Critique of: Crash Risk in General Aviation

Below is a first draft of my critique of this report. Any
suggestions, error corrections, or other critique is welcome.
=====================================


The report below was funded by federal grants, and created by medical
researchers at Johns Hopkins University. Their report's noble mission
is an attempt to provide guidance in mitigating fatalities resulting
from aviation operations conducted by other than military and airline
operators.

Because of the researchers' apparent unfamiliarity with the segment of
aviation they chose as the focus of their report, I personally find
the authors' implied causations and conclusions to be less than
astute. While the researchers may be adequately qualified to assess
medical issues, their report is flawed in its analysis due to their
apparent unfamiliarity with aviation.

Because of the hysterical treatment of General Aviation activities in
the news media prior to the terrorist attacks of September 11, 2001,
and its increase in shrill sensationalism post 9/11, I feel compelled
to rebut many of the notions put forth in this report. To distinguish
my critique from the text of the original report, I will enclose my
words within square brackets.


--------------------------------------------------
Crash Risk in General Aviation

Guohua Li, MD, DrPH
Susan P. Baker, MPH


IN THE AFTERNOON OFOCTOBER 11, 2006, A PRIVATE PLANE crashed into
an apartment complex in Manhattan, killing the pilot, New York
Yankees pitcher Cory Lidle, and his flight instructor Tyler
Stanger. The impact destroyed the 4-seat, single-engine aircraft
and set the building on fire. The crash scene brought aviation
safety back to national headlines.1

[The choice of this accident for the opening of this publicly funded
report created by ostensibly erudite academicians is unfortunate. This
accident was sensationalized in the news media because it involved a
celebrity, and it took the form familiar to those who witnessed the
9/11 terrorist attacks on the World Trade Center towers. In reality it
was not significantly more remarkable than any other pilot error
mishap. Publicly trotting out this tragic incident evokes emotional
reaction in a report that should be professional and factual, not
sensational, and may reveal a certain prejudiced mind-set or bias on
the part of the researchers.]

In this article, we examine the crash risk of private flights,
identify major factors influencing survival in aviation crashes,
and discuss possible approaches for improving the safety of
general aviation.

[While the report's stated mission is a noble, I take issue with the
authors' intent to limit it to "private flights." There is no formal
definition of a "private flight." There are FAA certified airmen who
are Private Pilots, and as such they are prohibited by federal
regulations from receiving compensation for flying; but the report's
scope is obviously not limited to flights conducted by them as is
evident by the paragraph below.]

Crash Rates
Civilian aviation generally can be divided into 2 groups:
commercial and noncommercial flights.2 Commercial flights
transport individuals and goods to generate revenue; they include
operations of major airlines, commuter air carriers, and air
taxis. Noncommercial flights, usually called general aviation,
encompass a wide array of activities—emergency medical services
(EMS), sightseeing, flight training, traffic reporting, aerial
surveys, search and rescue, crop dusting, firefighting, logging,
recreation, and personal or business use. General aviation
aircraft range from small private airplanes and business jets to
helicopters, hot-air balloons, and gliders.

[This paragraph reveals the researchers' lack of understanding of the
definition of General Aviation. Air Taxi, pipe-line and power-line
patrol, crop dusting, and air charter flights all generate revenue,
are piloted by airmen holding FAA Commercial or Airline Transport
Pilot certificates, and they are all General Aviation operations. In
fact, other than military aviation operations and airline (Code of
Federal Regulations Title 14 Part 121) operations, all aviation
operations are classified under the General Aviation designation. To
assert that medical rescue helicopter ambulance services, flight
training, traffic reporting, aerial surveys, and crop dusting are
noncommercial is ridiculous.]

