Vaccination reduces need for emergency care in breakthrough COVID-19 infections: A multicenter cohort study

Abstract

Background: While recent literature has shown the efficacy of the SARS-CoV-2 vaccine in preventing infection, it's impact on need for emergency care/hospitalization in breakthrough infections remain unclear, particularly in regions with a high rate of variant viral strains. We aimed to determine if vaccination reduces hospital visits in breakthrough COVID-19.

Methods: This observational cohort analysis compared unvaccinated (UV), partially vaccinated (PV), and fully vaccinated (FV) adult patients with SARS-CoV-2 infection requiring emergency care(EC)/hospitalization within an eight-hospital system in Michigan. Demographic and clinical variables were obtained from the electronic record. Vaccination data was obtained from the Michigan Care Improvement Registry and Centers for Disease Control vaccine tracker. Primary endpoint was rate of emergency care/hospitalization encounters among patients diagnosed with COVID-19. Secondary outcome was severe disease-composite outcome (ICU, mechanical ventilation, or in-hospital death).

Findings: Between December 15,2020 and April 30,2021, 11,834 EC encounters were included:10,880 (91.9%) UV, 825 (7%) PV, 129 (1.1%) FV. Average age was 53.0 ± 18.2 and 52.8% were female. Accounting for the SARS-CoV-2 vaccination population groups in Michigan, the ED encounters/hospitalizations rate relevant to COVID-19 was 96% lower in FV versus UV (multiplicative effect:0.04, 95% CI 0.03 to 0.06, p < 0.001) in negative binomial regression. COVID-19 EC visits rate peaked at 22.61, 12.88, and 1.29 visits per 100000 for the UV, PV, and FV groups, respectively. In the propensity-score matching weights analysis, FV had a lower risk of composite disease compared to UV but statistically insignificant (HR 0.84, 95% CI 0.52 to 1.38).

Interpretation: The need for emergency care/hospitalization due to breakthrough COVID-19 is an exceedingly rare event in fully vaccinated patients. As vaccination has increased regionally, EC visits amongst fully vaccinated individuals have remained low and occur much less frequently than unvaccinated individuals. If hospital-based treatment is required, elderly patients with significant comorbidities are at high-risk for severe outcomes regardless of vaccination status.

Keywords: COVID-19; SARS-CoV-2; coronavirus; intensive care unit; mechanical ventilation; mortality; vaccination; variants.

Figure 1

Screening and Categorization of all hospital-based COVID-19 cases into UV, PV, and FV groups Patients with SARS-CoV-2 infection with an ED encounter were eligible participants. Patients with secondary diagnosis of COVID-19 and age less than 18 years were excluded. Patients with primary COVID-19 diagnosis without reference to confirmed testing in the emergency provider note were also excluded. Included patients were then categorized into UV (unvaccinated), PV (partially vaccinated), and FV (fully vaccinated) groups. UV individuals had no record of immunization against SARS-CoV-2 or first-dose vaccination after symptom onset. PV individuals had symptom onset after a single dose of either mRNA (Pfizer, Moderna) vaccine, or < 14 days after the second dose of either mRNA vaccine (Pfizer, Moderna) or < 14 days after the administration of the single dose of viral vector vaccine (Johnson & Johnson). FV individuals had symptom onset >14 days since administered of second dose of either mRNA vaccine, or >14 days since administration of viral vector vaccine (Johnson & Johnson).

Figure 2

ED encounters of COVID-19 patients among vaccination groups Results shown are for the entire study cohort of adult COVID-19 patients presenting from December 15, 2020 thru April 30, 2021. Case rate of emergency encounters proportionated to the State SARS-CoV-2 vaccination population groups. Weekly crude and estimated trend of COVID-19 infection ED encounters (visits) for each vaccinated group are depicted as number of cases per 100000 over study period. The line graph illustrates the estimated trend of infection ED visits for each vaccinated group. When the state FV (fully vaccinated) population size was only 19 individuals between 12/27/2020 and 1/2/2021, one ED visit occurred in FV group which was not included in analysis due to the bias of an extreme outlier in trend analysis. Through joinpoint regression analysis for each vaccination group, the grid search of joinpoints by Monte Carlo permutation tests identify that statistically significant change of slope estimates occurred before and after specific time points in trend.