English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Similars in
SciELO 
Permalink
KEY POINTS
Current knowledge
• The beneficial effects of SARS-CoV-2 vacci nation in preventing the development of severe disease from COVID-19, reducing hospitalizations, intensive care unit admis sions and mortality are known, but few data exist on the evolution in vaccinated patients who develop severe illness from COVID-19.
Contribution of the article to current knowledge
• Our results showed no benefit in time in days from symptom onset and peak disease severity in vaccinated patients who develop severe COVID-19 disease. But the morta lity of vaccinated patients who develop COVID-19 disease was less than those not vaccinated.
Although social distancing and isolation were, initially, effective in limiting the spread of the virus, the absence of immunity left the people more susceptible to infection1. Vac cines have become an effective tool for pre venting SARS-CoV-2 infections and hospital ization2.
In Argentina, the greatest impact of the pandemic was in 2021, with overcrowding of beds in emergency services, general wards, and critical care units. At that time, the vac cination status of hospitalized patients varied from unvaccinated to incompletely or fully vaccinated. The vaccine campaign began on December 29, 2020, and, according to statis tics from an international database of vacci nation against COVID-19 (Our World in Data, dependent on Oxford Martin School, Univer sity of Oxford and Global Change Data Lab), until October 2022, 41 322 415 people have been vaccinated in Argentina with at least one dose, and 37 808 773 people were consid ered as fully vaccinated3.
So far, vaccines are known could be effec tive in preventing asymptomatic and symp tomatic infections, as well as hospitalizations, transfer to intensive care units, severe dis ease, and COVID-19-related death4-6. Second or booster doses of vaccines are associated with increased protection against symptom atic disease3,7,8. Furthermore, it has been prov en that, although the efficacy of COVID-19 vaccines decreased with the emergence of new variants, the prevention of hospitaliza tion and death has remained high compared to unvaccinated patients9,10. These findings were also observed in studies that included healthcare workers11.
Since the start of the pandemic, much has been studied about the impact of vaccination on the course of the COVID-19 disease. How ever, there remains a knowledge gap regard ing the specific effect of vaccination on the time between symptom onset and peak se verity. This data could be important to inform the clinical management of patients with CO VID-19, particularly in terms of hospitaliza tion and treatment decisions.
Our primary objective was to compare days from the onset of symptoms to the peak of disease severity, considering the day of Me chanical Ventilation (MV) or death (in patients who did not receive MV). Our secondary ob jective was to compare mortality in vaccinat ed and unvaccinated patients hospitalized for COVID-19 in the general ward.
Materials and methods
Study design and source of data
The Argentinian Multicenter Registry of CO VID-19 (in Spanish: Registro Multicentrico Argenti no COVID-19 [REMA-COVID-19]) is a multicenter registry of COVID-19 hospitalized patients that includes 36 hospitals from 6 Argentine provinc es12. The registry uses an ad hoc online platform that was created the 1st of March 2020 by an ini tiative of the Argentinean Society of Medicine to investigate the epidemiology of the SARS-CoV-2 pandemic in Argentina. The registry was created respecting the protection of the patient’s identity and data, following the current legal regu lations, and assigning them numerical codes. Each center could log in with its users and enter the information corresponding to each case. The data entry was carried out by physicians of each center who were in charge of the care of those patients hospitalized with COVID-19.
This registry was approved by the Institution al Review Board (Ethics Committee of the Italian Hospital of Buenos Aires #5602) and sent to all participating centers for individual approval by their respective Ethics Committees. The manu script adheres to the Strengthening the Report ing of Observational Studies in Epidemiology (STROBE) guideline12,13.
Setting and participants
Patients over 18 years of age with a confirmed diagnosis of coronavirus infection admitted to different hospitals in Argentina during the peri od from 1st January 2021 to 31st May 2022, were included.
Inclusion criteria were: 1) patients aged 18 years or older, 2) with a diagnosis of COVID-19 confirmed by reverse transcriptase polymerase chain reaction (RT-PCR) or other methods that detect viral RNA for SARS-CoV-2, 3) hospitalized in any general ward of a participating facility in cluding specific respiratory care wards but not primarily involving invasive mechanical venti lator support, and 4) COVID-19 as the primary diagnosis. Patients were excluded if they were admitted directly to Intensive Care Units. Pa tients were followed up until in-hospital death or hospital discharge.
Variables
We divided the original variables into 9 groups: 1) demographic data, body mass index 2) comorbidities; 3) the number of doses of vac cines applied, name of the vaccine, and date of application; 4) symptoms and signs at admission and time since symptom onset at admission; 5) complementary studies; 6) specific treatment for COVID-19 infection; 7) need for admission to Intensive Care Unit, need for MV, vasoactive support, and hemodialysis; 8) hospital discharge or death. For patients requiring admission to critical care, non-transfer was recorded due to decisions on the limitation of therapeutic effort or lack of available beds.
Statistical analysis
A descriptive analysis of all data was per formed. Means with their standard deviations and medians with their interquartile intervals, according to the distribution of the variable, and confidence intervals (CI) were described.
