Exposure measures, accident data

The subcommittee reviewed data that came primarily from human experimental studies and from toxicity studies in various animal species. The evaluation focused on inhalation exposure studies that measured respiratory irritation and tolerance to odor. Human case studies, accident reports, and epidemiologic studies of industrial exposures were extensive but of limited use to the subcommittee because they lack quantitative exposure measurements. Controlled human experiments were most important to the subcommittee for establishing the SEALs for ammonia. There appears to be a broad range of sensitivity to ammonia s pungent odor and in irritation caused by exposures to low concentrations... 

Although accident statistics are able to give valuable information about accidents as well as influencing factors (at least to some extent), the findings must still be interpreted with care, as their true meaning is only revealed when related to exposure [32]. Many studies using accident data do not consider risk exposure or discuss the correct measure for exposure with respect to the research question [34]. 

The reports reviewed used consistent measures of exposure. Epidemiologieal studies of aviation accidents typically incorporate some measure of exposure (Li, 1994), such as flight hours documented during an investigation or self-reported flying hours. It is important to have a measure of exposure to interpret the relative risks associated with particular activities. In the reports on GA accident data reviewed during this project, the common measure of exposure was the total number of hours flown by GA per 100,000 hours. 

One very detailed analysis of pedestrian crash, injury, and fatality risk that considered the different exposure measures was conducted by Keall (1995). For his analysis he combined travel exposure data from the 1989- 1990 New Zealand Travel Survey with pedestrian accident data from the New Zealand national Traffic Accident Report files for the period 1988-1991. The travel survey data for children 5-9 years old were obtained from interviews with the parents or other adults in the same household. Their casualty data as a function of the pedestrian age is plotted in Figure 15-2. If we look first at the absolute numbers of pedestrians injured or killed as a function of age and gender we see the expected high numbers of young (5-19 years old) pedestrians. We also see that more males than females are injured or killed. 

Of course, the absolute numbers present only part of the picture. Exposure to accidents has also changed. Passenger miles have declined, employee numbers have fallen drastically, and highway traffic has increased. Data for some important exposure measures are given in table B3 in appendix B. The next sections of this chapter combine exposure measures with the fatality and injury data to produce casualty rates for employees, highway-crossing users, trespassers, and passengers. 

All fire smoke is toxic. In the past two decades, a sizable research effort has resulted in the development of over twenty methods to measure the toxic potency of those fire smokes (6). Some methods have been based on determinations of specific chemical species alone. Values for the effect (e.g., lethality) of these chemicals on humans are obtained from (a) extrapolation from preexisting, lower concentration human exposure data or from (b) interpretation of autopsy data from accident and suicide victims. The uncertainty in these methods is large since ... 

Human toxicity data are limited to secondary citations. Because these citations provided no experimental details, they cannot be considered reliable. Deaths have occurred from aniline ingestion and skin absorption, but doses were unknown. Reviews of the older literature indicate that a concentration of 5 ppm was considered safe for daily exposures, concentrations of 7 to 53 ppm produced slight symptoms after several hours, a concentration of 40 to 53 ppm was tolerated for 6 h without distinct symptoms, a concentration of 130 ppm may be tolerated for 0.5 to 1 h without immediate or late sequalae, and 100 to 160 ppm was the maximum concentration that could be inhaled for 1 h without serious disturbance. In studies of accidents with unknown exposure concentrations, methemoglobin levels of up to 72% were measured. Recoveries occurred with a minimum of medical intervention following cessation of exposure. 

Recommendation 5c. The Army and its operating contractors should automate as much as feasible important medical information related to worker exposure to facilitate epidemiological studies. Automated information, available at the programmatic level, should include, but should not be limited to, results of medical examinations, evaluations of exposure to agents, measurements of cholinesterase levels, heat-stress data, and accident/injury information. 

