Retrospective occupational exposure assessment for case-control and case-series epidemiology studies based in Shanghai China.

To provide exposure information for epidemiology studies conducted in Shanghai from 2001 to 2008, we completed retrospective exposure assessments (EA) of benzene and other hazards. Interviewers administered questionnaires to subjects from Shanghai area hospitals. An initial exposure screening by EA staff members, blinded as to case-control status, stratified jobs into exposed, unexposed, or uncertain categories prior to review by a separate expert panel (EP). Resources for the EA included job/industry-specific questionnaire responses by subjects, short-term benzene area concentration measurements from a Shanghai regulatory agency database, Chinese literature for qualitative and short-term quantitative measurements, on-site investigations, summaries of technology changes, and selected task simulations with concurrent benzene concentration measurements. An EP in Shanghai completed semi-quantitative benzene exposure assignments, with categories of 0 to 4 corresponding to intensity ranges of none, <1, 1 to 10, >10 to 100, and >100 mg/m(3). For other hazards, sources included the EP's knowledge of the industries and Chinese and Western literature. For benzene, 20% of the EAs selected by a stratified random process were evaluated by two alternate methods. The study database of potential cases and controls included 18,857 jobs from the subjects' work histories. From 818 individuals initially screened as probably benzene exposed, 964 jobs underwent further review. From subjects with final diagnoses, 755 jobs qualified for inclusion in the final database for any study. For other exposures, the EA considered 17,893 jobs from 7654 subjects for possible exposures and were in the final study database. Of these, 2565 individuals had exposures of study interest from their 4909 exposed jobs. The prevalent exposures included agricultural chemicals, petroleum products, and metals. The EA involved extensive information assembly and exposure assignment by an EP and periodic reviews. The methods described went beyond those typically applied in past general population studies and may have provided improved information for the epidemiologic analyses. However, sufficient, reliable measured historical data are lacking to evaluate this conclusion.

Equipment, exposure, emission review--specification of process equipment for worker exposure control.

Worker exposure to chemical agents may occur from equipment leaks in closed processes and from open system activities and maintenance (e.g., sampling, tank gauging, line breaking, equipment flushing, and drainage). To prevent worker overexposure to chemical agents, an Equipment, Exposure, Emission Review (EEER) was developed to consistently match equipment properties to the leakage-related inhalation risk posed by each stream. For streams where gas or liquid leakage could cause worker exposure above occupational exposure limits (OEL), the EEER recommended a high performance, low leakage equipment category. Conversely, where standard engineering offered reliable containment for lower health risk streams, the EEER did not recommend over-control. The EEER matched equipment to stream health hazard as follows: (1) the composition of each stream was determined, with particular attention to chemical substances with stringent exposure limits; (2) a mixture exposure limit was calculated for each stream based on stream composition and the OEL for stream constituent chemicals; (3) each stream was classified as to its respiratory exposure hazard on the basis of the stream exposure limit; (4) equipment was recommended as a function of respiratory exposure hazard class using an Equipment Selection Matrix. Equipment options were based, in part, on the emission performance of the equipment and a near-field dispersion model that was used to relate equipment emissions to an OEL. Over a 5-yr implementation period, nearly 1700 streams of 78 refining process units were reviewed. Standard engineering practice was selected for about 70% of the streams reviewed. Benzene, hydrogen sulfide, ethanolamine, and high boiling aromatic oil streams were the primary chemical agents responsible for more stringent controls. Although the EEER criteria for stream classification and control options were arranged in order of magnitude--a form of control banding--the correct selection of control technology was based on detailed assessment of stream composition and risk, rather than on process unit name or function. The EEER and Equipment Selection Matrix provided reproducible recommendations and helped manage cost and complexity of equipment specification for health protection.