Predictors of hospitalization for COVID-19 in patients with autoimmune rheumatic diseases: results from a community cohort follow-up.

The identification of risk factors for COVID-19 adverse course in autoimmune rheumatic diseases (ARDs) is of the utmost importance when approaching patient management; however, data are scarce in relation to the Latin American population. The objective of this study was to determine predictors of hospitalization for COVID-19 patients from an ARD community cohort. A real setting longitudinal study (March to November 2020) in an ARD community cohort was carried out. Potential predictors of hospitalization for COVID-19 examined included (1) sociodemographic variables (age, gender, education, tobacco use, socioeconomic status, and co-inhabitants), (2) comorbidities, (3) time to COVID-19 diagnosis, and (4) ARD’s features: clinical (disease type, disease duration, activity), treatment [corticosteroids use/doses, use of synthetic DMARDs (cDMARDs, tsDMARDs, and bDMARDs)], treatment schedule and non-adherence, and the Multidimensional Health Assessment Questionnaire (MDHAQ). Univariable and multivariable regression analysis were conducted; OR and 95% CI (p < 0.05) were determined. One thousand and one hundred forty-eight patients with ARDs were included; 154 had COVID-19; of these 139 (90.3%) were women, aged 52.5 (13.7) years; 33.1% had hypertension and 61.0% an affected organ by ARD. Infection was detected 8.4 (10.1) days after symptoms started; there were 33 hospitalized patients (rate 21.4%). Predictors of hospitalization by multivariable analysis were age (OR: 1.06; CI: 1.01–1.10; p: 0.01), comorbidities: hypertension (OR: 3.95; 95% CI: 1.40–10.95, p: 0.01) and neoplasm (OR: 9.0; 95% CI: 1.6–52.3; p: 0.01), number of organs involved by ARD (OR: 2.26; 95% CI: 1.16–4.41; p: 0.02), and infection diagnosis delay (OR: 1.36; 95% CI: 1.03–1.80; p: 0.01). In our ARD patients with COVID-19, older age, comorbidities (neoplasm and hypertension), and a delay in COVID-19 diagnosis were predictors of hospitalization. The only ARD-associated predictor feature was the number of organs involved.