In this study, the primary outcome exam ined the duration between symptom onset and the peak of disease severity (measured as me chanical ventilation (MV) utilization or death for patients who did not receive MV), using a Cox regression model. The analysis was restricted to patients who experienced the peak severity of COVID-19 disease. All patients included in this analysis had the event of interest, and no censoring or competitive risk events were ac counted for. To enhance the comparability of the treatment groups, we reported both crude and adjusted coefficients using Propensity Score Matching (PSM). As a subgroup analysis we com pare patients with one and two doses.
In terms of our secondary outcome, which in volved evaluating mortality rates between vacci nated and unvaccinated patients, we calculated the Risk Ratio (RR). The RR was calculated first in the entire study population, and secondly in the subset of the PSM population. Each RR and the 95% confidence interval were calculated with the “cs” command in STATA v.16. For statistical significance, a p-value below 0.05 was adopted, and the estimated effects, accompanied by their corresponding 95% confidence intervals, were systematically reported.
To assess the robustness of the vaccine’s ef fect, we used PSM to reduce the confounding effects. The potential confounders in the caus al relationship between the COVID-19 vaccine and the severity of COVID-19 disease were: age, male sex, hypertension, Chronic Obstructive Pulmonary Disease (COPD), immunosuppres sion, cancer, diabetes, heart failure, chronic kid ney disease (CKD)14 and systemic corticosteroid therapy15. For the PS development, first, we es timated the individual PS for COVID-19 vaccine receipt with a multivariable logistic regression model that included all the potential confound ers. Matching was performed with the use of a 1:1 matching protocol without replacement (greedy-matching algorithm), with a caliper width equal to 0.2 of the standard deviation of the logit of the PS16. Standardized differences were estimated for all the baseline covariates before and after matching to assess pre-match imbalance and post-match balance. We consider standardized differences of less than 10.0% for a given covariate to indicate a relatively small im balance16,17. The data were analyzed with STATA® v16 software.
Study size
We included a consecutive sample of patients who met the inclusion criteria. Because this was a “live registry” of patients with COVID-19, we had a fixed sample size, thus we performed a power calculation for the primary and second ary outcomes.
For the primary outcome, we only evaluated patients who receive MV or died during hos pitalization. In the registry, 788 patients died or received MV, of which 181 were vaccinated. With this sample size, we have a power of 99%, with an alpha of 5%, to detect a minimal signifi cant difference (between days from the onset of symptoms to peak disease severity) of 2 days between groups (we expected 7 days in unvacci nated patients and 9 in vaccinated), with an SD of 2 in both groups.
Results
Participants
The Argentinian Multicenter Registry of CO VID-19 included 5895 patients, of which 3663 were included in this study (Fig. 1). We excluded patients younger than 18 years old and hospi talized before January 1st, 2021. Of them, 3001 (81.9%) were unvaccinated patients, and 662 (18%) were vaccinated. Among the vaccinated, 421 (63.6%) received one dose and 241 (36.4%) two doses.
Descriptive data
The median age was higher in the group of vaccinated patients: 66.8 (SD 18.9) vs. 51.8 (SD 21.8); and the proportion of women was similar in both groups (45.3% in the unvaccinated and 48.3% in the vaccinated group). As for comorbid ities, vaccinated patients had a higher burden of comorbidities, such as hypertension, diabetes, or cancer (Table 1).
Table 1. Baseline characteristics of patients hospitalized for COVID-19
Among vaccinated patients, Sputnik V was the most frequent (n = 304, 46%). The number of doses and type of vaccine received are shown in Table 2.
Outcome data and main results
Regarding our primary outcome, we observed that unvaccinated patients took a median of 7 days (IQR 4-11) from symptom onset to reach the peak of severity, defined as either requiring mechanical ventilation or experiencing death (in patients with do-not-intubate orders). Vac cinated patients also took a median of 7 days (IQR 4-12) to reach the same peak severity. In the crude Cox regression analysis, we observed a hazard ratio (HR) of 0.93 (95% CI 0.79-1.1; p = 0.43), with no significant difference between vaccinated and unvaccinated patients. After Pro pensity Score Matching (PSM), the HR was 1.08 (95% CI 0.82-1.4; p = 0.56), further supporting the absence of significant differences between the two groups. These findings are presented graph ically in Fig. 2, which displays the Kaplan-Meier curves, and Fig. 3, which illustrates the covari ate imbalance between vaccinated and unvac cinated patients before and after matching. In tables S1 and S2 we presented the descriptive data and outcomes of the PSM population. In the subgroup analysis patients with one dose and two doses were compared against unvacci nated patients. Patients with one vaccine dose had a HR of 1.04 (95% CI 0.85-1.28; p = 0.65) and with two doses 0.84 (95% CI 0.65-1.01; p = 0.21). In the PMS population, the HR of patients with one dose was 1.19 (95% CI 0.86-1.65; p = 0.27) and with two doses 1.20 (95% CI 0.80 - 1.8; p = 0.36).
