Fatal Accident Rate of Angel Flights Seven Times Higher than Other Private Flights

[APenNameAndThatA]

On ABC News Web Page 

 

https://www.abc.net.au/news/2019-08-13/atsb-releases-findings-into-fatal-angel-flight-crash/11407294

 

That would explain the talks about being much more strict with Angel Flights. I don't know about the statistical significance of the difference and have not read the original report. 

 

 

[facthunter]

 It's an emotive issue this one, but It's a situation with all sorts of possible weaknesses. I can't see 250 hours meaning much. Recency and a PIFR rating is what's needed as a minimum standard. Nev

 

 

[alf jessup]

At the end of the day it doesn’t really matter that it was an Angel Flight, the VFR pilot took off in conditions close to IMC and ended up IMC killing himself and 2 innocents that trusted him.

 

 

[APenNameAndThatA]

I was surprised to read that Angel Flight said that it was CASA who regulated things and not to blame them (Angel Flight), *and* they lobbied CASA to not regulate them or their pilots more fully! 

 

It seems that the problem is pilots feeling pressure and taking off into dicey weather. The obvious solution would be for a dispatcher of some type being involved. Perhaps volunteer flying instructors could have pilots describe the weather and etc to them and sign off on the trip. That would give the pilots an out. Or Angel Flight could do that themselves. There would be insurance issues to consider. 

 

One of the people who died had juvenile rheumatoid arthritis. That could have been managed by a GP and Skype I would have thought. 

 

 

[facthunter]

 I've been in similar situations during searches. You tend to go that bit extra because of the "importance" imagined or otherwise  of getting it done. Nev

 

 

[turboplanner]

I was surprised to read that Angel Flight said that it was CASA who regulated things and not to blame them (Angel Flight), *and* they lobbied CASA to not regulate them or their pilots more fully! 

 

It seems that the problem is pilots feeling pressure and taking off into dicey weather. The obvious solution would be for a dispatcher of some type being involved. Perhaps volunteer flying instructors could have pilots describe the weather and etc to them and sign off on the trip. That would give the pilots an out. Or Angel Flight could do that themselves. There would be insurance issues to consider. 

 

One of the people who died had juvenile rheumatoid arthritis. That could have been managed by a GP and Skype I would have thought. 

 

The flights were handled by Pilots in Command. They had no one to blame but themselves. CPLs are the ones who have the extra training, extra experience indecision making, and usually IMC qualifications.

 

 

[old man emu]

This caused me to do a double-take

 

Analysis by the Australian Transport Safety Bureau (ATSB) has found that community service flights conducted by charity Angel Flight Australia have a fatal accident rate more than seven times higher than other private flights.

 

This looks to me to be lying with statistics. I can't quite explain my suspicion, but to collect data based only on the motivation for a flight and compare that data with data for flights having different motivations seems dodgy.

 

What is the accident rate of $100 hamburger flights compared to that for flights to or from air shows? What about comparing the accident rate for Sunday afternoon local area flights with water bore inspection flights?

 

 

[turboplanner]

This caused me to do a double-take

 

Analysis by the Australian Transport Safety Bureau (ATSB) has found that community service flights conducted by charity Angel Flight Australia have a fatal accident rate more than seven times higher than other private flights.

 

This looks to me to be lying with statistics. I can't quite explain my suspicion, but to collect data based only on the motivation for a flight and compare that data with data for flights having different motivations seems dodgy.

 

What is the accident rate of $100 hamburger flights compared to that for flights to or from air shows? What about comparing the accident rate for Sunday afternoon local area flights with water bore inspection flights?

 

It’s valid to compare fatalities between charter flights to much the same destinations to volunteers-on-call carrying about the same number of people to roughly the same locations under the same time expectations.

 

 

[Downunder]

One of the people who died had juvenile rheumatoid arthritis. That could have been managed by a GP and Skype I would have thought.

 

The late night pessimist in me thinks the doctor couldn't make as much money that way......

 

 

[frank marriott]

The subjects of PVT, CHTR & medicals will be rehashed and debated again as usual.

