Error reporting and safety psychology

[wmdgrant]

I trust this is an appropriate question for this forum. If not, I apologize in advance (and would someone let me know so I don't bother people).

I am an associate dean at a medical university (SUNY Upstate Medical University).
We are currently engaged in a project to understand the genesis and development of medical errors and how they may be identified and prevented. It seems to me that somewhere in pilot training there is a fundamental change that occurs that helps pilots function within a precise set of operational rules. Increased numbers of rules seems to have the effect of improvement in error reduction and reoccurance. In medicine, the implementation of operational rules and guidelines and/or implementation of new safety devices seems to have little change on error occurance. Part of this may stem from the fact that in medicine, error identification and reporting is usually greeted with a punative response from regulatory systems. I would like to try to understand what goes on in pilot training that engenders such an appreciation of rule compliance and error reporting and try to understand how (or if) this might be translated into medical systems. One of my suspicions is that while errors both in medicine and aviation may lead to disastorous outcomes, the time and ability to 'correct' may be greater in medicine. Is there someone(s) who would like to discuss this further?
Thanks.

[John Farley]

A very interesting post. And I for one certainly have some views but they will have to wait a couple of days.

May I suggest that you also post an invitation on the Medical and Health forum to look at your topic on Flight Test. I think the subject should reach a wider group than those likely to read Flight Test.

Regards

[Droopy]

Perhaps the most difficult requirement for this to work is to generate the trust from your potential reporters that their reports will be handled in a mature and reasonable way. It's well established that the true value of an incident reporting system is not just about the major occurrences but also about picking up the undesirable trends which, when combined with say a technical failure and a training deficit then result in an occurrence.
Probably the greatest edge that aviation reporting has is that it has been integral to the industry for so long. I have no knowledge of the medical profession but I am closely involved with reporting of flight safety matters in police air support - both pilots and officers - and it's been interesting to see how with regard to the latter only time has engendered sufficient trust from a profession which also suffers from blame and disciplinary proceedings being management's common response.
I can see the "verbose" high level warning light flashing now but feel free to email if you want to discuss it further.

[Genghis the Engineer]

(Dr?) Grant,

A very interesting question, which I've debated on several occasions with Medical friends.

Speaking as both a pilot and an engineer, I'd say that the culture you are asking about pervades - has to pervade - all of professional aerospace. It's not specifically dedicated to pilots, although there are peculiarities to the piloting profession.

Firstly let me try and explain what the professional aerospace culture is. It is firstly about obsessive attention to detail, and continual doubt and questioning of ones own, and ones colleagues abilities. Anybody not seen to doubt their own infallibility is regarded with great suspicion. A co-pilot is encouraged to question his captain's actions, I praise any of my subordinates for spotting my own cock-ups, and they know I will point out their mistakes, then "file and forget". This is a hard culture to adopt, and most of us take a few years to get the hang of it. I'd say about 8 in my case.

How is this achieved. Well, probably a lot of it is because all of us are subject to constant monitoring and review, by senior pilots, by the authorities - in a continuous chain. Even the national authorities such as FAA and CAA are annually audited by an international body called ICAO. This is ingrained in the culture, but also you might next month be auditing the chap auditing you now (or visa versa), which encourages a lack of blame. Next month my office will be ransacked by the annual CAA audit, which I look forward to as a way to ensure I'm doing the job right - honest. If (or usually when) they find deficiencies, we will agree a programme to bring things back up to speed, they will NOT take disciplinary action except after repeated opportunities to correct problems (which has happened, but only when the CAA think it's necessary to protect the public).

However there are other more tangible elements to the system, one of the most significant (at least here in the UK) is something called CHIRP. CHIRP, or Confidential Human Factors Reporting Programme ( www.chirp.co.uk . CHIRP is an independent organisation to which anybody may send a report on a failing by themselves or anybody else. The receiving body carefully disidentifies the report, then passes it to a panel of experts for investigation. If the panel has questions they go through the receiving body (the only people to know your identity). Ultimately a report is issued in a way that ensures everybody within the industry knows a mistake was made and how to ensure that they don't repeat it. Exceptionally a company (say an airline) might be told "this problem occurred with one of your crews - amend your training procedures". Once the report is issued, the reporter's details are destroyed. CHIRP exists throughout civil aviation, both in flying and engineering, and the military have their own equivalents which work the same way.

