Prevalent comorbidities among young and underprivileged: Death portrait of COVID-19 among 235 555 hospitalized patients in Brazil

BackgroundCOVID-19 has been alarmingly spreading worldwide, with Brazil ranking third in total number of cases and second in deaths. Being a continental country, which comprises many ethnic groups and an engrained social inequality, the pandemic evidenced this heterogeneous discrepancy. We aimed to estimate the impact of associated risk factors, isolated or combined, on COVID-19 severeness, detecting specific epidemiological profiles for multiple age ranges in hospitalized Brazilians. MethodsIn this large retrospective cohort study, we used open-access data from the Ministry of Health of Brazil with COVID-19 confirmed hospitalized patients annotated in SRAG system between February and August 2020, a total of 235555 entries. The association of COVID-19 death with socio-demographic and clinical characteristics was analysed and presented as odds ratios adjusted by confounding co-variables. We also presented marginal mean aOR values for high-order interactions either by or not another fixed level or condition. We kept all other variables in the multivariate logistic models in their mean values or equal proportions. FindingsYounger individuals with one or more comorbidities had an adjusted odds ratio up to four-fold compared to those without it, in the same age interval. Younger diabetic patients either self-declared as brown ethnicity (aOR 5{middle dot}58, 95% CI 4{middle dot}97-6{middle dot}25; p<0{middle dot}0001) or with some other associated comorbidities, mainly chronic hematologic disease (21{middle dot}09, 13{middle dot}64-32{middle dot}6; p<0{middle dot}0001) and obesity (aOR 21{middle dot}7, 95% CI not calculated; p<0{middle dot}0001), resulted in outstanding death risk. Age over 60, particularly over 90 (28{middle dot}91, 24{middle dot}5-34{middle dot}11; p<0.001), usage of invasive ventilatory support (16{middle dot}23, 14{middle dot}05-18{middle dot}75; p<0{middle dot}001), admission to intensive care units (3{middle dot}14, 2{middle dot}82-3{middle dot}48; p<0{middle dot}001), multiple respiratory symptoms (3{middle dot}24, 2{middle dot}79-3{middle dot}75; p<0{middle dot}0001), black ethnicity (1{middle dot}78, 1{middle dot}52-2{middle dot}07; p<0{middle dot}05), and diagnosis previous to hospitalization (1{middle dot}32, 1{middle dot}19-1{middle dot}47; p<0{middle dot}05) were associated with higher death odds. As protective factors, with roughly one third less death risk, we found hospitalization duration of (4, 7] days and illness onset to hospitalization over 6 days. InterpretationWe found evidence for increased COVID-19 risk in two distinct groups: younger patients with prevalent comorbidities, especially in brown ethnicity, and patients with black ethnicity. We speculate that the pro-inflammatory synergism of COVID-19 and comorbidities, promoting an overproduction of cytokines, is partially the cause of higher mortality in this young group. Brazilian black and brown are underprivileged populations, with structural social inequality, limited healthcare access and, thus, remarkable disease vulnerability. Our study supplies essential data to patient stratification upon admission, optimizing hospital management, and to guide public policy determinations, including group prioritization for COVID-19 vaccination in Brazil. FundingNone. O_TEXTBOXResearch in context Evidence before this studyCOVID-19 is still very active, having spread to over two hundred countries and caused more than one million deaths worldwide. Its current situation requires large-scale studies to assess the impact of preexisting comorbidities, symptoms, and socioeconomic issues regarding mortality rate, especially where lack of control is evident. We searched PubMed, Google Scholar, medRxiv, and bioRxiv on Dec 12, 2020, for studies published in English or Brazilian Portuguese, estimating the impact of several risk factors in COVID-19 prognosis. We used the search terms "Brazil" or "risk factors" or "ethnicity" or "cohort" or "diabetes mellitus" or "mortality" or "symptoms" or "comorbidities", and related synonyms, combined with "SARS-CoV-2" or "COVID-19". Many pre-existing conditions have shown to directly impact patient prognosis, out of which cancer, chronic kidney disease, chronic obstructive pulmonary disease, cardiovascular disease, obesity, and diabetes, among others, are well established in SARS-CoV-2 infection severeness. Some studies reported an increased death risk for non-white Brazilians, but no large scale cohort analyzing the impact of one or more associated risk factors in younger Brazilians patients were found. Added value of this studyWe found that the impact of having one or two or more risk factors on mortality are progressively higher in ages (60, 80], (40, 60], (20, 40], and (0, 20], compared with people of the same age interval without comorbidities. We also found that young brown individuals with diabetes, as well as black ethnicity on its own, are population subgroups at remarkably higher risk for severe COVID-19 in Brazil. Furthermore, advanced age, usage of ventilatory support, admission to intensive care units, multiple respiratory symptoms, and diagnosis previous to hospitalization were associated with higher death odds. As protective factors, we found hospitalization duration of (4, 7] days and illness onset to hospitalization over 6 days. Implications of all the available evidenceWe identified multiple epidemiological profiles associated with death risk in different age ranges in Brazilian COVID-19 hospitalized patients. These findings unveil that a large part of Brazilian working-age population is at a higher risk for SARS-CoV-2 death, a neglected situation that is further exacerbating inequalities, leading to a striking sociodemographic and economic impact. We hope that our analysis aids patient risk stratification, hospital management optimization, and public policy determination, including prioritization for COVID-19 vaccination in Brazil. C_TEXTBOX