Case Series; Retrospective Cohort
Study population and setting
This study included children and adults (age ≥ 10 years) diagnosed with COVID-19 between January 20 and April 15, 2020 at 20 healthcare organizations belonging to TriNetX. TriNetX is a global health research network that provides real-time access to electronic medical records. Risk of mortality, mechanical ventilation, and hospitalization within 30 days after diagnosis were the primary outcomes of interest.
Summary of Main Findings
Among 5980 males and 7730 females in the study, 45% of males and 50% of females were from the US. Men and women exhibited significant differences in clinical and demographic information, including risk behaviors, comorbidities, and lab values. Propensity scores were generated using age, race, high-risk behavior, nicotine use, and comorbidities (diabetes, hypertension, chronic lung disease, cardiovascular disease, chronic kidney disease, etc.). Males and females were matched 1:1 using greedy nearest-neighbor matching, resulting in 5,350 within each group. Before and after propensity score matching, men were more likely to die, be hospitalized, and receive mechanical ventilation than women, though the differences were less pronounced after matching.
The study is a large, multi-site study with access to real-time data. The authors constructed and matched on propensity scores to control for potential confounding.
All data was queried from the electronic medical record, which is subject to coding errors and missing data. The authors do not mention missing data, so it can be assumed this was a complete case analysis, and this population may not fully represent those who are missing data on certain characteristics. Even with available data on laboratory parameters and symptom information, it does not seem this information was included in the propensity score or adjusted for in the analysis, which could result in incomplete control of confounding. It is unclear how the diagnosis of COVID-19 was made, and may have differed by site.
This propensity-score matched analysis attempts to address gender disparities in COVID-19 outcomes.