Currently, there are approximately 228 000 active private pilots
and 220 000 registered general aviation aircraft in the United
States; 93% of the aircraft are planes, 4% are rotorcraft, and 3%
are nonmotorized craft such as gliders.3 From 2002 through 2005,
general aviation, with an annual average of 1685 crashes and 583
deaths, comprised 91% of all aviation crashes and 94% of all
aviation fatalities.4 The fatal crash rate for general aviation,
1.31 fatal crashes per 100 000 flight hours, is 82 times the rate
for major airlines (0.016).3 This difference in crash rates has
persisted over many decades.

[This statement, while probably true, misleads the reader into the
belief that all aspects of General Aviation are _inherently_ more life
threatening than airline flight, because of its attempt to compare two
dissimilar classes of aircraft operation.

Consider airline flights:

* Long distance legs require lots of hours but only one takeoff
and one landing, those being accepted as the most hazardous
phases of flight.
* Two professional pilots at the controls
* An FAA certificated Dispatcher on the ground influencing flight
decisions.
* Able to fly above the weather
* ...


General Aviation flights:

* Short distance legs mean many more landings and takeoffs are
performed per hour than on airline routes.
* Usually only one pilot at the controls.
* Often the pilot holds only a student certificate.
* During training flights, which constitute a large percentage of
GA flight hours, the corners of the flight envelope are
routinely explored.
* Flights conducted entirely within the Troposphere where weather
exists.
* ...]

Risk Factors for Crash Involvement
Due to their relatively small aircraft size and low altitude,
general aviation flights are especially vulnerable to adverse
weather conditions.

[The above statement, while generally true, overlooks the fact that
many large aircraft are operated under the General Aviation
designation. For example, actor John Travolta flies a Boeing 707,
aviation advocacy groups fly B-17 Flying Fortresses, the air cargo
operators fly large airlines fitted for freight hauling; all are large
aircraft flown in General Aviation operations.]

Flight procedures vary with weather conditions. Visual flight
rules regulate procedures for flight under visual meteorological
conditions (defined as a ceiling of 1000 feet and 3 miles of
visibility), with the guiding principle of “see and avoid.”

[Actually, Instrument Flight Rules (IFR) apply to all flights on an
Instrument Flight Plan regardless of whether the weather meets Visual
Metrological Condition (VMC) minima or Instrument Metrological
Conditions (IMC). Further, see-and-avoid is mandated by federal
regulations for _all_ flights, IFR or VFR, conducted in VMC.]

Flight under reduced visibility is governed by instrument flight
rules, for which the navigation and control of the aircraft are
performed using instruments.

[That is true, but the vast majority of IFR operations occur in VMC.
Of course, the only time the aircraft is piloted solely by reference
to instruments is during IMC.]

Although commercial flights are almost always operated under
instrument flight rules, general aviation pilots often fly under
visual flight rules and may not have the necessary training for
flying under instrument meteorological conditions.

[As I have pointed out earlier, many General Aviation flights are
commercial in nature, because someone is paying to have them
performed. The researchers probably meant to use the word 'airline'
instead of 'commercial' here.

It is true that most airmen who hold a Private FAA certificate are not
instrument rated. However nearly all airman who hold a Commercial or
Airline Transport Pilot certificate are instrument rated by the FAA to
operate in IMC.]

For pilots without instrument training, flying from visual flight
rules into instrument meteorological conditions is a perilous
scenario.

[There are a miniscule number of airmen who hold FAA certificates,
that have not received any instrument training; instrument training is
not required to obtain a Glider certificate. The phrase the
researchers probably meant to use was 'instrument rating' not
'instrument training.'

Regardless, it is true that the average life expectancy of a pilot who
is not instrument rated and qualified (recent experience) is a bit
over a minute when unintentionally finding himself in a cloud that
totally obscures his outside reference.]