Another example (which, it should be said, works better in the UK than the USA, at least in civil aviation) is accident investigation. The UK Air Accidents Investigation Branch has a mandate to investigate accidents "so that there will be no more accidents". They are allowed to blame procedures, manuals, or recommend amendments to equipment. They are not allowed to blame people. I have actually sat in a meeting at AAIB where the Principle Inspector stopped the meeting and said "I'm sorry, we're getting to close to blaming somebody in this discussion."

So, when professionals know that their mistakes are investigated like that they are prepared to be open. Nobody wants to be bad at their job, so if they're allowed to go to their colleagues for help, or to report their own mistakes for others to learn from, they will.

I hope this helps, no doubt others will have their own views.

G

[John Farley]

Bill

Thanks for the email

You have received some good comments above which are much more relevant to making the medical system safer than what I am going to talk about.

I am going to offer comment about your actual question regarding why pilots have a different attitude to the rules than doctors. You wonder what it is in their training that brings this about. I suggest that the first time a pilot flies solo it is immediately apparent that he can die if he does not live by the rules. This is a good incentive to stick to the rules and also to accept extra rules in the future once they become developed through the experience of the aviation community.

Pilot breaks rules - pilot dies. Doctor breaks rules - patient dies. Not the same thing.

It gets better. When a pilot breaks the rules and might well have died but actually gets away with it (through good luck or skill or both) he now has a very good talking point at the bar and will certainly brag about what he has got away with. This sort of word tends to spread. He is not keeping the event to himself and feeling bad, he is singing his head off and feeling good.

Once a pilot gets a job ferrying 400 people behind him, we hope that he has a real grasp of the rules and can be truly relied upon to stick to them in spades, because if he doesn’t he will die and he knows this. (The others behind will die as well but that never crosses his mind. After all he does sit in the front and that is the worst place to be if you hit something)

The culture of total honesty and professional integrity is not easy to encourage in any trade. For the reasons I have mentioned I think such a culture is a tad more likely to be alive and well in the flying business.

I have used "he" throughout quite deliberately. In my experience men are more ego driven, more casual about danger (indeed may even seek it) and certainly are more given to brag than ladies. So some of my comments are not so generally applicable to the modern breed of very good lady pilots.

The bottom line though is that the systems that are used in the flying business to encourage safety (and rule observation) are now quite sophisticated and could easily be transplanted (sorry) to a medical profession that wanted to reduce the effects of the errors that we humans make and always will.

Good luck with your project

Regards

[aspinwing]

Hi doc,
You have asked an awful lot in a couple of sentences.
Let me also state that you should assume this post to be ‘gender neutral.’
Let me start with an old ‘saw.’ “ A surgeon buries his mistakes but a pilot arrives with his.
There is a complication facing both medicine and aviation and that is the litigious nature of the US society -- I say that from the perspective of a Canadian married to an American -- which is creeping across the border. But, that should be another thread.
There was a time, not too long ago, when every crash was ‘pilot error’ -- aircraft crashed -- pilot was in ‘command’ -- ergo ‘pilot error.’
Gradually crash investigators started to look at outside influences. Was the weather as forecast? Was there an aircraft warning that the pilot could have, should have seen? Was there a mistake made by someone in the Air Traffic Control organization? Were all maintenance procedures done correctly? Were there ‘systemic’ factors that contributed to the crash?
Should training be changed? Should an aircraft be modified?
Remember crash investigators and lawyers all have 20/20 (or is it 6/6) vision and all the time in the world. It stikes me that inquests - assuming that the case comes to that -- are more about lawyers protecting their client’s interests than really fixing the ‘system.’
Very few aviation accidents have a single cause; there is a lot of redundancy built into most systems / aircraft.
The Standard Operating Procedure for most airline malfunctions at altitude is to get the book out and work your way thorough the problem.
How many deaths had occurred before the fittings on various gas hoses in hospitals were changed so that the wrong gas could not be connected to the patient. Surgery was correctly done but patient dies from post surgical infection from the general hospital environment. Is that the surgeon’s ‘fault?’
The day is probably not too far off when surgeons will do many, certainly complicated, cases to a simulator that has been fed the data that relates directly the patient; CAT scan, blood pressure, ultra sound, etc.
“One of my suspicions is that while errors both in medicine and aviation may lead to disastrous outcomes, the time and ability to 'correct' may be greater in medicine.”
Answer: Not necessarily.
If a pilot is not happy with a landing approach, he can always ‘go-around’ and have another go at it -- although there is a lot of pressure not to -- while I would imagine that a heart surgeon who has an aorta - suddenly and unexpectedly -- rupture does not have time to open the text book - read, aircraft operating manual. Seems to me that ‘STAT’ is a word that I have heard more than once in a hospital setting.

