1972). Brit. J. industr. Med., 29,[394][395][396][397][398][399][400][401][402][403][404][405][406] Mortality of gasworkers-final report of a prospective study. The mortality experience of selected groups of gasworkers employed by four area Gas Boards and observed over a period of eight years was described by us in a report in 1965. The present paper adds a further four years' data to those previously collected for men having regular exposure in coal carbonizing plants and for men having exposure only to by-products of the gas-making process. Tothesewehave addeddata relating to men employed by four additional area Gas Boards who have been observed over periods of seven to eight years.

The mortality of selected groups of gasworkers has been observed over a period of eight years, and a comparison has been made of the mortality from different causes among different occupational groups. Men were included in the study if they had been employed by the industry for more than five years and were between 40 and 65 years of age when the observations began. All employees and pensioners of four area Gas Boards who met these conditions were initially included; but the number was subsequently reduced to 11,499 by excluding many of the occupations which did not involve entry into the carbonizing plants or involved this only irregularly. All but 0·4% of the men were followed successfully throughout the study. Mortality rates, standardized for age, were calculated for 10 diseases, or groups of diseases, for each of three broad occupational classes, i.e., those having heavy exposure in carbonizing plants (class A), intermittent exposure or exposure to conditions in other gas-producing plants (class B), and such exposure (class C). The results showed that the annual death rate was highest in class A (17·2 per 1,000), intermediate in class B (14·6 per 1,000), and lowest in class C (13·7 per 1,000), the corresponding mortallity for all men in England and Wales over the same period being slightly lower than the rate for class A (16·3 per 1,000). The differences between the three classes were largely accounted for by two diseases, cancer of the lung and bronchitis. For cancer of the lung the death rate (3·06 per 1,000) was 69% higher in class A than in class C; for bronchitis (2·89 per 1,000) it was 126% higher. For both diseases the mortality in class B was only slightly higher than in class C, and in both these categories the mortality was close to that observed in the country as a whole. Three other causes of death showed higher death rates in the exposed classes than in the unexposed or in the country as a whole, but the numbers of deaths attributed to them were very small. The death rate from cancer of the bladder in class A was four times that in class C, but the total number of deaths was only 14. Five deaths were attributed to pneumoconiosis, four of which occurred in bricklayers (class B). One death from cancer of the scrotum occurred in a retort house worker. For other causes of death the mortality rates were similar to or lower than the corresponding national rates. Examination of the data separately for each area Board showed that the excess mortality from lung cancer and chronic bronchitis in retort house workers persisted in each area. For two Boards the mortality from other causes was close to that recorded for other men living in the same region; in the other two Boards it was substantially lower. A comparison between the mortality of men who worked in horizontal retort houses and of those who worked in vertical houses suggested that the risk of lung cancer was greater in the horizontal houses and the risk of bronchitis was greater in the vertical houses, the differences being, however, not statistic...

Between February and May 1984, we conducted a pilot study to examine the methods for a larger study of a previously reported relation between Reye's syndrome and medications. Thirty patients with Reye's syndrome, whose diagnosis was confirmed by an expert panel, and 145 controls were matched for age, race (black or not black), and antecedent illness (respiratory infection, chickenpox, or diarrhea) and selected from the same hospital, emergency room, or school, or identified by random digit dialing. Significantly more cases (93 per cent, 28 of 30) than members of each of the four control groups or all controls combined (46 per cent, 66 of 145) had received salicylates during matched antecedent illnesses (odds ratio of all 30 cases vs. all controls = 16.1; lower 95 per cent confidence limit = 4.6). The prevalence and mean severity score of signs, symptoms, and selected events during the antecedent illness tended to be lower among cases than controls. Thus, differences in the severity of this illness between cases and controls did not explain differences in medication exposures. This pilot study suggests an association between Reye's syndrome and the use of salicylates during an antecedent illness.