Safety Belt Usage and Effectiveness in Reducing Fatalities in

Document Sample

```					    Safety Belt Usage and Effectiveness
in Reducing Fatalities
in Passenger Cars with Air Bags

Demet Ozkan, D.Sc Candidate

Carl E. Nash, Ph.D
FHWA/NHTSA National Crash Analysis Center, The George Washington University
Goal :

To determine safety belt usage and
effectiveness in potentially fatal crashes of
passenger cars and light trucks that have full
front air bags.

2
Overview:
Current Methodology – Nash 1998

•   the distribution of fatal crashes according to safety belt usage
and the potential for being killed as a function of safety belt
use.

•   the concept of a potentially fatal crash is one where an
unbelted occupant would be killed, but a belted occupant
would survive.

•   generalization of the “Double Pair Comparison” method.

3
Earlier Studies:
One of the more widely used is the Double Pair
Comparison (DPC) method by Evans (1986).
•   Aims to discover how some characteristic affects the occupant’s fatality
risk.

•   The method focuses on vehicles containing two specific occupants, at
least one being killed.

•   One as the subject occupant.

•   The other, the control occupant, serves to standardize conditions to
estimate risk to the subject occupant.

4
Earlier Studies: DPC Method
A Simple Example:

•   The aim: to determine how driver gender influences driver
fatality risk.

•   The subject occupant: a car driver.

•   The control occupant: a male passenger seated in the right-
front seat.

5
Earlier Studies: DPC Method
Two sets of crashes:
Subject Group   Control Group

The first contains cars with a female driver and
a male passenger, at least one being killed.

The second contains cars with a male driver and
a male passenger, at least one being killed.

6
Earlier Studies: DPC Method
From these crashes the following
quantities are calculated:                 Subject Group   Control Group

A: Number of female drivers killed in cars with
male passengers

B: Number of male passengers killed in cars with
female drivers

C: Number of male drivers killed in cars with
male passengers

D: Number of male passengers killed in cars with
male drivers

7
Earlier Studies: DPC Method
A: Number of female drivers killed
in cars with male passengers
Female driver to male passenger
B: Number of male passengers                        ratio:
killed in cars with female                     r1 = A/B
drivers

C: Number of male drivers killed
in cars with male passengers
Male driver to male passenger
D: Number of male passengers                        ratio:
killed in cars with male drivers               r2 = C/D

8
Earlier Studies: DPC Method
Dividing these two ratios gives :

• Measures the risk of death to a female driver compared with the
risk of death to a male driver, other factors being essentially the
same.

• The crash conditions are effectively standardized because the
female and male drivers experienced their injuries in a mix of
crashes that posed similar risks to accompanying male passengers

9
Current Methodology: Nash 1998
The Universe of Crashes :

Classification: based on what happens to driver only…

1.   Crashes in which restraint use has no effect because an occupant would survive
in any case.

2.   Crashes in which restraint use makes the difference between death and survival.

3.   Crashes that are so severe that a driver would not survive even using this
restraint.

Unit :

…is a passenger car, light truck, or van involved in a crash.
10
Current Methodology: Classification
that are restraint
Crashes in which so severe
All Drivers who survived                All Drivers who were killed
thatmakes effect because
use has no the difference
a driver would not
between death and
an occupant would the
survive even using                                                         Survivable
Survivable
if Restrained              Not Survivable

survival.
survive in any case.
restraint.
Unrestrained           Restrained        Unrestrained            Restrained

Unrestrained                                              DI        EI              FI        I

Survivable            Restrained                                              DII        EII             FII      II
Right Front Passengers
who survived                                      Restrained              No Fatalities                   DIII      EIII            FIII      III

Survivable if Restrained   Unrestrained       AIV            BIV         CIV        DIV       EIV             FIV       IV

Unrestrained        AV               BV       CV         DV         EV              FV       V
Right Front Passengers
who were killed             Not Survivable          Restrained        AVI            BVI         CVI        DVI       EVI             FVI       VI

No Right Front Passenger                                                       No Fatalities                   DVII      EVII            FVII      VII

A           B            C           D         E               F
1                    2                          3
11
Current Methodology: Classification
If a restraint system is more effective,                                All Drivers who survived                All Drivers who were killed
the result will be that some of the
crashes that would have been in                                                                       Survivable
Survivable                if Restrained              Not Survivable
column E or F move to column D or C,
respectively.
Unrestrained           Restrained       Unrestrained            Restrained

Unrestrained                                             DI        EI              FI        I

Survivable            Restrained                                             DII        EII             FII      II
Right Front Passengers
who survived                                      Restrained              No Fatalities                  DIII      EIII            FIII      III

Survivable if Restrained   Unrestrained       AIV            BIV        CIV        DIV       EIV             FIV       IV

Unrestrained       AV                BV      CV         DV         EV              FV       V
Right Front Passengers
who were killed             Not Survivable          Restrained        AVI            BVI        CVI        DVI       EVI             FVI       VI

No Right Front Passenger                                                       No Fatalities                  DVII      EVII            FVII      VII

A           B            C          D         E               F

12

Rt
Effectiveness = 1 -             x 100%
Ru

Rt = rate of loss or injury for the treated population.