 

The real issue from my observations is compliance with VMC, IMC & night limitations.

 

NVFR in other than ideal conditions is always a serious issue.  Lengthy flights at night without a CIR requires more in-depth planning & considerations and the current available electronic devices (although helpful) do not substitute for training and currency.

 

 

[APenNameAndThatA]

The late night pessimist in me thinks the doctor couldn't make as much money that way......

 

I should have been more clear. The patient and and GP could have skyped the specialist from the GP's office. 

 

 

[Jaba-who]

I should have been more clear. The patient and and GP could have skyped the specialist from the GP's office. 

 

Who knows what the medical service they were going to was for. 

 

Just cos they has Juv. rheumatoid doesn’t mean the service they were going to was related to that. 

 

And even if it was could have been anything from further tests ( of many possible types) to getting physio assessment, to assessment by a surgeon for joint surgery ( of the JR affected joints, to specialists in a number of fields where JR can be a cause or contribute to other system disorders. 

 

 there are  plenty of limitations on telemedicine that are make face to face assessments still the gold standard. 

 

 

[Jaba-who]

As someone who deals in stats (and trying to see through dodgy ones frequently), on a daily basis I too am pretty cynical about this announcement. 

 

How does two crashes in 10 years get morphed into any useable statistic. It’s just noise in the background from a statistical point of view. 

 

This is clearly CASA just grabbing at dodgy figures and throwing them out in the media to hopefully blind everyone to their knee jerk reaction. 

 

Sounds like like the  jabiru engine fiasco all over again. 

 

 

[pmccarthy]

I agree. You can't make a statistic out of two incidents.

 

 

[aro]

How many fatal crashes would you need to see to consider it significant?

 

 

[old man emu]

Yes, you can make a statistic out of two incidents. The statistic just won't have much meaning, unless it's used for wool pulling.

 

 

[turboplanner]

The statement was not based on two incidents: This is what it said: "The average likelihood of a fatal accident involving an Angel Flight organised passenger-carrying flight was more than seven times higher than other private flights," 

 

It's referring to the  complexities faced in an Angel Flight vs a normal private flight.

 

You can take safety action following one incident, or even based on a narrow escape.

 

I can remember a Coroner giving us a talk and holding the remote control of a hydraulic tailgate with a green plastic insert on it. It replaced the normal one where a person slipped between the tailgate and body, and pressed the stop button, but the wiring had been damaged and stayed in the down mode cutting the operator in half. The manufacturer had voluntarily replaced every remote control they had sold with the green plug version which fails in the stop position. No one has been killed since.

 

 

[Jaba-who]

The statement was not based on two incidents: This is what it said: "The average likelihood of a fatal accident involving an Angel Flight organised passenger-carrying flight was more than seven times higher than other private flights," 

 

It's referring to the  complexities faced in an Angel Flight vs a normal private flight.

 

You can take safety action following one incident, or even based on a narrow escape.

 

-snip snip - 

 

Mmm what does any of that mean?  

 

You can take action that does not disaffect anyone or you can knee-jerk react and throw everyone into chaos and then “justify” it with made up rationale. 

 

Deriving a likelihood of a future event ( and assigning a numerical value to it) can only truthfully be based on the statistics of past events. 

 

You can’t just make guesses about what you perceive to be risky and place a numerical likelihood on it causing some future event. Plenty of people perceive things to be risky and demand things be done to mitigate the tick,  yet we never actually see events happen from it. I see it everyday with “crusaders of causes” in the medical world. 

 

You can say “I perceive this to be a risk of some problem happening” but until it happens it’s just conjecture. You certainly can’t say it is 7 times more dangerous than something else purely based on cogitating over possible complexities and generating fallacious numbers to match them. 

 

That is in fact is “somebodies method”. I forget who the name is but it’s a psychological trick that is used by people to make it look as if they know what they are talking about. 

 

The basic principle is you must convert a general principle into a number even if you have no source of or evidence to support that number.  Numbers equal certainty. 

 

That gives the impression you actually have solid evidence.