Figure 2 Kaplan Maier time to peak of disease severity (death or MV) in ventilated patients against non-ventilated patients
Figure 3 Love plot for absolute standardized differences before and after propensity score matching comparing covariate values between vaccinated and unvaccinated COVID-19 patients
Regarding our secondary outcome, we ob served that mortality was higher in vaccinated patients than unvaccinated in the univariate analysis (23.3% vs. 17.7%, p < 0.001 (Table 3).
Therefore, the RR of death in vaccinated pa tients compared to unvaccinated patients was 1.31 (CI 95% 1.12-1.54; p < 0.001). However, in the PSM population, we observed that the re lationship between vaccination and mortality was reversed, with the vaccine being protective against mortality with an RR of 0.7 (IC95% 0.56- 0.84; p < 0.001). The RR in patients with one CO VID-19 vaccine dose in the PSM population was 0.7 (IC95% 0.45-1.03; p = 0.063), and with two doses was 0.6 (IC95% 0.46-0.79; p < 0.001).
Discussion
Our study has some limitations. First, we could not differentiate between SARS-CoV-2 variants. Previous studies have confirmed that the active immunity conferred by the vaccines decreased with the emergence of new variants, such as Delta, but even so, vaccinated patients had a lower risk of death compared to unvacci nated patients18. Secondly, we could not include booster doses of the vaccines. Current evidence has shown that patients who receive a booster dose, may have up to 90% lower risk of death compared to those who do not receive a booster dose8,18, even in the presence of emerging vari ants9. Third, we did not include High Flow Nasal Cannula (HFNC) therapy as a study endpoint be cause not all participating centers utilized this treatment strategy, and at the time of the study, there was insufficient evidence to support the benefits of using HFNC in patients with COVID.
In this study, no significant differences were observed in terms of the time elapsed between the symptom onset and peak of disease severity (MV or death) when comparing the vaccinated and unvaccinated groups. As expected, and in agreement with previous literature, we observed that the COVID-19 vaccine reduces mortality, af ter adjustment by confounders.
The peak of disease severity usually coincides with the cytokine storm. This is an excessive inflammatory response with the release of cy tokines and pro-inflammatory chemical media tors. Clinically, it can lead to respiratory distress syndrome requiring mechanical ventilator sup port, multi-organ failure, or even death17. Previ ous studies reported that the time between the onset of symptoms and the cytokines storm was 7 to 10 days19,20. In this sense, Khalili et al. pub lished a meta-analysis that describes the time between the onset of symptoms to ICU admis sion, that was 9.8 (95% CI: 8.8, 10.9) days and 15.9 (95% CI: 13.1, 18.8) to death20. In our study, we ob served also that the time between onset to the peak of severity was 9 days, and the difference between time to death is because the studies included in the revision of Khalili et al. evalu ated time to death in ventilated patients, while we evaluate time to mechanical ventilation or death in patients with do-not-intubate orders.
Regarding our secondary objective, we ob served that mortality was slightly higher in the vaccinated group, but these patients had a high er burden of comorbidities. In this sense, when a multivariate analysis was performed adjusting for age, sex, and comorbidities, we observed that vaccination is an independent protective factor against mortality (Table S2). In this way, Ten forde et al2, reported that vaccinated COVID-19 hospitalized patients had an adjusted OR of 0.33 (CI 95% 0.19-0.5) for the outcome progression to severe disease outcomes (including respiratory failure and death). Other studies that evaluated the incidence of mortality showed a decrease in the number of cases per 100 000 persons/day in favor of patients vaccinated with mRNA vac cines4 as well as with inactivated virus or viral vector vaccines5. Finally, in the study of Busic et al21 results, the initial analysis suggests height ened mortality rates among vaccinated patients. Yet, upon meticulous confounder adjustment, COVID-19 vaccines exhibit a remarkable poten tial to curtail mortality in the context of hospi talized patients.
Several published articles have examined the impact of SARS-CoV-2 vaccination on mortality in Argentina. Macchia et al5 conducted a study evaluating mortality in a group of patients over 60 years old who received either one or two dos es of 4 different vaccines. The results showed favorable outcomes for vaccination. Other study (Gonzalez et al22) assessed the benefits of a single dose of the Sputnik V vaccine on mortality in a cohort of patients aged 60-79 from the province of Buenos Aires; and a third one (Gonzalez et al23) examined a cohort of children and adoles cents aged 3-17 years, where vaccination dem onstrated benefits in reducing mortality and hospitalization. It is important to note that all the mentioned studies were conducted on co horts of patients without previous SARS-CoV-2 infection. In contrast, our current study focuses on a cohort of hospitalized patients with severe symptoms of COVID-19. Another local study24 has shown an absolute risk reduction of 15.1% in mortality in hospitalized patients with a com plete vaccination schedule vs. those not vacci nated or with an incomplete schedule.
In conclusion, in the present study it was ob served that the COVID-19 vaccine did not change the time between symptoms onset and peak of disease severity in COVID-19 patients. It also was found that hospitalized vaccinated patients due to COVID-19 have lower mortality compared to unvaccinated patients.