Previous posts have rightly indicated the safety is a culture; how you change the medical culture is a daunting task.
Good luck.
David

[john_tullamarine]

Another related point which was drawn clearly to me many years ago is that most experienced pilots (in regard to mental determination/motivation) are quite different to the average Joe in the community - and this fits quite nicely with a disciplined approach to operations.

In our CRM refreshers at one period, a party trick was to seat two people on opposite sides of a table but with a divider between so that neither could see the other. The one would have some arrangement of objects in front of him, the other the same objects but jumbled. The task was to communicate sufficient information from the one to the other (against a tough clock limit) to end up with the jumbled objects in the same arrangement as the ordered objects.

I was always rather amazed that pilot pairs would near invariably work it down to the alarm even after the goal had become patently unachievable. Non-pilots in the same role play exercise very often threw their hands up in despair as soon as the goal-time situation became difficult.

[Chunkee]

Interesting thread.

With regard to the way in which aerospace culture is formed and has changed, I have recently finished reading the report on Qantas aircraft that overran the runway in Bangkok, no casualties, but $100 million damage.

As was said earlier, it could have been called "Pilot Error" he selected wrong flap, and wrong reverse thrust setting, then compounded by a late decision change.

Yet true to form in the current culture and learning environment, it was found that the pilot had only ever been trained to use that much flap/thrust setting. In fact the company had almost demanded this be the norm for all approach and landings.
All pilots have been retrained and the organisation has learnt as a whole. The pilot still flies today, it was not his fault.
The report makes very interesting reading and was published by the Bureau Air Safety Investigation in September 2000.

I also highly recommend the book "Human Factors in Multi-Crew Flight Operations" by H.Orlady and L.Orlady published by Ashgate 1999.

Several chapters covering Non-punitive incident reporting, selection and training, crew resource management and other aviation psychology issues.

[Nick Lappos]

wmdgrant,

You question about the rule-making process in the "programming" of a pilot is a subject I have thought about at some length. I am sure the issues are universal to training of humans in most fields. Let me share some thoughts with you.

There are rules to assess risks, define actions and predict outcomes that grow increasingly complex with increased operational flexibilities (ie, more environments, more types of short term operations, more types of missions). The determination of proper human conduct through training seems to have 4 levels, similar to the steps in a typical trade:

1) Novice or apprentice - have a limited set of hard fast rules, simple, direct, deterministic, describing a sub-set of the total required environment and tasks. The student and private pilot fall in here. We limit environments and tasks to capture safety ,(day, visual), and we limit exposure (fewer seats), and we reduce pressure to accomplish the task (non-commercial recreational operations with little pressure to perform).

2) Journeyman - an accomplished capable achiever. Knows more sets of rules, each set adapted to specific environments, all still deterministic. The rule sets are complex, and the journeyman knows which to use when, but may reach limits of operational capabilities in unusual corners, or when the environments become mixed, creating a minor conflict (fuzzy correlation) between competing rule sets. Low time commercial pilots or high time private pilots are journeymen.

3) Experts - have a more complex set of rules, with increasingly greater breadth of operational capability. They have determined the cross-overs between rule sets that bridge the conflicts that stifle a journeyman. Their rule sets are not separate and parallel, they tend to cross and meet in places where the journeyman is stimied. They can fly nightand instruments and with icing successfully, where the compounding of the rule sets is quite complex. They are still rule driven, and occasionally fail to perform because they chose the "wrong" rule set, usually by believing inaccurate information and failure to identify that fact in time (American Airlines Chicago, engine dropped off, aircraft was climbing successfully, the crew slowed down to Vy and lost roll control. Air Florida at DCA obeyed max EPR and failed to climb, struck 14th st bridge. Three Mile Island crew did not identify incorrect reactor pressure/temp readings and misidentified the situation). Most failures in this class of pilot are lumped into a breakdown of "situational awareness." High time commercial pilots and airline pilots are in this catagory.

