hospitals can be explained through the process of negative reinforcement. Moreover, it is likely that the suppression of bad news was negatively reinforced across all levels of the organisation (from ward to Board), whereby members of the Board were also negatively reinforced to avoid censure, which may have conflicted with a desire for transparent and timely reporting of statistical data.

Exhortations for an ‘end [to] NHS blame games’,17 to be replaced by ‘a culture of openness and learning from mistakes…’,17 would indeed change the current environmental contingencies operating in the NHS, such that staff would no longer be negatively reinforced for misreporting or taking comfort from inaccurate figures. However, the emphasis on meeting targets in modern healthcare systems, and the negative consequences associated with not doing so, may preclude such a change in culture. Indeed, until this changes, the recent introduction of ‘whistleblowing’ procedures in the NHS are unlikely to effect any change. Thus, according to behavioural theory, where aversive control is used to manage a system, inevitably, self-preservation (ie, avoidance of blame and shame), through negative reinforcement, becomes an organisational norm.

Learned helplessness

Learned helplessness,18 describes the finding that animals, when exposed to uncontrollable electric shocks, show (A) reduced behavioural initiation (ie, passivity), (B) deficits in subsequent escape/avoidance learning and (C) emotional stress. In one of their earliest experiments, Seligman and Maier,18 examined the effects of escapable shocks (eg, pressing a button terminates shock) compared with inescapable shocks (eg, pressing a button did not terminate shock) on subsequent escape and avoidance learning in dogs. Results showed that dogs in the escapable shock condition learnt to avoid shock by pressing a button; that is, reinforcement (ie, termination of the shock) was contingentⁱⁱⁱ upon their response (ie, pushing the button). Conversely, dogs in the inescapable shock condition eventually ceased responding. Compared with dogs in the ‘escapable’ condition, these dogs learnt that their environment was uncontrollable—button pressing was independent of the shocks they received. In further experiments, dogs that had been exposed to inescapable shock subsequently failed to learn shock-termination responses when placed in escapable shock conditions.19 Seligman and Maier18 argued that the mechanism responsible for impairments in learning was the non-contingent relationship between responding and reinforcement—if a person learns that their responses have no effect upon subsequent reinforcement in their environment, they may display behaviours associated with learned helplessness, such as passivity, detachment and apathy.

While learned helplessness theory has typically been applied to explain the development and maintenance of psychopathology in clinical populations, it has more recently been applied to organisations.20 For example, a recent study21 revealed significant relationships between self-reported learned helplessness and employee perceptions of procedural and distributive justice, emotional exhaustion and cynicism in a sample of 217 banking employees. Earlier models of helplessness in the workplace22 suggest it leads to negative attitudes and behaviours, such as job dissatisfaction and turnover intentions.

Many factors highlighted in each of these inquiries suggest learned helplessness had become a significant factor in these hospitals. The detachment, apathy and passivity of staff was clear—general disengagement from key IPC audits (hand hygiene, environmental cleaning, intravenous line care, etc), disenfranchised clinicians reluctant to take on senior managerial roles, high levels of sickness absence and a withdrawal from basic patient care—patients were not fed, were left in soiled bedding and call bells went unanswered. As Francis6 noted, many factors such as lack of resources and an uncompromising focus upon targets make it difficult to provide appropriate care. These factors may operate to ‘numb’ staff into believing that nothing can ever be done, lapsing into ‘fatalistic acquiescence’.6 This ‘numbing’ process may also take hold when employees feel not listened to—a feature common to all of these inquiries. Indeed, the non-contingency between staff behaviour through the repeated raising of concerns—be they regarding mortality rates, unsafe practices or high HCAI figures—and environmental outcomes (eg, Boards failing to respond to these concerns in a timely manner) in these hospitals potentially precipitated and maintained learned helplessness—staff had tried, but failed to be heard. Unfortunately, it isn't hyperbole to state that staff had tried for years to raise concerns regarding poor care in their workplaces, only to be met with silence or ‘threats and humiliation’8 from senior managers, which ultimately may have led to the perception of uncontrollability and learned helplessness.