A case-control study revealed that having been initially
licensed after age 25 years and not having an instrument rating
(ie, not being qualified for flying under instrument flight rules)
are each associated with a 4-fold increased risk of being in a
general aviation crash in instrument meteorological conditions.5

[I'm having difficulty parsing the above sentence. The use of the
verb 'are' above implies the plural. Are the researchers saying, both
being licensed after the age of 25 years and not having an instrument
rating _each_ increase the risk of a mishap by 400% if they should
attempt to conduct a flight in IMC for a total increase in risk of
800%. I hesitate to presume to infer the intended meaning of the
researchers statement.]

Partly reflecting inadequate training and flight experience, pilot
error is a contributing factor in 85% of general aviation crashes
compared with 38% of airline crashes.6

Other environmental factors (eg, airport features, wires, and
terrain) also play an important role in general aviation safety.
Flying is especially hazardous in Alaska, where the crash rate per
flight hour for general aviation is nearly 3 times the national
average.7

A considerable body of research literature on pilot
characteristics and crash risk exists.8 Alcohol-impaired flying is
a well-established risk factor for general aviation crashes.

[Alcohol poses the same risk for airline operations, but because there
are usually two pilots at the controls and occasional medical testing,
it is less likely to occur.]

In the 1960s, alcohol reportedly was involved in more than 30% of
fatal general aviation crashes.9 Experimental studies conducted in
flight simulators indicate that alcohol, in doses as low as 0.02
g/dL, can impair piloting skills, such as the ability to detect
angular motion and changes in the oil pressure gauge.10 The
impairment in pilot performance increases with blood alcohol
concentrations (BACs) in a doseresponse fashion. In a study
involving actual flights, Billings et al11 reported that when BACs
reached the level of 0.12 g/dL, pilots lost control of the
aircraft in 16 of 30 flights. Consequently, the Federal Aviation
Administration (FAA) has implemented alcohol education programs
and adopted a zero-tolerance alcohol policy.

[I am unaware of any FAA alcohol education programs for other than
airline pilots, and I've held an airmans certificate since 1970.]

Currently, federal aviation regulations prohibit any person from
acting as a crew member within 8 hours after consuming any
alcoholic beverage or while having a BAC of 0.04 g/dL or higher.
The legal alcohol limit for pilots, 0.04 g/dL, was promulgated in
1985 when it was considered the lowest level that could be
reliably measured by testing equipment.9

In 1990, the FAA amended regulations regarding background checks
on pilots for alcohol-related motor vehicle convictions, requiring
pilots to provide a written report of each alcohol-related traffic
offense within 60 days of the conviction. Flying privileges can be
suspended or revoked if a pilot has had 2 or more convictions for
driving under the influence in the past 3 years. A recent cohort
study indicated that a history of driving while intoxicated is a
valid risk marker for general aviation pilots. After adjusting for
age, sex, and flight experience, the study showed that a history
of driving while intoxicated was associated with a 43% increased
risk of aviation crash involvement.12 Following intensive research
and interventions, the proportion of alcohol involvement in fatal
general aviation crashes has decreased progressively from more
than 30% in the early 1960s to about 8% today.13

Sudden incapacitation of the pilot is a critical safety issue for
general aviation flights, which, unlike commercial flights,
usually do not have a co-pilot who could take control of the
aircraft in an emergency. Cardiovascular disease, particularly
acute myocardial infarction, is the leading cause of in-flight
incapacitation. The incidence rate of sudden incapacitation due to
cardiovascular disease for general aviation pilots is estimated to
be 1.7 crashes per 100 000 pilots per year and increases with
pilot age.14 However, less than 1% of general aviation crashes are
attributable to sudden incapacitation resulting from medical
emergencies.14

Other known or suspected risk factors for general aviation crashes
include pilot inexperience,5,12 older age,12 male sex,12
nonconformist flying behavior (measured by an index of
noncompliance with generally accepted flying procedure),15 and
prior aviation crash and violation records.5,16