 

Next - Then you must make sure that number is not a round number because round numbers seem fake but unusual numbers seem likely to be real. So 7 sounds more real than 10 times. 

 

Its even better if the number has a decimal place because that increases the perception of accuracy. 

 

So I disagree absolutely  with you that it is reasonable for CASA to look at complexities in isolation of past events and come up with a  number of 7 times the risk compared to another private flight. It is just statistical absurdity! 

 

 

[old man emu]

At the other end of the stick, at least 98% of pilot/passenger fatalities involved an aircraft in all stages of flight. The other 2% accounts for walking into propellers or jet blast, and similar incidents when the aircraft was not involved in flight.

 

 

[Jaba-who]

How many fatal crashes would you need to see to consider it significant?

 

“Statistical significance” is an entity that has nothing to do with cultural or personal or societal significance. 

 

A persons death in an accident is significant - to them, their relatives etc etc. 

 

but in terms of determining whether one persons death in the other  million people (or whatever number it is) who die today in the world and whether that one persons death can be used to prevent others etc is dependant in its “statistical significance.”

 

One in how many?, how many were caused by the same thing, how many were caused by similar but different etc etc etc.

 

”Significance”  is has many meanings. 

 

 

[turboplanner]

Mmm what does any of that mean?  

 

You can take action that does not disaffect anyone or you can knee-jerk react and throw everyone into chaos and then “justify” it with made up rationale. 

 

Deriving a likelihood of a future event ( and assigning a numerical value to it) can only truthfully be based on the statistics of past events. 

 

You can’t just make guesses about what you perceive to be risky and place a numerical likelihood on it causing some future event. Plenty of people perceive things to be risky and demand things be done to mitigate the tick,  yet we never actually see events happen from it. I see it everyday with “crusaders of causes” in the medical world. 

 

You can say “I perceive this to be a risk of some problem happening” but until it happens it’s just conjecture. You certainly can’t say it is 7 times more dangerous than something else purely based on cogitating over possible complexities and generating fallacious numbers to match them. 

 

That is in fact is “somebodies method”. I forget who the name is but it’s a psychological trick that is used by people to make it look as if they know what they are talking about. 

 

The basic principle is you must convert a general principle into a number even if you have no source of or evidence to support that number.  Numbers equal certainty. 

 

That gives the impression you actually have solid evidence.

 

Next - Then you must make sure that number is not a round number because round numbers seem fake but unusual numbers seem likely to be real. So 7 sounds more real than 10 times. 

 

Its even better if the number has a decimal place because that increases the perception of accuracy. 

 

So I disagree absolutely  with you that it is reasonable for CASA to look at complexities in isolation of past events and come up with a  number of 7 times the risk compared to another private flight. It is just statistical absurdity! 

 

So using your logic, what would trigger action by CASA? How many patients have to die? 10, 50, 100, or in frequency of PPL flown fatal Angel Flight crashes, 5, 10, 20, 40, 100?

 

 

 

[mnewbery]

How many fatal crashes would you need to see to consider it significant?

 

Ask @Manwell

 

 

[Yenn]

They counted each flight and compared it with the number of flights done by the rest of GA.

 

I don't know how long the average Angel Flight flight is but I am sure it would be a lot longer than my average flight. A quick look at my MR for the Corby shows 413 hours and 1023 landings. 25 mins per flight.

 

 

[turboplanner]

They counted each flight and compared it with the number of flights done by the rest of GA.

 

I don't know how long the average Angel Flight flight is but I am sure it would be a lot longer than my average flight. A quick look at my MR for the Corby shows 413 hours and 1023 landings. 25 mins per flight.

 

I don't know the average, but would they would be around three hours. Closer than that and it's just about faster by car.

 

 

[facthunter]

They analyse the nature of the accident in detail. You could get good data out of only one. Like for instance RPT had cancelled flights from that location due fog and low cloud too degraded for qualified pilots to operate in. at that time.  Air Ambulance were having a bad run due single pilot and long duty times  and very marginal weather at one stage. Low numbers make comparisons a bit dicey but the situation is open to criticism with Angel flight since it has obvious deficiencies that need to be addressed. Nev