Historical cohort investigation of spontaneous abortion in the Semiconductor Health Study: epidemiologic methods and analyses of risk in fabrication overall and in fabrication work groups.

The risk of spontaneous abortion (SAB) in the semiconductor industry was examined in a historical cohort study of pregnancies at 14 companies. We identified female employees who had worked for at least 6 months and whose ages ranged from 18 to 44 years during the 1986-1989 study period. Using company records, we included all fabrication-room (fab) employees and an approximately equal number of nonfabrication (nonfab) employees, for a total sample of 7,269. Telephone interviews with 6,088 women (84%) identified 904 eligible pregnancies and 113 SABs. Exposure classification was based on questionnaire and industrial hygiene assessments of tasks the women performed during the first trimester of pregnancy. Using logistic regression to control for age, smoking, ethnicity, education, income, year of pregnancy, and stress, we found a higher risk of SAB in fab employees than in nonfab employees (15.0% of fab pregnancies ended in SAB vs. 10.4% of nonfab pregnancies, adjusted relative risk [RR] = 1.43, 95% CI = 0.95-2.09). Analysis of fab work groups showed that the highest relative risk was in masking employees (17.5% SAB rate, adjusted RR = 1.78, 95% CI = 1.17-2.62 in comparison with nonfab employees). Within masking, the highest risk was found in etching-related process employees (22.2% SAB rate, adjusted RR = 2.08, 95% CI = 1.27-3.19 in comparison to nonfab employees.