Children s Susceptibility. A limited number of human studies have examined health effects of CDDs in children. Data from the Seveso accident suggest that children may be more susceptible to the dermal toxicity of 2,3,7,8-TCDD (chloracne), but it is not known if this would be the case for other effects. Follow-up medical surveillance of the Seveso children (including measurement of serum 2,3,7,8-TCDD levels) would provide information on whether childhood exposure would pose a risk when the individual matures and ages. The available human and animal data provide evidence that 2,3,7,8-TCDD can cross the placenta and be transferred to an infant via breast milk. Although information on the developmental toxicity of CDDs in humans is limited, there are extensive animal data that the developing... 

Fisher et al. (1991) Biokinetic Model A modified biokinetic model for uranium was developed for inhaled soluble uranium based on human data from an accidental release of uranium hexafluoride in Oklahoma. Urinary excretion data from 31 exposed workers were used to test two previously published compartmental models for inhalation exposure to uranium (ICRP 1979 Wrenn et al. 1989). Urinary uranium was measured periodically for 2 years following the accident. Statistical analysis showed that the Wrenn et al. (1989) model produced a better fit to the excretion data than the ICRP (1979) model. 

The early phase is defined by the time period during which there is the threat of a significant release. The time interval between the recognition of an accident sequence and the start of the release can extend from about half an hour to about a day and the duration of the release may be between half an hour and several days. In this phase the information based on the analysis of data and predictions being from the nuclear installation and some limited environmental measurements of off-site exposure rates and airborne concentrations from the plume may become available. Thus, the... 

The first step in designing a road safety development index (RSDI) is to come up with a comprehensive set of indicators, which includes as far as possible aU the main parameters in road safely of human-vehicle-road-enviromnent-regulation, instead of considering a few factors such as accident rates per population or per kilometoe driven. In addition, this index should be as relevant as possible for different countries, especially in developing countries. The choice of accident risk and exposure variables is necessary to what is available in international data and what is considered necessary for meaningfiil comparisons. Commonly, frequencies of aimual numbers of vehicles, accidents, injured and killed people are some kind quantification and relatively easy to define and to measure in different countries. But differences in definitions, noncollection of data, non-rehabUity of data and under-reporting are problems for effective measurements of road safety. 

As classical data collections are limited to accidents, one way to get data on non-collision events is a stochastic variation of accident reconstruction data in a way that the single event does not necessarily result in an accident anymore. These non-collisions are then used in the simulation in order to assess the balance between desired and undesired effects of a measure in traffic [ 15]. As a consequence, validating the non-collisions regarding the distribution of key parameters and their representa-tivity for overall traffic is vital. The basic data concerning exposure are not as well known as accident related data. 

In the event of an accident at a nuclear or other facihty, environmental sampling and subsequent measurements should be performed to provide data on the levels, time dependence and spatial distribution of radionuclides in air, soil, plants, foodstuffs and feedstuffs so as to assess doses to critical groups of the population and to support decisions on mitigation and protective actions. Sampling should therefore be representative with regard to the exposure conditions of the critical group (see paras 6.18-6.22). 

ABSTRACT Occupational risk rates per hour of exposure are quantified for 63 occupational hazards for the Dutch working population. Data were obtained from the analysis of more than 9000 accidents that occurred over a period of six years in the Netherlands and resulted in three types of reportable consequences under Dutch law a) fatal injury, b) permanent injury, c) recoverable injury requiring at least one day hospitalisation. A Bayesian uncertainty assessment on the value of the risk rates has been performed. Annual risks for each of the 63 occupational hazards have been calculated including the variability induced by the variability of the annual exposure of the working population to these hazards. Suitability of three risk measures, individual risk rates, individual annual risk and number of accident occurred is examined and commented upon. 

One shortcoming of the Maycock et aL (1991) and the Forsyth et al. (1995) studies was that experience was measured in months of driving rather than the more commonly accepted measure of exposure miles or kilometers of driving. However, in an analysis of data collected in the Netherlands approximately ten years later, and using accidents per billion kilometers driven, Vlakveld (2004) obtained a very similar pattern of results, which are presented in Figure 6-7. Another advantage of the Vlakveld results as presented in Figure 6-7 is that the line plots are of actual empirical crash data rather than of functions derived from a data-based model. 

---