NHS Improvement's Just Culture Guide: good intentions failed by flawed design

Healthcare is a risky business, whenever and wherever it is practised. The Code of Hammurabi (ca. BC 1750) identified poor clinical outcomes considered avoidable and recorded the punishments to be administered to physicians and surgeons when they occurred. ¹ A gravestone in the Shetland Islands records an early medication error, the death in 1848 of Donald Robertson, whose apothecary ‘sold him nitre instead of Epsom salts by which he was killed in the space of 5 hours after taking a dose of it’. ² Now, 10% of NHS inpatient episodes lead to harmful adverse events, around half of them preventable, ³ and medical error has been estimated to be the third leading cause of death in the United States. ⁴

Many enterprises carried significant risk in their early days, including, most obviously, heavy manufacturing, engineering, mining and chemical industries, and transportation by land, sea and air. Almost all of these enterprises are now characterised by their safety, but medicine has singularly failed to achieve the same transformation. There is simply no doubt about what should happen: every time something goes wrong (or there is a near-miss), it should be investigated to discover why, by examining systems and human factors, and how improvements can be made to reduce the risk of recurrence.^3,5–8 This is how other enterprises have transformed their safety record; why has healthcare lagged behind?

Clinicians have high expectations of themselves and of each other. Despite the obvious ubiquity of human error, nobody likes to think they have erred, and the stakes are higher when somebody may be harmed, perhaps seriously or fatally, as a result. But there is also an undercurrent of professional culture that is based on the notion of infallibility and characterised by intolerance of any deviation from it. ⁹ This is not taught in any lecture theatre, but may be seen in the day-to-day behaviour of leaders and role models. In consequence, the predominant reaction of a clinician to a safety incident that signals departure from this notional ideal is a sense of deep shame and embarrassment that makes it difficult to be open about what has happened. ¹⁰

Adding to this already personally threatening situation, clinicians face the likelihood of blame, both publicly and from health organisations. Underlying the public and media reaction is the human characteristic that blaming individuals is emotionally more satisfying than targeting institutions, ¹¹ but the tendency of employers and professional regulators to pursue technical notions of individual competence rather than systems and human factors is less easy to understand. Some recent cases of criminal prosecution for gross negligence manslaughter, although relatively uncommon, have ‘led to an increased sense of fear and trepidation, creating great unease within the healthcare professions’. ¹²

It is, then, no surprise that clinicians face significant personal difficulty in being open when errors have occurred, nor that this may lead to a desire to find ways to gloss over what has happened and explain it away. Fear of blame is a principal reason not to report clinical incidents. ¹³ The consequence is profound: Lucian Leape has said that the single greatest impediment to error prevention in the medical industry is ‘that we punish people for making mistakes’. ¹⁴

One response has been the move by NHS Improvement to introduce the Just Culture Guide, its aim to ‘treat staff fairly, making them confident to speak up when things go wrong rather than fearing blame’. ¹⁵ The intention of this policy is laudable; the execution lamentable. The problem arises from the disconnect between the logical sequence set out in the guide, a series of questions to identify a series of possible causes of a safety incident, and the relative frequency with which they occur.

It is difficult to be precise about the contribution of individual acts of negligence or deliberate harm to preventable morbidity and mortality, because safety incidents are not reported that way. ¹⁶ All of the major contributors to knowledge on patient safety, however, are clear that individual acts are responsible for only a small minority of the incidents that occur.^3,6,11 Some are associated with clinicians who are unwell, ¹⁷ but very few are due to acts of deliberate harm. ¹⁸

Yet the first question in the Just Culture Guide asks whether the clinician concerned intended to cause harm. Applied to all of the safety incidents estimated to occur to NHS inpatients every year, this would mean asking clinicians involved in around one million incidents if they intended to cause harm, when we know that the number of true instances of deliberate harm would probably not even reach double figures. And if the answer to the first question is negative, the next questions ask whether the clinician is misusing drugs or is physically or mentally unwell.

These are intrusive and challenging questions even in prospect. Far from making staff confident to speak up when things go wrong, the approach could hardly be better designed to increase their fear of blame. However logical the notion of excluding each possibility in turn, it is perverse in practice and likely to have the opposite effect to that intended. It is not until the very last question, which asks whether there were any significant ‘mitigating factors’, that the guide enters the territory where the overwhelming majority of incident causes lie. The Just Culture Guide is, to all intents and purposes, upside down.

It may be argued that the guide is not intended to be applied in strict sequence, but it is very clearly set out as a sequential series of tests. It may be argued that clinicians would not be asked the sequential questions directly, but they are bound to be aware that the guide lies behind the approach. Whatever the logic of the sequence of questions, they look and feel like a series of tests of culpability, and only if these can be passed satisfactorily will attention switch to the learning about the systems and human factors that underlie the great majority of safety incidents. This is not the way to promote a just culture.

Healthcare remains a risky business, but fostering a culture of openness is a necessary step in improvement, ¹⁹ and reducing the fear of punitive responses improves openness. ²⁰ As Dalton and Williams put it, it is ‘both the right thing to do, and hugely important for reinforcing a trusting relationship between the public and the organisations that care for them’. ²¹ There are other things that need to be done, including putting an end to the professional culture of infallibility and training clinicians how to respond to error and its emotional consequences. But first, we need to stand the Just Culture Guide on its head or risk making worse the problem we are trying to address.