4.2 Extent of healthcare error

In 2000, an influential report entitled To Err is Human: Building a safer health system (Kohn et al., 2000) suggested that across the United States somewhere between 44 000 and 98 000 deaths each year could be attributed to medical error. A pilot study in the UK demonstrated that approximately one in ten patients admitted to healthcare experienced an adverse event.

Healthcare has been able to learn from a number of other high‐risk industries, including the nuclear, petrochemical, space exploration, military and aviation industries, about how team issues have been managed. These lessons have been slowly adopted and translated to healthcare.

Specialist working groups and national bodies have been instrumental in promoting awareness of the importance of human factors in healthcare. They aim to raise awareness and promote the principles and practices of human factors, identify current human factor activity, capability and barriers, and create conditions to support human factors being embedded at a local level. One such example of this in the UK is the Human Factors Clinical Working Group and the National Quality Board’s concordat statement on human factors.

4.3 Causes of healthcare error

• Q. What one thing links all of these errors?
  
• A. The humans involved – these are all examples of human errors.

Humans make mistakes. No amount of checks and procedures will mitigate this fact. In fact the only way to completely remove human error is to remove all the humans involved. It is vital therefore that we look to work in a way that, wherever possible, minimises the occurrence of mistakes and ensures that when they do occur the method minimises the chance of it resulting in an adverse event.

4.4 Human error

It has been suggested that these human errors can be further categorised into: (i) those that occur at the sharp end of care by the treating team and individuals; and (ii) those that occur at a blunt or organisational level, typically through policies, procedures, staffing and culture.

These errors can be further subdivided as shown in Table 4.1.

It is typically found that the latent/organisational issues often coexist with the sharp errors; in fact it is rare for an isolated error to occur – often there is a chain of events that results in the adverse event. The ‘Swiss cheese’ model demonstrates how apparently random, unconnected events and organisational decisions can all make errors more likely (Figure 4.1). Conversely, a standardised system with good defence mechanisms can capture these errors and prevent adverse events.

Each of the slices of Swiss cheese represents barriers that, under ideal circumstances, would prevent or detect the error. The holes represent weaknesses in these barriers; if the holes align, the error passes through undetected.

Reconsider the example of drug error using the Swiss cheese model. The first slice is the practitioner writing the prescription, the second slice is the organisation’s drug policy, the third is the practitioner who draws up the drug and the fourth is the practitioner who second checks the drug.

Now consider the following: What if the first practitioner is very junior and not familiar with that area or the drugs used? – her slice of cheese has larger holes. What if the organisation has failed to develop a robust drug policy that is fit for purpose? – this second slice is considerably weakened or may even be removed completely. What if the practitioner does not normally work on that ward and is not familiar with the commonly used drugs? – his slice has also got larger holes. What if this area is always short of staff, so staff do not routinely attempt to get the drug second checked? – this slice is completely removed.

The end result is that multiple defences have been weakened or removed and error is more likely, and the error is more likely to cause harm. Also be aware of the different types of error with potential gaps in knowledge, a latent/organisational error (no effective policy and possibly an issue with nurse staffing) and a routine violation.

4.5 Learning from error

Historically, those making mistakes have been identified and singled out for punishment and/or retraining, in what is often referred to as a culture of blame. With our example, drug error blame would most likely have fallen on the shoulders of the nurse administering and/or the doctor incorrectly prescribing. Does retraining these individuals make it safer for other or future patients? That clearly depends on the underlying reasons. If it was purely a knowledge gap, possibly, but does the same knowledge gap exist elsewhere? Potentially all the other issues remain unresolved. Moreover such punitive reactions make it less likely for individuals to admit mistakes and near misses in the future.

The focus is now on learning from error and in shifting away from the individual, is much more focused on determining the system/organisational errors. Once robust systems, procedures and policies that work and are effective are in place, then errors can be captured. Of course, issues will still need to be addressed where individuals have been reckless or lacked knowledge – but now reasons why the individuals felt the need to violate procedures or had not been given all the knowledge required, can be looked at.

For this to work health services need to learn from errors, adverse events and near misses. This requires engagement at both the individual level, by reporting errors, and the organisational level, investigating and feeding back the error using a systematic approach. It is also key that information is cascaded through the organisation and across the health service to raise awareness and prevent similar situations occurring.

Violation may be indicative of the failure of systems, procedures or policies or other cultural issues. It is important that policies, procedures, roles and even our buildings and equipment are all designed proactively with human factors in mind so things do not have to be fixed retrospectively when adverse events occur. This means that all members of the organisation must be aware of human factors, not just the front‐line clinical staff.

4.6 Communication

Poor communication is the leading cause of adverse events. This is not surprising; to have an effective team there needs to be good communication. The leader needs to communicate with the followers, and followers communicate with the leaders and other followers. Communication is not just saying something – it is ensuring that information is accurately passed on and received. We all want to ensure effective communication at all times. Remember there are multiple components to effective communication (Table 4.2).

When communicating face‐to‐face a lot of the information is transmitted non‐verbally, which can make telephone or email conversations more challenging. Communication can be more difficult when talking across professional, specialty or hierarchal barriers as we do not always talk the same technical language, have the same levels of understanding, or even have a full awareness of the other person’s role.

There are a variety of similar tools to aid communication, such as SBAR (situation, background, assessment and recommendation). Find out what tool your organisation has in place and practise using it; look out for other staff using it too. SBAR is designed for acute clinical communications. It facilitates the sender to plan and organise the message, make it succinct and focused, and provide it in a logical and expected order. It is also an empowerment tool allowing the sender (who may be more junior) to request an action from a more senior individual. While these tools are useful, they tend to be reserved for certain situations, whilst we want to establish effective communication as the routine not the exception. One method to routinely improve communication is to incorporate a feedback loop.