Efficiency of autocoding programs for converting job descriptors into standard occupational classification (SOC) codes

Abstract

Background Existing datasets often lack job exposure data. Standard Occupational Classification (SOC) codes can link work exposure data to health outcomes via a Job Exposure Matrix, but manually assigning SOC codes is laborious. We explored the utility of two SOC autocoding programs. Methods We entered industry and occupation descriptions from two existing cohorts into two publicly available SOC autocoding programs. SOC codes were also assigned manually by experienced coders. These SOC codes were then linked to exposures from the Occupational Information Network (O*NET). Results Agreement between the SOC codes produced by autocoding programs and those produced manually was modest at the 6-digit level, and strong at the 2-digit level. Importantly, O*NET exposure values based on SOC code assignment showed strong agreement between manual and autocoded methods. Conclusion Both available autocoding programs can be useful tools for assigning SOC codes, allowing linkage of occupational exposures to data containing free-text occupation descriptors.

Determining occupation for National Violent Death Reporting System records: An evaluation of autocoding programs

“…For analysis, we treated the manually coded occupation on the death certificate as if it were a gold standard, though we acknowledge that coding procedures may vary from state to state. In a prior comparison of manually coded versus auto‐coded job titles, SOC codes were assigned independently by two coders with differences resolved by consensus 7 . Our study observed similar agreement to this previous validation study, supporting the use of the death certificate code as a comparison standard.…”

“…Overall, we found good agreement for both the NIOCCS and SOCcer autocoding programs for assigning occupation codes that match manual coding. A previous study validating autocoding systems in two occupational cohort studies with self‐reported text‐based occupations found similar agreement with 64.0%–67.4% for major occupation codes for NIOCCS and 62.4%–72.3% for SOCcer 7 . For detailed occupation codes (6‐digit), we found a higher percent agreement than the former study and a lower percent agreement for major occupation (2‐digit) occupation codes.…”

“…For detailed occupation codes (6‐digit), we found a higher percent agreement than the former study and a lower percent agreement for major occupation (2‐digit) occupation codes. The previous study found percent agreement for detailed occupation codes less than 50% for both autocoding programs 7 . The underlying populations used to validate the autocoding programs were different in these two studies.…”

“…Autocoding of occupation with NVDRS data has been completed previouslypreviously 2 but has not been validated. Previous validation studies have examined agreement between occupation codes assigned by autocoding programs and occupation codes assigned by expert manual coders 7 . However, there is no published validation for using autocoding programs with information collected in NVDRS.…”

Not just a research method: If used with caution, can job-exposure matrices be a useful tool in the practice of occupational medicine and public health?

“…Users of JEM must consider the source of exposure data – expert assessments, data collected from individual workers, or environmental sampling. The coding of occupations is time consuming and can introduce error ( 9 ), and more testing of and comparison with automated job coding systems is needed ( 10 ). JEM reflect an “average” level of exposure within a job at the expense of individual variation.…”