Risk Factors for Crash Fatality
Most general aviation crashes do not result in fatalities. Factors
influencing occupant survival in aviation crashes have been
studied extensively.17-20 Emerging from these studies are 4 major
environmental and pilot-related risk factors for crash fatality:
aircraft fire, instrument meteorological conditions, off-airport
location, and failure to use safety restraints. Aircraft fire is
the single most important determinant of occupant survival in
aviation crashes, regardless of the type of flight or aircraft. In
one study, the crash fatality rate (defined as the proportion of
crashes resulting in
1 deaths) for general aviation crashes was
15% in the absence of fire and 69% when there was a fire.19
Aircraft fire is involved in 13% of general aviation crashes but
accounts for 40% of crash fatalities; the adjusted odds ratio of
pilot fatality associated with aircraft fire is 14 in general
aviation crashes.19

The risk of fire after a crash can be reduced through appropriate
aircraft design. Crash-resistant fuel systems, designed to sustain
high-impact forces without rupture and leakage, have virtually
eliminated postcrash fire and thermal fatalities in US Army
helicopter crashes.21 This technology is also effective in
preventing fires when applied to civil helicopters, although to a
lesser extent than in US Army helicopters due to a weaker standard
for civil helicopters.22

[If aircraft fuel systems can be made crash resistant without imposing
too great a cost or weight penalty, they should be. But I don't seem
them being offered on the most recently certified GA aircraft.]

Adverse weather conditions increase the chance of a crash and are
important determinants of crash outcome. General aviation crashes
occurring in instrument meteorological conditions are more likely
to be fatal than crashes in visual meteorological conditions.

[This is likely due to loss of control resulting in speed that exceeds
the design limits of the aircraft structure or the rapid onset of
icing conditions.]

Although representing only 9% of general aviation crashes,
instrument-condition crashes account for 28% of pilot
fatalities.19 Adverse weather may increase the risk of fatality in
aviation crashes in several ways. First, crashes occurring in
conditions of degraded visibility may involve considerably greater
impact forces than crashes in visual conditions because the pilot
has less warning of impending impact. Second, instrument
conditions may hamper search and rescue efforts. And third,
extremely low or high temperatures may pose a significant risk to
crash survivors, particularly those injured, while waiting for
rescue.

The risk of fatality following a crash also depends on the crash
location. Overall, 46% of general aviation crashes occur at
airports.19 The crash fatality rate for general aviation crashes
occurring away from airports is 36% compared with 6% for crashes
at airports.19 Like weather, location may influence survival
through several pathways. Off-airport crashes are more likely to
involve high-velocity, uncontrolled impacts than on-airport
crashes.

[I'd have to see credible evidence that supports that allegation
before I'd accept it.]

Locations away from airports may lengthen and severely complicate
search and rescue attempts, including fire-fighting and EMS.

Not wearing safety restraints, including lap belts and shoulder
restraints, is another risk factor for pilot fatality. A study of
commuter and air taxi crashes found that those not wearing
shoulder restraints were nearly 4 times as likely to die as those
wearing them.17 Research has confirmed that safety restraints are
also a significant protective factor for pilots in general
aviation crashes.18 Recently, seatbelt airbags have become
standard equipment in many new general aviation aircraft. The
devices, available also as retrofit kits, combine airbags with
restraint systems that have integrated lap belts and shoulder
belts and offer improved protection for the head and neck.23

[A pilot who flies without the use of shoulder restraint belts is a
fool.

It is curious that the researchers failed to mention ballistic
parachute recovery systems like those currently mandated for the
recently FAA certified Cirrus aircraft.]

The general aviation crash fatality rate has remained at about 19%
for the past 20 years while the overall airline crash fatality
rate has declined from 16% from 1986 through 1995 to 6% from 1996
through 2005.4,24

[Due to the reduction in airline operations due to the September 11,
2001 terrorist attacks, increased airport security, and general
decline in airline ticket sales, that statistic may be misleading.]

The higher fatality rate for general aviation crashes may be
because such aircraft are not as able to withstand impact forces
and protect occupants from death and severe injury as commercial
aircraft are.

