Demographic Characteristics
The demographic characteristics of vaccinated and non-vaccinated patients are displayed in Table 1. Significant differences were observed between both groups for age (95% CI [1.048; 1.102]), virus variant (95% CI [2.09; 33.00]), pre-existing diseases (95% CI [2.87; 27.14]), immunodeficiency (95% CI [1.96; 12.97]), type 2 diabetes (95% CI [1.29; 6.43]), hypertension (95% CI [2.12; 7.96]), oncological disease (95% CI [2.59; 18.68]), cardiac disease (95% CI [2.50; 10.26]), vascular disease (95% CI [1.48; 9.27]), pulmonary disease (95% CI [1.56; 6.85]), neurological disease (95% CI [1.58; 12.18]), and organ transplant (95% CI [2.85; 70.90]): vaccinated patients were older and were suffering from pre-existing diseases more frequently. There were six cases of pregnancy in the non-vaccinated group and none in the vaccinated group.
Virus variant, pre-existing disease, immunodeficiency, oncological disease, vascular disease, neurological disease, organ transplant, and pregnancy were rare characteristics in our dataset and therefore not suitable as confounding variables in a large multiple logistic regression model. Consequently, we used age, type 2 diabetes, hypertension, cardiac disease, and pulmonary disease as confounders in a comprehensive model and conducted a sub-analysis for virus variant, pre-existing disease, immunodeficiency, oncological disease, vascular disease, neurological disease, organ transplant, and pregnancy.
Vaccination and ARDS
Without adjusting for confounding variables, vaccinated patients developed ARDS significantly less frequently than non-vaccinated patients (RR: 0.40, 95% CI [0.21; 0.62]). The NNT was three patients. When adjusting for confounding variables such as age, type 2 diabetes, hypertension, cardiac disease, and pulmonary disease, the effect of vaccination remained significant (RR: 0.64, 95% CI [0.29; 0.94]) (Table 2, Figure 2).
The subgroup analyses excluded patients with the omicron variant, healthy subjects, immunodeficiency, oncological disease, vascular disease, neurological disease, organ transplant, and pregnancy. The results remained significant in all subgroup analyses (Supplementary Table 1, Supplementary Figure 1).
Vaccination and ARDS severity
Without adjusting for confounding variables, COVID-19 vaccination showed increasing protective effects with increasing severity of ARDS (RR: 0.63, 95% CI [0.41; 0.82]). After adjusting for confounding variables such as age, type 2 diabetes, hypertension, cardiac disease, and pulmonary disease, the effect remained significant (RR: 0.61, 95% CI [0.37; 0.92]) (Table 3, Figure 3).
Younger patients developed ARDS significantly more often than older patients (OR: 0.956, 95% CI [0.934; 0.976]). This difference persisted when controlling for vaccination status (OR: 0.966, 95% CI [0.943; 0.089]) (Table 4).
Therefore, we evaluated the effect of vaccination in younger and older patients separately. Younger patients (< 60 years) developed ARDS less frequently when they were vaccinated (RR: 0.51, 95% CI [0.20; 0.90]), whereas older patients (> 60 years) did not show a significant effect of the vaccination on the onset of ARDS (Table 5, Figure 4). The NNT was three in younger patients, whereas it was non-significant for older patients.
ARDS and predictive risk factors
In order to evaluate the risk factors for developing ARDS in COVID-19 patients, we performed a logistic regression between ARDS and the patients' characteristic variables, controlling for vaccination status (Table 4). After adjusting for vaccination status, the following characteristics remained significant for developing ARDS: older patients had lower odds (OR: 0.966, 95% CI [0.943; 0.989]), patients with the omicron variant had lower odds (OR: 0.16, 95% CI [0.00; 0.95]), patients with oncological disease had lower odds (OR: 0.26, 95% CI [0.05; 0.84]), and patients with vascular disease had lower odds (OR: 0.15, 95% CI [0.02; 0.57]). All six pregnant women developed ARDS, making it impossible to compute ORs with CIs, however, this suggests a strong relationship. Therefore, younger patients, patients with the delta variant, patients without oncological disease, patients without vascular disease, and pregnant patients were more prone to developing ARDS. Conversely, older patients, patients with the omicron variant, patients with oncological disease, patients with vascular disease, and non-pregnant patients were less prone to developing ARDS.
Other characteristics had a significant relationship with ARDS in univariate analysis, but the relationship disappeared after adjusting for vaccination status: obese patients had higher odds (OR: 2.35, 95% CI [1.03; 5.78]), patients with immunodeficiency had lower odds (OR: 0.35, 95% CI [0.11; 0.91]), patients with hypertension had lower odds (OR: 0.48, 95% CI [0.24; 0.93]), patients with cardiac disease had lower odds (OR: 0.32, 95% CI [0.14; 0.67]), and patients with renal disease had lower odds (OR: 0.37, 95% CI [0.14; 0.83]). There was no association between ARDS and type 2 diabetes mellitus (OR: 1.02, 95% CI [0.44; 2.23]).
This standard approach of assessing each variable separately has some limitations because these variables have unknown amounts of correlation among each other. Therefore, we performed a regularized logistic regression using LASSO regression incorporating all variables, the model confirmed our prior results that younger patients, patients without immunodeficiency, patients without oncological disease, patients without cardiac disease, patients without vascular disease, patients who were non-vaccinated, and pregnant women were associated with a higher frequency of ARDS.
Furthermore, we investigated clinical features associated with COVID-19 ARDS in our dataset and adjusted the results for vaccination (Table 6): COVID-19 ARDS was associated with increased dyspnea (OR: 16.8, 95% CI [3.3; 307]), amount of oxygen therapy (OR: 1.20, 95% CI [1.08; 1.34]), NIV (OR: 110, 95% CI [34; 473]), high flow oxygen therapy (OR: 36, 95% CI [13; 118]), intubation and invasive ventilation (OR: 71, 95% CI [25; 246]), ECMO (all 25 ECMO patients were in the ARDS group, therefore computing OR and CI was not possible, but the association is likely very strong), tracheotomy (OR: 18, 95% CI [5; 84]), ICU therapy (OR: 139, 95% CI [37; 918]), ICU duration (OR: 1.13, 95% CI [1.08; 1.20]), higher pO2 (OR: 1.025, 95% CI [1.012; 1.041]), higher pCO2 (OR: 1.06, 95% CI [1.03; 1.10]), higher mortality (OR: 12.81, 95% CI [5.18; 35.80]), more sepsis (OR: 33, 95% CI [9.1; 219]), higher CRP level (OR: 1.005, 95% CI [1.001; 1.01]), more pulmonary superinfections (OR: 12.67, 95% CI [5.64; 31.06]), higher D-Dimer levels (OR: 1.09, 95% CI [1.03; 1.17]), more coagulopathy (OR: 18.7, 95% CI [3.36; 353]), and more renal failure (OR: 2.76, 95% CI [1.38; 5.61]). ARDS was associated with higher PCT (OR: 1.31, 95% CI [1.05; 1.73]), but this effect was not significant when adjusting for vaccination status.