4) Masters - have passed through a gate that breaks down the multiple rule sets of the expert through unification of the rules. The master pilot sees the situation as more simple than the expert, because the master's rules are reduced into fewer, simpler rules with "fudge factors" that allow the rules to apply while jumping between operational environments. Many very high time pilots with broad backgrounds have achieved master level. Test pilots are there too. The master explains decisions using standard fuzzy logic phrases such as "that depends" and "it's simple, really". If we plotted the master's rules, they would be three dimensional. The master can predict outcomes that are robust, and see through the conflicts brought about by inaccurate information, because the master can see and eliminate the inaccurate data and inconsistancies before an inaccurate answer is produced.

I would be glad to discuss this with you further, if you'd like. I am a high time experimental test pilot, and have created training programs for introduction of new equipment and new missions.

Regarding the error reporting issues of aviators, there is a general cultural conditioning that I have observed that might shed some light on the problem. In most cultures, pilots tend to be self-deprocating, and generally credit luck and chance for specific achievements. They also tend to describe screw-ups in hangar talk to help spread the word on what not to do. There is an almost opposite pilot culture to the character of "Maverick" in Top Gun, the swaggering ego, because this ego driven person will not learn from his mistakes, and will end up repeating them, disasterously. The best pilots I know are confident, and can assert themselves with the machine, but they can easily admit mistakes, tell what went wrong, and help all learn from the error. I wonder if medicine has the same culture, or do doctors tend to hold their most intimate self appraisals inside?

[ 29 November 2001: Message edited by: Nick Lappos ]

[ 29 November 2001: Message edited by: Nick Lappos ]

[wmdgrant]

I can't tell you all how incredibly helpful this has been and I deeply appreciate the information that you have shared (and the personal exchanges that have resulted.)

In addition to the Orlady's book reference, I have also been pointed to a one act play "Charlie Victor Romeo." They have their own web site at www.charlievictorromeo.com. It is based on cockpit voice recordings and discusses decision making. Note that they do not use only fatal incidents. The play won several major awards. I am trying to track down a copy of the script.

I am intrigued by the co-pilot analogy. Physicians have 'co-pilots' only during medical school and residency training (after medical school for training in specific specialties). Later they are on their own often with little on-going 'supervision.'

I do agree that there may be something to the distancing of the physician from the event - - it happens to someone else. Also, with medical care, it is often possible to 'rectify' and error on the run. A wrong medication dose can often be counteracted. And there is often a prolonged warning that something is going wrong.

As I write this I continue to be struck by the similarity in language that physicians and pilots use. A patient going bad is said to be 'crashing.' A patient who had a bad, unexpected outcome is said to have 'crashed and burned.' When things are not going well 'alarm bells are going off.'

With your collective permission, I hope to continue this thread as things develop here.

Bill

[go with the flow]

Thanks to all so far. A less established angle follows. I've recently started flying (past solo, but not much) and been struck by how badly we do this sort of stuff in medicine. At this stage all I've done in my own practice is to establish proper checklists for some areas where omissions were either frequent, or alternatively unremediable or at least difficult. Just as in aviation there are some things from which you can recover, and other things that you simply can't, and therefore can't let yourself get into.

In terms of the comments about the co-pilot issue I think that the medical culture, at least in australia, is still very hierachical. The "right" of a registrar (advanced trainee specialist: about 4-8 years out) to speak out about a case or a procedure is still not at all accepted (having been there not long ago), despite the fact that more often than not they're in a better position to know than someone who trained 25 years ago, and hasn't retrained since. Now I'm on top there's very poor currency assessment. God knows how we change this culture, because if you're a surgeon, you're GCMG (God calls me God), and physicians (my area) are only marginally better. That's not to say that I entirely believe the propaganda about the spiritual harmony at the pointy end (!), but at least many of the people there believe that they should be trying even if they aren't perfect, because it will actually help outcomes. We medicos haven't got to that point yet where it's accepted that it should be accepted!

