Manufacturing processes at two French rapeseed oil companies: possible relationships to toxic oil syndrome in Spain.

The toxic oil syndrome (TOS) epidemic that occurred in Spain in spring 1981 has been associated with the consumption of rapeseed oil that was denatured with aniline for industrial use but diverted for human consumption. The precise aetiologic agent in the oil responsible for the outbreak has not been identified. To learn more about possible contaminants and how the contamination might have occurred, we visited two French companies that process rapeseed oil and that were identified in Spanish administrative and judicial records as the ones exporting aniline-denatured rapeseed oil to Spain in 1981. With the apparently full and voluntary co-operation of personnel at both companies, we reviewed the processes involved in manufacturing, treating and transporting rapeseed oil, and we have summarized the information provided to us. Of particular importance is the finding that oil exported to Spain was taken from stock, the rest of which was sold for human consumption in the French domestic market, apparently without any adverse health effects. The differences between the oil exported to Spain and the oil sold as food in France were that aniline equivalent to 2% of the weight of the oil was added to most of the Spanish oil but not to that sold in France, and that contamination of the Spanish oil may have occurred in the tank trucks used for transportation to Spain, which had previously carried industrial chemicals. There is no assurance that the trucks were cleaned appropriately for transporting a food product before the oil was loaded for the journey to Spain. Since the clinical manifestations of TOS are not those of aniline toxicity, we conclude that the aetiological agent of TOS is likely to be one of the following: (1) a contaminant in the aniline, (2) a contaminant introduced during transportation, (3) a reaction product of normal oil components or materials used in refining with either aniline or the potential contaminants mentioned under (1) or (2) above.

An epidemiological description of lightning-related deaths in the United States.

To better quantify and update the health impact of lightning and to compare potential sources of information, we reviewed data from the National Centre for Health Statistics (NCHS) database for 1968 through 1985 and from the North Carolina Medical Examiner (NCME) computerized database for 1972 through 1984. We epidemiologically characterized all lightning-related deaths identified in these databases. Results of our analysis are presented together with previously published information from the National Oceanic and Atmospheric Administration and from the National Institute for Occupational Safety and Health. In the US, Wyoming has the highest average annual lightning-related death rate (0.196/100,000), whereas Florida has the largest cumulative incidence of deaths (200) for this 18-year period. Death rates are the highest for males and for people 10-19 years old. There is a tendency toward a slight, but steady, decrease in the yearly incidence of lightning-related deaths. One-third of the fatalities are job-related. Both the NCHS and NCME databases are limited in describing the circumstances and aetiologies of these deaths.

Lightning-related mortality and morbidity in Florida.

Cases of lightning-related deaths and injuries that occurred in Florida in 1978-87 were reviewed to determine the factors involved, to quantify the morbidity and mortality related to lightning strikes, and to describe epidemiologically the injuries and circumstances involved. Statewide information on deaths was obtained from death certificates, autopsy reports, and investigative reports. Information about morbidity was obtained from the Florida Hospital Cost Containment Board data base and the National Climatic Data Center data base for all Florida counties, as well as from hospitals in selected counties. Lightning-related deaths totaled 101 in Florida during the period 1978-87, an annual average of 10.1. Eight percent of the victims were from other States. The overall yearly death rate for State residents was 0.09 per 100,000 population, with the highest rate being that for men aged 15-19 years, 0.38 per 100,000. Thirteen percent of victims were females. The ratio of lightning-related injuries to deaths in Florida was estimated at about four to one. Thirty percent of all deaths were occupationally related. The first strikes of lightning from a thunderstorm may be the most dangerous, not in terms of impact, but because of the element of surprise. During thunderstorms, people may seek shelter under isolated trees because they believe erroneously that a tree offers protection from lightning, or perhaps because their top priority is to escape from rain rather than lightning. People may not seek adequate shelter during thunderstorms because they do not know the dangers of remaining outdoors or their judgment is impaired by drugs or alcohol.

Skin test reactions to a phenol-soluble antigen of Brucella abortus among veterinary students, Lyon, France, 1984.

In December 1984, 424 students at the National Veterinary College, Lyon, France, were skin-tested with a phenol-soluble antigen of Brucella abortus B19. Of all students, 2.6% had a positive intradermal reaction indicating previous contact with Brucella. Prevalence of positive reactions was significantly lower among students from the three first school years (less than 2%) compared with students in their last school year (5.9%). These results are discussed and compared to the prevalence of brucellosis among cattle. Specificity (94%) of the intradermal testing resembles that of serological testing but its sensitivity (75%) seems to be better.

Injuries and risk factors for injuries from the 29 May 1982 tornado, Marion, Illinois.

On 29 May 1982, a tornado struck Marion, Illinois. As a result, ten people were killed, 38 were hospitalized, and 196 were examined in hospital emergency rooms. In order to describe injuries and to identify risk factors that might have increased or reduced the risk of injuries, we (1) conducted a telephone survey of all persons we could identify who were in the path of the tornado and (2) reviewed the emergency room logs and records and admission files of hospitals that provided initial care. Overall, 19.3% of people who were in the path of the tornado were injured during the impact. Only 39.0% of people examined in emergency rooms were injured as a direct result of the impact. The remainder were injured while rescuing, cleaning up, or just walking in the disaster area. Behaviours that were most protective were being in a house on the lowest floor, staying underground in a basement, protecting one's body with something like a blanket, and staying away from windows. Neither lying down nor opening doors or windows seemed to decrease the risk of being injured. Our results emphasize the need for early warnings and public education about protective behaviours.