Making Mistakes

It has been said that "to err is human" yet there are endeavours in which error cannot be accepted. The airline industry is an obvious example, and the nuclear industry is another.  Dose this mean that to be safe we need to remove humans from the equation?  The short answer is 'Yes' and that is exactly what has been done in the above examples.  Humans are still involved, but systems have been created that monitor for errors and correct or prevent them.

Now to health care.  In many ways, errors must be avoided because they cause illness or injury, increased length of stay in hospital, or death.  Everyone accepts that errors are bad, but for some reason it is expected, by regulators, managers, the public, and health care practitioners themselves, that health care will be error free. In effect we expect that health care workers are not human! This is patently absurd because it is their humanity that makes them effective.  For all our faults and propensity for making mistakes, humans are still better at caring for other humans than machines are. Granted, there may be some individuals who do feel they have a relationship with 'Siri' on their iPhone or iPad, but very few people would actually choose to have nursing or medical care delivered by a machine. The therapeutic relationship is a crucial component of caring.

The expectation of perfection is therefore unrealistic and unreasonable, and in fact is not supported by case law.  The law accepts the concept of 'reasonableness' which therefore allows for mistakes.  So why does everyone else, including health care workers themselves, expect perfection? I think it is because mistakes are taboo, and that is dangerous.  Very few health care workers go to work intending to harm someone that day, and if someone is harmed they feel tremendous guilt and self-doubt. We are typically not trained to deal with mistakes, as the unspoken rule is that if you are 'good enough' you won't make any.  So we hide our errors, possibly even from ourselves. If you do report an error and the report is then mishandled by managers who take the easy approach of blaming you for it, you will be far less likely to report any future errors.  What about near misses?  If no harm was caused, forget about anybody reporting it.  This robs the organisation of any chance to prevent future events that might not be near misses.  In many causes, managers only have themselves to blame for this culture of avoidance.  You can't encourage a person to be honest by beating them every time they tell the truth!

That's the problem, and it's a big one.  What can we do about it?  Firstly, we must adopt a no-blame approach to incident reporting.  Mistakes happen all the time in health care, yet very few are reported especially if the patient was not obviously harmed or not aware of the mistake.  Then we must investigate the incident reports to work out what system faults caused the error or allowed it to progress.  In some cases the person is the problem, but that should never be the first or primary conclusion because no-one works in isolation from the systems.  Disciplinary matters have their place, a long way down the priority list.

We need to educate health care workers that mistakes are inevitable and our only hope of preventing them is to work together.  'Two minds are better than one' rings true as cross-checking is a valuable tool to detect and prevent mistakes.  To make this work properly, all participants have to have a voice, and be heard. Part of investigating an incident report should be about how cross-checking failed, and whether the error could have been prevented if someone had spoken up.  If so, what prevented them from doing so? Shifting blame to the person who didn't speak up is unhelpful, so the focus is on addressing the culture of the workplace so that everyone has the right to speak.

In summary, errors are inevitable in health care, and will remain so because humans are involved in it. Systems need to be created that help fallible humans to detect and prevent them before harm is caused, and these systems cannot be based on blaming the person who made the mistake. System change is driven by incident reporting, and the focus of investigation must be heavily biased towards finding system flaws that facilitated the mistakes. Finally, we all need to stop expecting the impossible s that just perpetuates the problem.

Thinking by proxy

At my workplace we have recently implemented the Modified Early Warning Score (MEWS) to assist clinicians to decide if their patients are unwell or not.  On the surface of it, you would think this is a great idea, especially as it empowers junior staff to call senior staff to review patients they are concerned about.  Given my interest in the role or rational thinking in healthcare, my view is less positive.

I have no problem with the MEWS being used as a tool to quantify your concern about a patient, even though there are times when you are concerned despite the observations appearing OK (and hence the MEWS is OK).  What I have serious doubts about is staff using MEWS to decide whether they should be concerned!!  If you need a score on a sheet of paper to decide if your patient is unwell, you probably should find another line of work.

Using the MEWS in the Emergency Department makes no sense.  It merely adds another task to the list, with little likelihood of impacting on patient outcomes.  This is especially true when there are no policies in place to govern the use of MEWS in this setting.  Say a patient has a MEWS of 8, so what?  We will already be implementing emergency care to deal with whatever the problem is, so what are we adding to the equation?  If the MEWS is supposed to guide the level of response such as Triage Category, them I'm getting really worried.  Allocating triage categories and directing ED resources is a specialised skill, and should not be done by inexperienced staff, so I can't see how using MEWS is going to improve anything in the ED.

However, because someone higher up the food chain than the frontline staff has decided that MEWS is the best thing since sliced bread, it seems that it is here to stay.  Its use is even being audited, which is quite funny is a sad way.  Since it has no bearing on patient outcomes in the ED, auditing its use is a waste of resources.  If we're going to audit it at all, we should audit its efficacy not just whether it is being used!!

Cookbook clinical practice

For many years I have been dismissive of the trend towards what I call "cookbook practice".  This is clinical practice in which all the decisions are pre-set, and the clinician just fills in the blanks and follows the resulting algorithm for treatment or referral.  My objection was based on the idea that clinical judgement is an expected skill/attribute exhibited by practitioners, and they should not have to be spoon-fed.

Recently, I have begun to have my thinking modified by some new data (well, 'new' to me anyway!).  Safety and quality in healthcare has been expensively studied over the years, and it has been found that having clinical guidelines in place significantly reduces the incidence of adverse events.  This may suggest that even good practitioners can benefit from having robust well-founded clinical guidelines to help with decision-making.  I'm still not completely comfortable with the idea, as I worry that guideline-based practice actually reduces the need for clinical judgement.  Maybe clinical guidelines work because practitioners need help making good decisions?  What if we looked at improving their decision-making, and their critical thinking?  Would that not achieve the same or better results?

In some ways, it seems analogous to the issue of young drivers on the roads.  This group are horribly over-represented in road crash statistics, so it seems logical that something needs to be done.  Is it more training and guidance for a longer period of time that is required, or is it more effort into creating drivers who think actively about what they are doing and learn good attitudes and habits?  If having L-plates for 2 years proves to be safer, what about 3 years?  If the restrictions of provisional drivers licences make young drivers safer, what's wrong with applying those restrictions to all drivers.  After all, if we did not allow anyone to drive until they were 21, the road crash data for 16-20 year olds would improve dramatically!

I think the answer is the have reasonable guidelines, then teach people how to think properly.  Some sort of attitude test if you like.  If you don't pass, you don't get a licence until you grow up a bit more.  I know some 16 year olds who are more mature than some 46 year olds, and much safer drivers as a result.  Likewise with clinical practice, I think it should be mandatory to pass some sort of critical thinking test before you are allowed near patients.  How that would look in practice, I'm still thinking about.  Watch this space ...

Critical thinking presentation

I recently accepted an invitation to speak to a group of staff at TCH about critical thinking in clinical practice. The attached file is the one I used for the presentation. ...  Well, that was the plan anyway.  Seems the file didn't attach!  Maybe this will work instead:  Try this link