Use of Private Sector Workforce Respiratory Disease Short-Term Disability Claims to Assess SARS-CoV-2, Mexico, 2020

We examined respiratory disease short-term disability claims submitted to the Mexican Social Security Institute during 2020. A total of 1,631,587 claims were submitted by 19.1 million insured workers. Cumulative incidence (8.5%) was 3.6 times higher than that for January 2015‒December-2019. Workers in healthcare, social assistance, self-service, and retail stores were disproportionately affected.

threshold by the 1-sided 95% CI around the arithmetic mean of the 30 highest preepidemic/postepidemic weekly rates and the EI (low, medium, high and very high intensity) by using the 1-sided 95% CI around the geometric mean of 30 highest weekly epidemic rates. We set CI limits at 40%, 90%, and 97.5%.
We estimated industry-, sector-and subsectorspecific cumulative RD-STDC incidence by dividing the sum of RD-STDC by the worker population registered at IMSS for each year and compared 2015-2019 (baseline) with 2020 by using risk ratios. Cumulative COVID-19 incidence was calculated by applying the percentage of workers with a positive SARS-CoV-2 test with the RD-STDC weekly number and dividing the cumulative sum of COVID-19 cases (n = 1,029,122) by the worker population. We calculated lost workdays (LWDs) by adding all authorized days and excess LWDs as the observed minus the expected number of LWDs in weeks above the threshold (90th percentile) for each week. We calculated proportions of LWD rates by dividing the number of workdays lost to RD-STDC per year by the total number of available workdays. This project was deemed a public health surveillance and did not require institutional review board approval. We used Stata version 15.1 (https://www.stata.com) for all graphical and statistical analyses.
Weekly RD-STDC rates increased sharply ( Figure  1) and reached a medium EI (first peak 35,707 RD-STDC reported); claims decreased but remained above medium EI level until May, when the slope increased to a maximum in week 29 (July 18, second peak 62,542 RD-STDC reported), crossing into the EI level of very high intensity. From this date, RD-STDC frequency decreased but remained 5-6 times higher than historical values until week 38 (September 13), when RD-STDC increased, reaching a third peak (71,785 RD-STDC reported) during week 51 (December 20). Up to week 53, a total of 1,543,600 cumulative RD-STDC were reported, with an estimated cumulative incidence of 8.0%. The EI of epidemic curves varied by industry sector and subsector ( Figure 2).
A total of 20% of workers who had an RD-STDC were tested for SARS-CoV-2; 59% were positive. The estimated number of symptomatic COVID-19 cases among IMSS-affiliated workers was 910,073 (cumulative incidence 5%). The median RD-STDC granted was 10 days (interquartile range 4-14 days), representing an excess of 17.4 million workdays lost compared with that during 2015-2019.
We compiled the cumulative RD-STDC incidence for different industry sectors (Table), which ranged from 3% for agriculture, forestry, fishing, and hunting to 10% for retail trade and manufacturing. When we compared these data with those for 2015-2019, we found that the total RD-STDC during 2020 were 2.9-5.0 times higher in all sectors. The amounts of total workdays lost (WDL) attributed to RD-STDC (Appendix Table, https:// wwwnc.cdc.gov/EID/article/28/1/21-1357-App1. pdf) were highest among the manufacturing sector (0.44%; WDL 6.10 million). Among subsectors, WDL ranged from 0.14% (WDL 0.24 million) for educational services, arts, entertainment, and recreation to 0.80% (WDL 0.64 million) for health care and social assistance.

Conclusions
Workers in multiple industry sectors and subsectors were disproportionately affected by the SARS-CoV-2 pandemic. The above-average cumulative incidence of RD-STDC among some subsectors suggests an occupational link to infection. Healthcare workers had a high cumulative RD-STDC incidence (15.6%). This incidence was similar for persons who worked near or frequently with other workers or who have frequent customer contact but are not exposed to SARS-CoV2 directly, such as workers in manufacturing, food, and retail. In contrast, lower attack rates were observed within the educational services, arts, entertainment, and recreation subsectors. In Mexico, after their closure on March 16, 2020, schools continued with distance learning activities until August 30, 2021, when in-person instruction was reestablished.
Real-time analysis of RD-STDC by sector and subsector enables detection of outbreaks and rapid implementation of mitigation measures such as contact tracing, reinforcing hand hygiene, universal masking, and social distancing. RD-STDC rates might also be used to guide vaccination priorities aimed at protecting workers at higher risk for infection.
Our study's first limitation is that COVID-19 cases among workers might have been underestimated because not all symptomatic workers requested RD-STDC. In a recent serosurvey of IMSS insured workers, 26%-40% of workers with SARS-CoV-2 antibodies and symptoms compatible with COVID-19 sought RD-STDC (D. Barros-Sierra Cordero et al., IMSS, pers. comm., 2021 Jan 10), suggesting that real symptomatic COVID-19 attack rates could be 3-4 times higher. Second, asymptomatic or minimally symptomatic workers infected with SARS-CoV-2 are not reflected in this analysis. Third, some RD-STDCs might have been related to diseases other than COVID-19. Fourth, not all workers included in our denominator sought care within IMSS facilities, leading to incomplete capture of RD-STDC and a potential underestimation of the magnitude of the epidemic. Fifth, we did not examine geographic differences in RD-STDC claims. It is possible that economic sectors and subsectors less affected were concentrated in areas of low epidemic intensity in the country, resulting in lower RD-STDC rates, which could account for the low rates observed in agriculture, forestry, fishing, and hunting.   Predictive models based on the RD-STDC surveillance system are currently used by the IMSS to alert large manufacturing sites and other businesses of possible workplace outbreaks. Information provided by this system complements other surveillance systems for monitoring epidemics and inform decision-making by health authorities. Our data highlight the need for including respiratory disease claims in pandemic preparedness planning.
This study was supported by the IMSS of Mexico.

About the Author
Dr. Hernandez-Avila is an epidemiologist and director of Economic and Social Benefits of the Mexican Social Security Institute, Mexico City, Mexico. His primary research interests are occupational health and disease prevention.