Skip to main content

Clinical Characteristics and Mortality of Healthcare Workers with SARS-CoV-2 infection in Mexico City

Our take —

This large study using publicly-available data from Mexico City revealed a lower risk of death among HCW compared to non-HCW, with nurses experiencing the lowest risk. These findings were independent of confounding factors and matching on important covariates. These results could also be explained with frequent use of PPE among HCW in and outside health care institutions, easy access to health care, higher quality of health care after being admitted to the hospitals compared to the general public, and better health status among HCW compared to the general public, as residual confounding was present. Risk factors for mortality among HCW were similar to those observed among non-HCW in other studies (male sex, age, and the presence of co-morbidities).

Study design

Retrospective Cohort

Study population and setting

This was a retrospective cohort study characterized the mortality risk and factors associated with mortality among health-care workers (HCW) relative to non-health care workers (Non-HCW) with laboratory-confirmed SARS-COV2 infection in Mexico City between February 27 and August 31, 2020. Data were ascertained passively from Mexico City’s Government database on September 9, 2020. Confirmed cases were defined as those who had evidence of SARS-COV2 by PCR plus 2 of the following: fever, cough or headache within 7 days prior being tested and had at least one of the following: chest pain, rhinorrhea, dyspnea, arthralgia, myalgia, sore throat or conjunctivitis. Potential risk factors for mortality (age, sex, occupation, type of institution of care, municipality of residency, date of symptom onset, date of first evaluation, and several comorbid conditions; diabetes, hypertension (HTN), chronic obstructive pulmonary disease (COPD),asthma, immunosuppression, cardiovascular disease (CVD), chronic kidney disease (CKD), obesity, smoking status, treatment setting (outpatient versus inpatient), and outcome (invasive ventilation, intensive care unit admission or death)) were assessed using multivariate logistic regression. Analyses were restricted to hospitalized participants to prevent potential bias associated with HCW having better access to screening and diagnostic services (and thus render less severe cases more likely to be included) than non-HCW. To ensure appropriate adjustment for important covariates, the authors also matched a group of hospitalized non-HCW with hospitalized HCW on age, sex, comorbidities, smoking status, state of residency and time from symptom onset to first evaluation and repeated the analysis using multivariate logistic regression. Analyses were conducted separately for type of HCW (physicians, nurses, other HCW).

Summary of Main Findings

125,665 patients were identified through the database. 16,445 (13.1%) were HCW and 109,219 (86.9%) were non-HCW. HCW were more likely to be female, younger in age and free of comorbid condition (61.1% vs 46.6% , median age of 38 years vs 45 years, 66.5% vs 60.6%. 25,771(20.5%) subjects were treated as an inpatient, 24,461(94.4%) were non-HCW VS 1310 (5.1%) HCW. 9,951 (38.6%) of the hospitalized subjects died, 39.6% were in non-HCW VS 19.3% were in HCW (p<0.001). In the adjusted model, HCW were less likely to die compared to Non-HCW (OR 0.53).This finding was consistent within different types of HCW (nurses, OR 0.29, physicians, OR 0.6,other HCW, OR 0.61). Overall, the likelihood of mortality was significantly more among males than females(OR 1.6),among those>50 years of age (OR 2.31), and among those with DM, HTN, or obesity (OR 1.26). Smoking and place of residency were not associated with mortality. After matching 1310 hospitalized non-HCW with the hospitalized HCW on age, sex, place of residency, smoking status, and several comorbid conditions, the analysis revealed similar results in reduction in the mortality risk among HCW compared to non-HCW with a greater protection among nurses compared to other HCW.

Study Strengths

The study used a large public database of confirmed SARS-COV2 infection over a six months period to evaluate the differences in several characteristics between infected HCW and non-HCW. Since this database included data pertaining to many counties inside Mexico City, the study results could be generalizable to the rest of Mexico City and could be possibly extrapolated to countries with similar structures and settings. The study used laboratory-confirmed cases only which prevents the likelihood that non-diseased participants were misclassified as having COVID-19.


Though the study utilized a large public database, the quality of the data was not validated actively and potential errors in data entry could have occurred. The data had many missing variables, which could alter the results of the study if these variables were very different from the included ones. A presidential order in Mexico City in March, 2020 offered a paid leave for high risk health care workers. This could preferentially render most of the COVID-19 HCW in low risk groups. The authors considered some important confounding factors as binary rather than continuous in their analyses such as HTN, DM and obesity, which could have led to inadequate adjustment. Though the authors mentioned that they had data available in regards to several other comorbid conditions such as COPD, HIV, cancer, CVD and CKD, they didn’t mention if they accounted for these comorbidities in their analysis or in their matching. Finally, the authors mentioned that COVID-19 related mortality differs between the healthcare institutions in Mexico city, but they didn’t mention if they matched HCW and non-HCW on the health care institution.

Value added

The study provides important data about the risk of death among SARS-COV-2 infected HCW compared to non-HCW using data from a large public database in Mexico City.

This review was posted on: 20 November 2020