Study population and setting
The study objective was to describe changes in death trends related to COVID-19 by occupational status among California residents. The study obtained data from the California Department of Public Health for all deaths occurring between January 2016 and October 2020 for California residents ages 18 to 65 years. They used the occupation text field and abstracted data via an automated program from the National Institute for Occupational Safety and Health, converting them to 2010 US census occupational codes across 13 sectors. They also described 4 racial/ethnic groups: non-Hispanic Asian, non-Hispanic Black, Latino, and non-Hispanic white. Individuals who identified as multiracial for Asian, Black, or white were excluded. The study used time-series analyses using dynamic harmonic regression with autoregressive integrated moving average errors for each occupational sector and stratified by race/ethnicity. They forecasted the number of deaths for unit of time and obtained the total number of excess deaths in the time window. They also generated risk ratios for mortality during the pandemic vs. pre-pandemic time.
Summary of Main Findings
In California, authors estimated 10,047 excess deaths from March to October 2020, with a risk ratio of 1.22 (95% Predictive Interval [PI]: 1.20 – 1.24) during the pandemic compared to the non-pandemic period. The absolute excess deaths were highest in the unemployed or occupation missing category (1,969 deaths, 95% PI: 1718 – 2225), but the highest risk ratio was among food or agricultural workers (Excess deaths: 1,050, Risk ratio: 1.39, 95% PI: 1.32 – 1.48). Disaggregated by race, Latino populations had the highest risk ratio (1.36, 95% PI: 1.29 – 1.44), followed by Black populations (1.28, 95% CI: 1.24 – 1.33) comparing pandemic to non-pandemic times. Finally, cooks had the highest risk ratio among census occupational codes (RR: 1.60), followed by packaging and filling machine operators (RR: 1.59) and miscellaneous agricultural workers (RR: 1.55).
The study used data from the California Department of Public Health to track deaths for nearly all California residents during the study period. They were also able to examine deaths among specific occupations with a high amount of granularity, for instance, drawing inferences between cooks as opposed to bakers and bartenders. This is important for public health planning around specific conditions workers face in an occupational setting that may impact their infection risk. Additionally, they were able to disaggregate this by both occupation and race/ethnicity, which provides important insights into the health disparities Californians face. They also had many years of data in the past (since 2016) to establish a baseline with which to compare the current death trends from 2020 during the pandemic.
There may be misclassification in occupational status when abstracting from an open text field. The primary occupation is usually reported by the next-of-kin and there may be errors in recall or in what is recorded by officials. Additionally, retirement was not considered in this analysis, and so reported occupation may not reflect current occupational risks among retirees, as occupation is defined as “type of work done most of working life.” Multiracial individuals were also removed from the analysis, which means these results do not describe a wholly accurate picture of racial/ethnic disparities given the diversity of the population. Also, it is impossible to determine if exposure and infection occurred in occupational settings or if other factors associated with both occupation and mortality may account for some of the differences between occupational groups, such as where each socializes or depending on how “tight-knit” a community is and whether infections spread among similar occupations outside of work. Similarly, the study examines excess deaths due to any reason, not specifically due to COVID-19. Therefore, there may be other factors leading to increased excess deaths during the pandemic such as disruptions to healthcare and livelihood.
This study is one of the largest to examine occupational differences in excess deaths with further stratification by race/ethnicity and occupational code. It provides one of the most comprehensive pictures to date of potential occupational risks associated with the pandemic.
This review was posted on: 26 March 2021