Ru = rate of loss or injury for the untreated population.

Applies to a situation where we measure a characteristic in
two randomly parts of a population, one of which is treated
and one is not.

13
Effectiveness:
Based on these…
…we define the effectiveness of the restraint system as:

In with the full drivers had been
Or other words, if allsummations: restrained (treated), only those
in columns C and D would have survived, while if none had been
restrained, all drivers in columns C, D, E, and F would have died.

14
Effectiveness:
…Let us define the ratio of restrained to unrestrained occupants within
the two most serious classes of crashes:

Where: ά = F/E
ß = C/D

If ά = ß we get:

Where:
K = no. of killed drivers
D = total no. of drivers
subscripts indicate belted or unbelted status

This is essentially the Double Pair
Comparison Formula…
15
Current Methodology:
Limitations:
Data used for estimating effectiveness of safety belts comes from
the Fatal Accident Reporting System (FARS).

Fatalities by Characteristics as Reported in FARS :

All Drivers who survived          All Drivers who were killed

Unrestrained       Restrained      Unrestrained      Restrained

Unrestrained                                                                         1

Right Front Passengers who survived       Restrained         No Fatalities - Not in FARS                                      2

Unrestrained                                                                         3

Right Front Passengers who were killed     Restrained                                                                          4

No Right Front Passenger                       No Fatalities - Not in FARS                                      5

K                 L                M                 N
A B+C                  D+E                        F
16
Current Methodology:
A New Solution to the Effectiveness Problem:

…Let us partition the data in a way that makes use of information about
restraint use:

i.     u = proportion of drivers who were restrained

ii.    v = proportion of r.f. passengers who were restrained
when with an unrestrained driver

iii.   w = proportion of r.f. passengers who were restrained
when with a restrained driver

Note: the u’s, v’s and w’s will be different for each set of cells
(uo : survivable, ui : survivable if belted, u : not survivable)

Note: u(1-w) + uw + (1-u)(1-v) + (1-u)v = 1
17
Current Methodology:
A New Solution to the Effectiveness Problem:                                                                         N1

All Drivers who survived                      All Drivers who were killed

Survivable

L3                              Survivable                      if Restrained            Not Survivable

Unrestrained               Restrained                Unrestrained            Restrained
(1-ui)(1-v)      (1-u)(1-v)    u(1-w)
Unrestrained                                                                                              I
(1-ui)v          (1-u)v        uw
Survivable      Restrained                                                                                               II
Right Front Passengers                                                                                 (1-ui)v          (1-u)v        uw
who survived                         Restrained                   No Fatalities                                                               III
Survivable
(1-uo)(1-v)       uo(1-w)         ui(1-w)      (1-ui)(1-v)      (1-u)(1-v)    u(1-w)
if Restrained   Unrestrained                                                                                              IV
(1-uo)(1-v)       uo(1-w)         ui(1-w)      (1-ui)(1-v)      (1-u)(1-v)    u(1-w)
Unrestrained                                                                                              V
Right Front Passengers       Not
(1-uo))v          uow             uiw          (1-ui)v          (1-u)v        uw
who were killed        Survivable      Restrained                                                                                               VI

A                  B             C              D               E             F

18
K                           L                           M                      N
Current Methodology:
A New Solution to the Effectiveness Problem:

(1 – ui) N5                 (1 – u) N5
E =1-                                =1-
uM5 + (u – ui) N5               uM5

…original paper for detailed explanation…

19
Data Analysis using Current Methodology:
•   Data for this study comes from the Fatal Accident Reporting
System (FARS)

•   Accident years 1995 through 2006.

•   The vehicles studied were 1995 and later passenger cars and
1999 and later light trucks.

•   With few minor exceptions, this includes only vehicles that are
equipped with both driver and right front passenger air bags.

•   Cases with ambiguous belt use or vehicle identification were
not included in the study.

20
Data Analysis using Current Methodology:
Results:
Numbers of 1995-2006 PASSENGER CARS in which people were killed as drivers
and/or right front passengers.

unrestrained        restrained         unrestrained    restrained
unrestrained                               805       1932      1066      405            I
7394      772        426     3719            II
restrained        793        8412         1641      171        142     1238           III
1985        916          179       429        355      135           IV
unrestrained      1020        471          116       279        656      249            V
restrained        408        4325         1069      112        262     2289           VI

20372      11571     6720    18570       no r.f. pass

A          B            C         D          E        F

21
Data Analysis using Current Methodology:
Results:
Numbers of 1999-2006 LIGHT TRUCKS in which people were killed as drivers
and/or right front passengers.

unrestrained        restrained     unrestrained           restrained

unrestrained                               941   1140        95   144                  I
7458   625         52   1142                II
restrained
477        3919        1934   162         21   455                 III
870        494          244   295         38    57                 IV
unrestrained
150        85            99   120         41    62                 V
restrained        82         673          784    66         22   488                 VI

2639   7351       555   4831        no r.f. pass.