Another problem as stated is that error reporting in medicine is suppressed by the design to protect one's own, given that virtually all error is punishable in medicine, and 'there but for the grace of god'....It should also be realised that I pay directly for my colleagues errors, because our insurance funds are collectively based. The short term view therefore is to shut up, especially about lucky escapes as far as any broader reporting goes. In relation to John Farley's comments, don't think that medicine doesn't have bar talk about stuff-ups, I think every profession has its' insiders-only nights, and one actually takes a lot away. However, unlike the aviation safety data, there's no investigation and publication service of these incidents. It's the careful informed examination then dissemination of the knowledge that's the problem. I KNOW as a C150 pilot with ~16 hrs (sad, isn't it, when the last digit's relevant)that the american data for the last 5 years is that what's likely to kill me is fuel exhaustion, accidental IMC, CFIT from inadequate performance (uphill, downdrafts in mountain lees etc), stalling while not paying attention (i.e. overflying your house) or with too much to do (a complicated go around), a fair few midair collisions, not that many engine failures without warning and the occasional one where the word suicide is not actually in the report. Easily less than half would be difficult to avoid (at this level of aviation). Knowing that, I can do something: dipstick or inspect fuel every time (the drag racers siphoning your full tanks overnight made me decide that), get an IR ASAP, etc. It may surprise you that I know all sorts of diseases that can get my patients, and their tests and treatments, but despite 13 odd years since starting I know virtually nothing hard about what are the common causes of error in my field that degrade outcomes.

Enough ranting. To Bill and anybody else, I would very much appreciate it if you could give me an idea of any initial attempts you've made, and any publications of your own or of otheres that you think are especially worthwhile. Links could be posted, or if you have them as a pdf or other file would be welcomed (along with any commments) at [email protected]

stephen

[ 03 December 2001: Message edited by: go with the flow ]

[GlueBall]

J. Farley's assessment was most correct. Comparing pilots' decisions with doctors' decisions is not valid because pilots' decisions are always based on personal survival instinct! Doctors are not confronted with personal life and death decisions when attending to patients.

[Genghis the Engineer]

I don't think I entirely agree with you. The safety related decision making process is not exclusive to pilots; everybody in aviation is expected to operate to the same approach. This runs from ground handlers to air trafficers.

I work, professionally, as both an aviator and an Engineer. I actually find that I tend to be even more obsessive about getting things right in an engineering context, where it's somebody else's neck on the line, than I do when flying an aircraft. I don't think an Engineer whose mistake has killed somebody is going to feel any differently about it to a surgeon in the same position. But, the engineer does seem to operate in a far better culture for double-checking his actions than the surgeon.

G

[GlueBall]

There is no question about an engineer's or doctor's concern, commitment, devotion and due diligence to their work. But it does not compare to the adrenaline rush of cheating death when barreling halfway down a wet runway at 140 Kts with 360,000 Kg of airplane strapped to your butt. If an engine decides to go on vacation at Vee One it's split second decision time.

[Genghis the Engineer]

Yes, but I think that the subject under discussion is error reporting and supervision - not behaviour in a crisis. I'm sure any surgeon could relate similarly urgent and stressful circumstances, even if it is not their own life on the line at the time.

G

[wmdgrant]

It's interesting how the non-medical folks and the medical folks differentially view eachother. This is great, actually. I like Nick's analysis of the individual form novice to master. In the world of congnitive psychology, novices are found to process in a linear fashion. They tend to go from one step to the next and often have difficulty 'seeing the forest for the trees.' Masters tend to process information through 'chunking.' That is, the simultaneous input and processing of data.

In the airline industry, you have responded to this by limiting the activities and responsibilities of the novice. Medicine tends to do this but with much less vigor and oversight.

It is also correct that in medicine virtually all error is punished and that error recognition is not 'rewarded.' I think that part of this is due to the medical profession itself which over the past 40 years or so has moved from a position of 'we will do the best that we can' to 'we can fix anything.' Given that premise, when outcomes are not ideal, then the patient and the system go hunting for someone to blame.

Over the past couple of days here, there have been a series of news stories related to 'wrong side' surgeries. Folks are running around trying to figure out how to prevent these. It seems to me that this situation is ripe for 'pre-flight' checklists conducted by more than one person and cross-checked repeatidly.

I think that you folks have gone a long way toward distinguishing system and human errors and that the medical system needs to take a look at this.

As we develop our efforts here, I will be pleased to share our findings, literature and anything with anyone who might be interested. Thanks for the continued converstation.

Bill