[A more robust airframe requires increased weight. There is a
tradeoff of safety for performance.]

In recent decades, while major airlines have improved seat
strength, revised exit row configurations, and used more fire
retardant materials, few improvements have been made in general
aviation aircraft, in part, because federal regulations only
require safety improvements for entirely new aircraft models. A
corresponding policy for automobiles would have meant that
Volkswagen Beetles could have been sold without seatbelts for
decades after federal regulation required them in all new cars.

[The Volkswagen analogy is flawed. The ubiquitous Cessna 172 aircraft
have had should restraints for decades despite their first being FAA
certified in the 1950s.]

General Aviation and Public Safety
General aviation accounts for the vast majority of aviation
crashes and casualties.

[That is because there are over ten times the number of GA aircraft
than there are airline aircraft.]

Although crash rates have decreased somewhat, the crash fatality
rate of general aviation has not changed in the past 20 years.
Since the September 11, 2001, attacks, aviation safety efforts
have centered on improving aviation security, including the
security of small airports and airstrips used primarily by general
aviation.

Besides being a public safety concern, general aviation intersects
with medicine directly in at least 2 ways. First, transporting
patients from crash sites and between medical facilities is more
hazardous than generally recognized, and EMS flight crew members
have an occupational injury death rate that is 15 times the
average for all occupations.20 Despite 1 EMS helicopter in 3 being
likely to crash during a life span of 15 years, few EMS
helicopters have crash-resistant fuel systems.20 Second, physician
pilots crash at a higher rate per flight hour than other pilots.25
It is possible that physicians are more likely than other pilots
to buy high-performance aircraft that require more time for
mastery than their schedules may allow. In addition, physicians
may take risks (eg, fly when fatigued or in bad weather) in order
to meet the demands of a busy medical practice. From 1986 through
2005, a total of 816 physician and dentist pilots were involved in
general aviation crashes; of them, 270 (33%) were fatally injured.
Physician and dentist pilots accounted for 1.6% of all general
aviation crashes and 3.0% of pilot fatalities (Carol Floyd, BS,
National Transportation Safety Board, written communication,
February 2, 2007).


Conclusions
In summary, general aviation crashes are a little-recognized
public safety problem even though they account for the great
majority of aviation deaths.

[Little recognized by whom? Ask the average lay citizen, and he'll
tell you "them little planes are dangerous."]

To improve the safety of general aviation, interventions are
needed to improve fuel system integrity and restraint systems,
enhance general crashworthiness of small aircraft,

Those are only viable measures if their added weight and cost do not
so negatively impact aircraft performance and affordability so as to
render General Aviation operations impractical.]

and reduce weather-related crashes through pilot training and
avionics technology.

[When someone invents a better method of training pilots, I am
confident it will be implemented. Avionics technology, especially
Global Positioning System equipment, has already begun to provide
pilot situational awareness enhancements, and the NASA/FAA Capstone
and Small Aviation Transportation System projects are poised to
revolutionize GA flight operations.]

The FAA and the National Transportation Safety Board should place
high priority on reducing general aviation crashes and allocate
adequate resources for developing and implementing effective
intervention programs.

[Faced with limited budget and exponential growth in airline
operations, the FAA has other priorities that take precedence.]

Financial Disclosures: None reported.

Funding/Support: This work was funded in part by grants R01AA09963
and R01AG13642 from the National Institutes of Health and grant
CCR302486 from the Centers for Disease Control and Prevention.

Role of the Sponsor: The funding agencies had no role in the
preparation, review, or approval of the manuscript.