A          B            C     D          E     F

22
Data Analysis using Current Methodology:
Results     for Passenger Cars:

Belt effectiveness:

•   58 percent for drivers who had a belt use rate of 76 percent.
•   59 percent for right front passengers who had a belt use rate of 75 percent.

Over the 12 years of this study:

•   Safety belts saved a total of 44,000 passenger car front seat occupants, but could

•   A total of 47,000 died in passenger cars who could not have survived even had
they been wearing safety belts.

•   Car driver belt use was only 68 percent when they were traveling with no right
front passenger.
23
Data Analysis using Current Methodology:
Results for Light Trucks:

Belt effectiveness:

•   84 percent for drivers who had a belt use rate of 82 percent.
•   77 percent for right front passengers who had a belt use rate of 79 percent.

Over the 8 years of this study:

•   Safety belts saved a total of 21,000 light front seat occupants, but could have

•   A total of 10,600 died in light trucks who could not have survived even had they
been wearing safety belts.

•   Light truck driver belt use was only 49 percent when they were traveling with no
right front passenger.

24
Discussion:
Comparison:
Effectiveness                      Belt Use Rate

Driver       Right Front   Driver with    Driver w/o     Right Front
Passenger     Passenger      Passenger      Passenger

Passenger     with air bags   58%               59%         76%            72%            75%
Cars
w/o air bags    63%               57%         45%            42%            42%

Light Trucks   with air bags   84%               77%         85%            82%            79%
& Vans
w/o air bags    81%               71%         49%            48%            42%

Effectiveness
…was not significantly different between vehicles with and without air
bags.

25
Discussion:

•   The high belt use rate is probably a function of the relatively
young age of the vehicles involved. It is well known that belt
use in older vehicles, particularly pickups and SUVs, is
significantly lower than belt use in newer ones.

Effectiveness                      Belt Use Rate

Driver       Right Front   Driver with    Driver w/o     Right Front
Passenger     Passenger      Passenger      Passenger

Passenger     with air bags   58%             59%           76%            72%            75%
Cars
w/o air bags    63%             57%           45%            42%            42%

Light Trucks   with air bags   84%             77%           85%            82%            79%
& Vans
w/o air bags    81%             71%           49%            48%            42%

26
Discussion:
The high belt effectiveness in light trucks is probably due to two factors:

•    the higher proportion of rollovers in these type vehicles where
safety belts can reduce ejection.

•    the fact that in crashes between light trucks and other vehicles, the
light truck generally suffers a lower change in velocity (ΔV) and
less occupant compartment intrusion.
Effectiveness                       Belt Use Rate

Driver        Right Front   Driver with    Driver w/o     Right Front
Passenger     Passenger      Passenger      Passenger

Passenger     with air bags   58%              59%           76%            72%            75%
Cars
w/o air bags    63%              57%           45%            42%            42%

Light Trucks   with air bags   84%              77%           85%            82%            79%
& Vans
w/o air bags    81%              71%           49%            48%            42%

27
Discussion:
• These results are consistent with other estimates of belt use and
effectiveness, and give confidence that belt use is as important in vehicles
with air bags as in pre-air bag vehicles.
Effectiveness                      Belt Use Rate

Driver       Right Front   Driver with    Driver w/o     Right Front
Passenger     Passenger      Passenger      Passenger

Passenger      with air bags   58%             59%           76%            72%            75%
Cars
w/o air bags    63%             57%           45%            42%            42%
Pre-Airbag

Light Trucks   with air bags   84%             77%           85%            82%            79%
& Vans
w/o air bags    81%             71%           49%            48%            42%

• These results do not, however, provide any indication of the
effectiveness of air bags themselves. However, other studies suggest that
passenger cars have become safer in the last twenty years despite little
reduction in rollover fatalities where air bags would provide little or no
protection.
28
Discussion:
…results show:
…that an additional 30,000(total) lives could have been saved
had all light vehicle occupants been wearing safety belts.

…the economic consequences of these deaths amounted to
more than \$100 billion.

…improved safety belt use reminder systems are needed in all
new vehicles – which would have cost a total of \$4 billion
dollars at \$25 per vehicle – as one of the most cost-effective
means of improving motor vehicle safety even if such systems
only cut non-use in half.

29
Thank you…

30
Current Methodology:
Since we have                                             All Drivers who survived                All Drivers who were killed

limited knowledge…                                          Survivable
Survivable
if Restrained              Not Survivable

Unrestrained           Restrained       Unrestrained            Restrained

Unrestrained                                             DI        EI            FI          I

Survivable            Restrained                                             DII        EII             FII      II
Right Front Passengers
who survived                                      Restrained              No Fatalities                  DIII      EIII            FIII      III

Survivable if Restrained   Unrestrained       AIV            BIV        CIV        DIV       EIV             FIV       IV

Unrestrained       AV                BV      CV         DV         EV              FV       V

Right Front Passengers
who were killed             Not Survivable          Restrained      AVI              BVI        CVI        DVI       EVI           FVI         VI

No Right Front Passenger                                                       No Fatalities                  DVII      EVII         FVII         VII

A           B            C          D         E               F
31

```
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
 views: 5 posted: 2/23/2012 language: English pages: 31