Crash causes

Anon., CISHC Chem. Safety Summ., 1974-5, 45—46, 2-3 Two horns after a road tanker had crashed, causing the load of liquid ethylene to leak, one of the tyres of the tanker burst and ignited the spill, eventually causing the whole tanker to explode. The tyre failed because it froze and became embrittled, and it is known that such failure of steel-braced tyres gives off showers of sparks. This could therefore be a common ignition source in cryogenic transportation spillage incidents. 

The role of speeding as a direct crash cause was probably first analyzed in a detailed and comprehensive manner by Treat et al (1977). In this study, described in detail in Chapter 17, a representative sample of more than 2,000 police-reported crashes was analyzed by crash investigators at the crash sites, and 420 of them were further analyzed by multidisciplinary teams. A cause was defined as an event or action whose absence would have prevented the crash, all other things being equal. Furthermore, a human cause was cited if the causal behavior was a deviation from the normal or expected behavior of the average driver. Thus, speed would not be cited in a crash of a speeding vehicle imless the speed deviated from the speed expected at that site under the conditions that prevailed and the crash would not have occurred had the speed been as expected. With this approach to causation, the study estimated excessive speed to be a definite cause in 7-8 percent of the crashes and a probable cause in an additional 13-16 percent of the crashes. 

Further analyses on U.S. and European data have all confirmed the very strong associations between speed and injury severity. Bowie and Waltz (1994) analyzed U.S. national data and showed that while speed is cited as a crash cause in 10.2 percent of property-damage-only crashes, and in 14.6 percent of crashes with minor (non-incapacitating) injuries, it is cited in 34.2 percent of all fatal crashes. More objective analyses that do not rely on possibly-biased clinical assessments of the role of speed have also demonstrated the relationship in a much more direct manner, as discussed below. 

Figure 14-11. The incidence of various types of crashes caused by sleep or fatigue in Norway (fi om Sagberg, 1999, with permission fi om Elsevier).

Causes of child and teen pedestrian crashes. On the basis of the information in Table 15-5 and other findings we can characterize the crash-causing behaviors at different ages as follows. The youngest pedestrians, 1-2 years old, rarely venture out into the street by themselves, and when they are involved in an accident they are more likely to be struck by cars backing up on driveways and private property. Slightly older children, 3-9 years old, are more likely to collide with vehicles when they dart out into the traffic stream (often from between parked cars) because they do not yet have safe pedestrian skills. 

Table 15-5, The primary pedestrian crash causes for pedestrians at different ages (based on Hunter etal.y 1995).

Table 16-3. Percent of primary crash-causing human errors attributed to the riders and the drivers in the MAIDS (ACEM, 2004).

Because most clinical investigations of crash causation have assumed the information processing model and not the motivational model, most crash causes described below will be framed in terms of failures in information processing, though some references to motivational factors will occasionally be made as well. 

The study yielded several significant findings, which at the time were quite unexpected. First, the researchers discovered that in approximately 50 percent of the crashes more than one factor caused the crash. This means that in a crash caused by two or more factors, the absence of any one of the factors would have prevented the crash. This also means that most often two factors or more had to occur at the same time for a crash to happen. 

Figure 17-5, The percent of crashes caused by human, environment, and vehicle causes, at the probable and definite levels, based on the on-site and in-depth crash analyses, and (bottom) the relative proportions of combinations of these causes. Because crashes can have more than one cause, the totals add up to more than 100% (from Treat et aL, 1979).

Figure 17-6. A comparison of the percent of crashes caused by human, environment, and vehicle defects or failures in England (upper number in each circle) and in the U.S. (lower number in each circle) (from Rumar, 1985, with kind permission of Springer Science and Business Media).

Even the use of police reported violations should be considered with some reservations. Often summary statistics of these violations are misrepresented as reasons for the crash or crash causes. For example, Israel s National Traffic Police publishes an annual report that contains the reasons for crashes . The most fi equent reasons cited (in that order) are deviation from lane, failure to maintain safe distance, failure to yield right of way, failure to stop at a stop sign, and failure to yield at a yield sign (Israel Police, 2001). Inattention - the most common cause of crashes according to in-depth investigations - does not appear as a cause because it is not a directly observable violation of the traffic code. 

Using speed as a crash cause example was not done by chance. It was picked because speed is one behavior or human factor that can be (relatively easily) observed both in crash data and in traffic flow data. Environmental and vehicle factors can also be obtained in both types of data and are prime candidates for evaluation via the statistical approach. Other measures are more difficult. For example, alcohol involvement in crashes can be assessed if alcohol levels are obtained from the crash involved drivers (not too difficult to do) and from the general traffic population (much more difficult since it requires random stopping of drivers to check for their alcohol levels - even if they did not commit any traffic violation). Still other behaviors are practically impossible to evaluate statistically. Unfortunately the behaviors that fall into this category are the most frequent ones to emerge in the clinical evaluations attentional, perceptual, and decision failures. They are difficult to ascertain in the crash data (and are typically absent in police crash reports), and absolutely impossible to determine from the traffic population data. 

Crash causes derived from epidemiological studies... 

With these limitations in mind, we can consider the role of various factors for which there is some theoretical basis to support their over involvement. The epidemiological approach, because it typically relies on existing data bases, does not account for all crash causes, or all... 

Given the implication of driver inattention as a major human crash cause in clinical in-depth analyses, and the recent demonstration of the crash risk from cell-phone related inattention (see Chapter 13), an interesting challenge in the 100-car study was to extract information on the causal roles of inattention and distraction fi-om these data without resorting to the drivers ... 

---