REFERENCES
1. Baron J. Plane crash in Manhattan. New York Times. October 12,
2006; 1A.
2. McCormick BW, Papadakis MP. Aircraft Accident Reconstruction
and Litigation. Tucson, Ariz: Lawyers & Judges; 1996:501-522.
3. Bureau of Transportation Statistics. National transportation
statistics.

http://www.bts.dot.gov/publications/...ble_01_13.html.
Accessed January 17, 2007.
4. National Transportation Safety Board. Accidents, fatalities,
and rates, 1986 through 2005, US General Aviation.
http://www.ntsb.gov/aviation/Table10.htm. Accessed January 17,
2007.
5. Groff LS, Price JM. General aviation accidents in degraded
visibility: a case control study of 72 accidents. Aviat Space
Environ Med. 2006;77:1062-1067.
6. Li G, Baker SP. Factors associated with pilot error in aviation
crashes. Aviat Space Environ Med. 2001;72:52-58.
7. Kearney PJ, Li G. Georgraphic variations in crash risk of
general aviation and air taxis. Aviat Space Environ Med.
2000;71:19-21.
8. Li G. Pilot-related factors in aircraft crashes: a review of
epidemiologic studies. Aviat Space Environ Med. 1994;65:944-952.
9. Gibbons HL. Alcohol, aviation, and safety revisited: a
historical review and a suggestion. Aviat Space Environ Med.
1988;59:657-660.
10. Cook CC. Alcohol and aviation. Addiction. 1997;92:539-555.
11. Billings CE, Wick RL Jr, Gerke RL, Chase RC. Effects of ethyl
alcohol on pilot performance. Aerosp Med. 1973;44:379-382.
12. Li G, Baker SP, Qiang Y, Grabowski JG, McCarthy ML.
Driving-whileintoxicated history as a risk marker for general
aviation pilots. Accid Anal Prev. 2005;37:179-184.
13. Li G, Baker SP, Lamb MW, Qiang Y, McCarthy ML. Characteristics
of alcoholrelated fatal general aviation crashes. Accid Anal Prev.
2005;37:143-148.
14. Booze CF Jr. Sudden inflight incapacitation in general
aviation. Aviat Space Environ Med. 1989;60:332-335.
15. Urban RF. Comparative analysis of social, demographic, and
flight-related attributes between accident and nonaccident general
aviation pilots. Aviat Space Environ Med. 1984;55:308-312.
16. Li G, Baker SP. Prior crash and violation records of pilots in
commuter and air taxi crashes: a case-control study. Aviat Space
Environ Med. 1994;65:979-985.
17. Li G, Baker SP. Crashes of commuter aircraft and air taxi
crashes: what determines pilot survival? J Occup Med.
1993;35:1244-1249.
18. Rostykus PS, Cummings P, Mueller BA. Risk factors for pilot
fatalities in general aviation airplane crash landings. JAMA.
1998;280:997-999.
19. Li G, Baker SP. Correlates of pilot fatality in general
aviation crashes. Aviat Space Environ Med. 1999;70:305-309.
20. Baker SP, Grabowski JG, Dodd RS, Shanahan DF, Lamb MW, Li G.
EMS helicopter crashes: what influences fatal outcome? Ann Emerg
Med. 2006;47: 351-356.
21. Shanahan DF, Shanahan MO. Injury in US Army helicopter crashes
October 1979-September 1985. J Trauma. 1989;29:415-422.
22. Hayden MS, Shanahan DF, Chen L-H, Baker SP. Crash-resistant
fuel system effectiveness in civil helicopter crashes. Aviat Space
Environ Med. 2005;76:782-785.
23. AmSafe Aviation. Inflatable restraint technology.
http://www.amsafeaviation.com/inflatablega.htm. Accessed February
23, 2007.
24. National Transportation Safety Board. Accidents, fatalities,
and rates, 1986 through 2005, for US air carriers operating under
14 CFR 121, scheduled service (airlines).
http://www.ntsb.gov/aviation/Table6.htm. Accessed January 17,
2007.
25. Booze CF Jr. Epidemiologic investigation of occupation, age,
and exposure in aviation accidents. Aviat Space Environ Med.
1977;48:1081-1091.