Introduction The UK Sellafield workforce is important for studying potential health risks of plutonium (Pu) exposure. However, several hundred early workers, employed during the period 1952–63, have been excluded from epidemiological studies because their urinalysis results were insufficiently reliable to provide accurate exposure assessment. This project aimed to develop a job-exposure matrix (JEM) that would enable future inclusion of these workers in epidemiological studies.

Methods 630 plutonium workers without reliable Pu urinalysis data for 1952–63 were identified within fourteen ‘homogeneous’ plutonium exposure groups. For each job/work location/year, ‘exposure analogues’ with reliable urinalysis data were identified (n=330). The JEM was based on 4487 work history records and 6899 urinalysis results. Intake assessments were produced using the ‘PuMA’ plutonium mass assessment code employing the latest conventional assessment methodology.

Results The JEM provided estimates for the median plutonium intake in becquerel (Bq) per year for each job/work location/year combination, and ranged from ”no intake” to 175 Bq/yr. Cumulative plutonium intakes for these workers ranged from ”no intake” to 990 Bq. Internal cross-validation indicated moderate-to-good correlations (r>0.4) and relative differences between JEM and validation sample <10%. Probabilistic evaluation indicated robust estimates of cumulative intake. Median cumulative JEM intake was 50 times lower than for conventional assessment methodology and much better aligned with prior expectation.

Conclusions The ‘exposure analogues’ methodology in JEM-development is a novel approach and has the potential to be a valuable tool for future epidemiological studies of the risks that may arise from plutonium exposure at Sellafield and potentially other similar cohorts.

Declaration of potential conflict of interest: Dr MacGregor, Mrs Wilson, Mr Peace and Mr Herdnan are employed by Sellafield Ltd. Professor Wakeford does consultancy work, including for the UK Compensation Scheme for Radiation-linked Diseases. The authors declare that they otherwise have no actual or potential competing financial interests.