Tuesday, 25 September 2012

Scheduled airlines are safe, just like out patient clinics

There are constant comparisons between aviation and healthcare especially in terms of how safe aviation is.  There is no doubt that aviation in general has a low mortality rate for passengers. I have already written about the need to learn from how aviation achieved it, I have also pointed out to the limitations of the comparisons ( The term aviation or air transport in my view, includes many things, which starts from the booking process, airport formalities, baggage, catering, flying, etc. It also includes transportation of animals and goods.

In general, the whole of aviation is considered arguably to be better than healthcare. What is not arguable is that commercial scheduled airlines have a very low mortality rate for passengers. Here is my problem, death is not one of the eventual natural outcomes of transportation when transporting essentially healthy persons from one place to another; quite rightly in aviation is mortality is unacceptable. Hospitals on the other hand are not in the business of transporting passengers, people come in with illnesses and diseases many of which are really serious; mortality is one of the eventual outcomes of serious illness and disease. In other words healthcare routinely battles against death and sometimes death wins.

To compare error rates could be valid as error is often a measurable part of process failure but to compare the impact of those errors is probably a false comparison. In aviation all mortality is avoidable mortality, in healthcare it is not. So to put it in context the comparison if we must is between all mortality in aviation and avoidable mortality in healthcare (i.e. the result of process failure). That is what I mean by impact. The impact of errors that result 'morbidity' is of course hugely different like losing a bag vs losing a leg. Hence let us not compare impacts of errors such as mortality morbidity between aviation and healthcare which skews the public discourse. Let us look at error rates and see what we can learn.

Variation the enemy of quality

People talk about variation of care across hospitals and locations; it is true that there is wide variation and reducing the variation will improve outcomes.

Aviation which is often looked upon as a beacon of safety also has variation. Looking at 2004 accident rates for North American airlines Delta scored 0.30 and Value Jet/Air Tran scored 5.88 – well, you work out the how wide the variation is even in an ultra safe industry. If you start looking at international comparisons the variations are of course much worse. ( There is also a five times variation of fatalilties per million flight hours with scheduled airlines being lowest compared with general aviation.

When there is human to machine interaction as in aviation, there is such a large degree of variation. Healthcare is human to human interaction so it is hardly surprising that variation exists and could be expected to be more than other industries. Looks like variation in performance is not a problem exclusive to healthcare industry; variation is a human problem or to put it better, variation is a function of human performance.

That does not mean we must accept variation especially when it causes harm, we should work very hard to reduce it to ensure safe healthcare.

Fatalities in Aviation

The human fatality rate is very low indeed in the scheduled airlines part of the aviation industry. The air transport of animals, however has suffered bad press. Airlines are apparently not even required to report animal deaths. The mortality rate of animals in air transport is thought to be 0.2%.(

The post surgical 30 day mortality for day case surgery in humans which is about 0.01%. In-hospital mortality for day cases is probably as low as the scheduled commercial aviation segment.

CRM and Simulation are of course extremely valuable tools and has a lot to teach us in healthcare. The number of air accidents and the number of fatal air accidents have remarkably decreased over the past few decades. That is truly fantastic. What is interesting though, is that the pilot error rate has been at about 50% since the 1950s to the 2000s, percentage of accidents attributed to pilot errors has not shown a significant decrease. The proportions of various reasons for crashes have also remained more or less the same. This is in a way a tribute to the aviation industry, since the planes have become technologically very superior it would not be surprising if pilot/human error played a bigger part and it has not, that is creditable. However, I wonder if it would be valid to argue that if CRM and simulation were indeed really powerful should the pilot error rates be falling?

Survival rates of passengers in aircrafts involved in fatal accidents has not improved (and averages about 25% since the 1930s to 2000s) ( Military aircraft, fighter planes, aircraft engaged in warfare and private planes are thought to have much higher accident and fatality rate.

Some Aspects of Aviation are Safer than Others; Some Aspects of Healthcare are Safer than Others

I suppose in clinical healthcare delivery terms, scheduled airlines are possibly the equivalent of out patient care – not many patients die in out patient clinics. The risk to life is also very low for elective investigations, day case surgery and obviously the risk increases with emergencies and trauma.

In healthcare we talk about morbidity as well. Airlines perhaps should take into consideration DVTs, respiratory illnesses, musculo-skeletal problems and other health issues that happen after a flight. Non-health related morbidity for aviation perhaps include lost baggage, wrong meals............... no let me stop there before it gets silly. Hold on, why not, non-flying errors are also errors and results in 'airline industry morbidity' to passengers, perhaps not that silly.

What has to be said is in healthcare there is clearly much avoidable mortality - that is unacceptable. In healthcare the error rates in day to day activities are simply too high, that is again unacceptable. That is where learning meaningfully from other industries will help.

There is a fundamental problem with my writing here. I am not comparing like for like, I am comparing apples to pears. In my defense, I did not start that comparison. Comparing aviation to healthcare was not my original idea. There is a second problem with this manner of writing, it may sound like I am being defensive of healthcare and its practices, I am certainly not defending any poor healthcare practice or result. I acknowledge the superior results that aviation has had as a result of dedicated persistent efforts in the field of human transportation in scheduled airlines. I recognise the need for healthcare to learn from every source possible including commercial scheduled airlines. All I am saying is, let us stop comparisons and let us focus on learning. Let us look for clinical adaptations of these techniques rather than attempted direct transfer of airline techniques. Let us recognise the uniqueness and the intimacy of human to human interaction that healthcare involves.

A word of warning: This is not a 'hate aviation' piece of writing, this is not aviation versus healthcare writing. This is a plea to learn the lessons in a way that is appropriate to healthcare - clinicians and patients.

Follow me on twitter @HemadriTweets


Anonymous said...

To extend your comparison...earlier this year I encountered a woman who had paid for a transatlantic crossing with a cruise company in order to take her bulldog with her when emigrating. Why? Because responsible airlines won't fly with them because of their potential airway difficulties and thus mortality risk in flight. Of course, if a human with a similar degree of airway risk needed surgery, the NHS wouldn't have the luxury of refusing to let them "fly"...

Indian doctor said...

Enjoyed reading that. I completely support the viewpoint that though we can learn from other industries we cannot copy the model because of the inherent differences.

For Example- some NHS trusts are trying to copy the 'Toyota' model of efficiency in work without understanding that toyota made cars and NHS trusts are servicing human beings which do not come with standard sizes, shapes, type of illness etc etc

Shankarabarathi said...

I actually enjoyed reading through this posting.Many thanks.
Hospitals in Chennai
Health Care in Chennai