ABSTRACT
Social and political polarization, which sometimes is the result of misinformation, is a common obstacle that can be harmful at the moment of communicating government policies. Intelligent tools that aid critical thinking in the light of different opinions and standpoints available in social media can help ameliorate this obstacle. This paper presents preliminary research work toward developing such tools by proposing a methodology for building stance trees based on tweets collected from social media. Stance trees are hierarchical structures where nodes represent arguments pro, anti, or uncertain about a target issue and edges stand for attack relations between those arguments. The proposed methodology includes retrieving tweets relevant to the target issue, manually labeling a sample set of the collected tweets, developing and applying a model for stance detection, and finally building a stance tree. We illustrate the expected results through a case study on the politically polarized "COVID-19 vaccine"issue. Our preliminary results demonstrate the feasibility of the proposal and highlight the utility of stance trees as a tool for aiding critical thinking. © 2022 ACM.
ABSTRACT
Background: In the current situation of the SARS-CoV-2 pandemic, the Spanish Society of Rheumatology recommends vaccination of patients with chronic infammatory diseases (CID) under treatment with biological DMARDs (bDMARDs). However, the data regarding the generation of protective antibody titers after mRNA vaccines in patients with CID is limited. Objectives: To determine the seroconversion rate and safety after the SARS-CoV-2 vaccine in patients with CID under treatment with bDMARDs Methods: Cross-sectional observational study of 81 patients with CID from the HURS in Córdoba, who have received full vaccination for SARS-CoV-2 (without having previously suffered from COVID-19 disease) according to national guidelines. A determination of specifc IgG-type antibodies against the trimeric spike protein of SARS-CoV2 was performed on all of them. The chemilumines-cence technique with the kit was used in serum samples taken 4-5 weeks after administration of the second dose of the vaccine. Information about sociode-mographic characteristics, disease, type of bDMARDs, concomitant treatments and adverse effects after the second dose of the vaccine were collected in each patient. Results: 81 patients were included (mean age 59.5, 72.8% females). 50.6% of patients had RA, 17.3% SpA, 11% PsoA and 18.5% other CID. 23.5% were under treatment with Rituximab, 38.8% antiTNF, 13.6% Tocilizumab, 9.9% abatacept, 5% anti-JAK and 14.2% under other treatments. Anti-SARS-CoV-2 antibodies and neutralizing activity were detected in 80% of study participants. Rituximab treatment was signifcantly associated with negative seroconver-sion in comparison with patients under antiTNF treatment (OR 84.0 (95%CI 12.9-1709.2)). No interaction was found between the bDMARDs treatment and the type of vaccine with regard to the seroconversion, nor between bDMARDs and concomitant synthetic DMARD. When we evaluated IgG titers against the spike protein of SARS-CoV2, we found that patients under treatment with Rituximab showed the lowest titers levels in comparison with patients with other treatments (Figure 1, Table 1). In addition, patients who received AstraZeneca vaccine developed lower titers of antibodies in comparison with patients who received Pfzer (Table 1). Interestingly, among patients with antiTNF treatment, AstraZeneca was associated with lower IgG titers in comparison with Pfzer and Moderna [405.9 (553.0) vs. 1084.1 (791.2) vs. 1264.0 (1012.6), p=0.016, respectively]. No differences between vaccines were found in patients with the other type of bDMARDs. Only 18.9% presented mild adverse effects. No serious adverse effects were observed and no patient experienced a disease fare after vaccination. Conclusion: Our results show that SARS-CoV-2 mRNA vaccines produce sero-conversion in most patients with CID, except in the case of patients with rituxi-mab. No severe adverse effects or CID reactivation were found. Despite the small number of patients included, this study suggests the need for revaccination in the group of patients treated with rituximab or vaccinated with Astrazeneca.