Mary Wilson
Well-known member
Published online April 6, 2023.
doi:10.1001/jama.2023.5348
Yan Xie, PhD[SUP]1[/SUP]; Taeyoung Choi, MPH[SUP]1[/SUP]; Ziyad Al-Aly, MD[SUP]1[/SUP]
In the first year of the COVID-19 pandemic, 2 US studies suggested that people hospitalized for COVID-19 had nearly 5 times the risk of 30-day mortality compared with those hospitalized for seasonal influenza.[SUP]1[/SUP][SUP],2[/SUP] Since then, much has changed, including SARS-CoV-2 itself, clinical care, and population-level immunity; mortality from influenza may have also changed. This study assessed whether COVID-19 remains associated with higher risk of death compared with seasonal influenza in fall-winter 2022-2023.
Methods
We used the electronic health databases of the US Department of Veterans Affairs (VA). Between October 1, 2022, and January 31, 2023, we enrolled all individuals with at least 1 hospital admission record between 2 days before and 10 days after a positive test result for SARS-CoV-2 or influenza and an admission diagnosis for COVID-19 or seasonal influenza. We removed 143 participants hospitalized with both infections. The cohort was followed up until the first occurrence of death, 30 days after hospital admission, or March 2, 2023. Differences in baseline characteristics between the groups were evaluated through absolute standardized differences (<0.1 indicating good balance).
We evaluated the risk of death in people hospitalized for COVID-19 vs influenza through inverse probability-weighted Cox survival models. Logistic regression was used to generate a propensity score, which was then applied in inverse probability weighting to balance the 2 groups; covariates were defined based on prior knowledge and ascertained in the 3 years preceding admission (eMethods in Supplement 1). We also estimated the absolute risk as the percentage of excess deaths (difference in death rates between COVID-19 and influenza groups at 30 days). Risk was also examined in prespecified subgroups (age ≤65 years vs >65 years; COVID-19 vaccination; SARS-CoV-2 infection status; and use of outpatient COVID-19 antiviral treatment before admission) (eMethods in Supplement 1). Interaction analyses were undertaken to assess statistically significant risk differences between subgroups.
Analyses were performed with SAS Enterprise Guide version 8.2 (SAS Institute Inc). Statistical significance was defined as a 95% CI that did not cross 1.0 on a relative scale. The study was approved with a waiver of informed consent by the VA St Louis Health Care System institutional review board.
Results
There were 8996 hospitalizations (538 deaths [5.98%] within 30 days) for COVID-19 and 2403 hospitalizations (76 deaths [3.16%]) for seasonal influenza (Table). After propensity score weighting, the 2 groups were well balanced (mean age, 73 years; 95% male).
The death rate at 30 days was 5.97% for COVID-19 and 3.75% for influenza, with an excess death rate of 2.23% (95% CI, 1.32%-3.13%) (Figure). Compared with hospitalization for influenza, hospitalization for COVID-19 was associated with a higher risk of death (hazard ratio, 1.61 [95% CI, 1.29-2.02]).
The risk of death decreased with the number of COVID-19 vaccinations (P = .009 for interaction between unvaccinated and vaccinated; P < .001 for interaction between unvaccinated and boosted). No statistically significant interactions were observed across other subgroups (Figure).
https://jamanetwork.com/journals/jama/fullarticle/2803749
doi:10.1001/jama.2023.5348
Yan Xie, PhD[SUP]1[/SUP]; Taeyoung Choi, MPH[SUP]1[/SUP]; Ziyad Al-Aly, MD[SUP]1[/SUP]
In the first year of the COVID-19 pandemic, 2 US studies suggested that people hospitalized for COVID-19 had nearly 5 times the risk of 30-day mortality compared with those hospitalized for seasonal influenza.[SUP]1[/SUP][SUP],2[/SUP] Since then, much has changed, including SARS-CoV-2 itself, clinical care, and population-level immunity; mortality from influenza may have also changed. This study assessed whether COVID-19 remains associated with higher risk of death compared with seasonal influenza in fall-winter 2022-2023.
Methods
We used the electronic health databases of the US Department of Veterans Affairs (VA). Between October 1, 2022, and January 31, 2023, we enrolled all individuals with at least 1 hospital admission record between 2 days before and 10 days after a positive test result for SARS-CoV-2 or influenza and an admission diagnosis for COVID-19 or seasonal influenza. We removed 143 participants hospitalized with both infections. The cohort was followed up until the first occurrence of death, 30 days after hospital admission, or March 2, 2023. Differences in baseline characteristics between the groups were evaluated through absolute standardized differences (<0.1 indicating good balance).
We evaluated the risk of death in people hospitalized for COVID-19 vs influenza through inverse probability-weighted Cox survival models. Logistic regression was used to generate a propensity score, which was then applied in inverse probability weighting to balance the 2 groups; covariates were defined based on prior knowledge and ascertained in the 3 years preceding admission (eMethods in Supplement 1). We also estimated the absolute risk as the percentage of excess deaths (difference in death rates between COVID-19 and influenza groups at 30 days). Risk was also examined in prespecified subgroups (age ≤65 years vs >65 years; COVID-19 vaccination; SARS-CoV-2 infection status; and use of outpatient COVID-19 antiviral treatment before admission) (eMethods in Supplement 1). Interaction analyses were undertaken to assess statistically significant risk differences between subgroups.
Analyses were performed with SAS Enterprise Guide version 8.2 (SAS Institute Inc). Statistical significance was defined as a 95% CI that did not cross 1.0 on a relative scale. The study was approved with a waiver of informed consent by the VA St Louis Health Care System institutional review board.
Results
There were 8996 hospitalizations (538 deaths [5.98%] within 30 days) for COVID-19 and 2403 hospitalizations (76 deaths [3.16%]) for seasonal influenza (Table). After propensity score weighting, the 2 groups were well balanced (mean age, 73 years; 95% male).
The death rate at 30 days was 5.97% for COVID-19 and 3.75% for influenza, with an excess death rate of 2.23% (95% CI, 1.32%-3.13%) (Figure). Compared with hospitalization for influenza, hospitalization for COVID-19 was associated with a higher risk of death (hazard ratio, 1.61 [95% CI, 1.29-2.02]).
The risk of death decreased with the number of COVID-19 vaccinations (P = .009 for interaction between unvaccinated and vaccinated; P < .001 for interaction between unvaccinated and boosted). No statistically significant interactions were observed across other subgroups (Figure).
https://jamanetwork.com/journals/jama/fullarticle/2803749