Your browser doesn't support javascript.
Show: 20 | 50 | 100
Results 1 - 3 de 3
Add filters

Document Type
Year range
Clinical Cancer Research ; 27(6 SUPPL 1), 2021.
Article in English | EMBASE | ID: covidwho-1816940


Background: The COVID-19 pandemic had enormous consequences in Brazil and worldwide. Patients with cancer affected by COVID-19 are at a higher risk of developing complications and worse outcomes compared to a non-cancer population, particularly the ones on active systemic treatment. Considering the COVID-19 high transmissibility in asymptomatic and presymptomatic patients, we sought to determine the prevalence of COVID-19 infection in patients with solid cancers receiving systemic therapy in a Brazilian public health hospital. Furthermore, we interrogated if socioeconomic status (SES) was associated with prevalence. Methods: Consecutive asymptomatic patients undergoing treatment for solid tumors at the chemotherapy and infusion center of Hospital de Base were enrolled. Patients were prospectively tested for SARS-CoV2 RNA real-time polymerase chain reaction with nasal and oropharyngeal swabs immediately prior to treatment. A socioeconomic survey was performed prior to testing. Demographic and socioeconomic characteristics were summarized in means, medians, and proportions. Results: From October 6 to 13, 2020, 148 asymptomatic patients were identified. Of those, 41 were excluded (16 had hematological malignancies, 15 declined testing, 10 were not on active systemic treatment) leaving 107 eligible patients. The mean age of the population was 58 years-old (SD± 12.6);55% were female and 90% were self-identified as White. The most common cancer sites were gastrointestinal tract (37%) and breast (25%). Most patients had metastatic disease (62.9%) and were on a anticancer treatment involving chemotherapy (62.9%). Regarding to SES, 70% of our population had either primary school or were illiterate as their highest educational level. In terms of monthly income, 88% had a personal income inferior to U$390 and 92% a household income inferior to U$585. Of 107 patients tested, only one (0.9%) was positive for COVID-19. This is a 48 years-old man living in an urban area, with primary school educational level and a monthly personal income inferior to U$390. Conclusion: Despite a high prevalence of COVID19 in Brazil, our cohort demonstrated a low prevalence of COVID19 (0.9%) amongst asymptomatic patients with cancer. We hypothesize that patients with cancer, independently of their SES, are aware of the increased risk of developing severe disease and are adherent to physical distancing, masking, and hygiene measures. LF and BB are co-senior authors.

Clinical Cancer Research ; 27(6 SUPPL 1), 2021.
Article in English | EMBASE | ID: covidwho-1816902


Background: Patients with cancer, both active and previously treated, are at a higher risk of developing severe outcomes from COVID-19. During the pandemic, health care systems (HCS) have adapted the delivery of care, and disparities between private and public systems became even more striking. In Brazil, where 70% of the population depends on the public system, ICU demands largely exceed the capacity in most public centers, whereas in private centers the situation is less challenging. Herein we compare outcomes of patients with cancer and COVID-19 treated in the public and private HCS in Brazil. Methods: We used data from adult patients with solid malignancies who tested positive for COVID-19 and were admitted to two tertiary centers in the state of São Paulo. Patients who tested positive for SARS-CoV2 RNA real-time polymerase chain reaction (RT-PCR) were included. We collected data on baseline clinical conditions, cancer and treatment. Patients were classified by HCS: public system (public) versus (vs) private insurance coverage (private). The co-primary endpoints were all-cause mortality and a composite endpoint consisting of intensive-care-unit (ICU) admission, mechanical ventilation or death (ICU-MV-D). Chi-square, Fisher's exact test and Mann-Whitney U test were used when appropriate. We assessed the association between outcomes and HCS using logistic regression analyses, adjusting for age, sex, current anticancer treatment and comorbidities. Results: From March 16 to October 20 2020, 124 patients were identified. Of those, 90 (72%) were from the public and 34 (28%) from the private HCS. There were no statistical differences in sex, smoking, primary tumor siteand staging between patients from both HCS. Conversely, patients treated in the private system were older [66 (SD 13.8) vs 74 (SD 15.1), p=0.004], had more comorbidities (64.7% vs 37.8% p=0.009), and were on anticancer treatment more frequently (64.7% vs 34.4% p=0.004) compared to public patients. There were no differences in all-cause mortality (public 40% vs private 44.1% p=0.69) between patients treated at the different HCS. Nevertheless, in the composite outcome, private system was significantly associated with increased risk of ICU-MV-D compared to the public system (79.4% and 57.8% p=0.030, respectively). The median time from COVID-19 diagnosis to ICU-MV-D was 13 vs 8 days (p=0.031) and to death was 25 vs 24 days (p=0.24), respectively for public and private HCS patients. In the multivariable logistic regression, HCS was not associated with death [adjusted odds ratio (aOR)=1.16 p=0.75] or ICU-MV-D (aOR=0.55, p=0.27). Conclusion: While patients in the private system were older and had more comorbidities, there were no differences in inpatients all-cause mortality between private and public systems. However, private patients were associated with increased ICU-MV-D. We hypothesize that these findings may reflect disparities in ICU availability, known to be higher in the private system. Further studies investigating this hypothesis are warranted. EDR and DVA co-senior authors.

AMIA ... Annual Symposium Proceedings/AMIA Symposium ; 2021:526-535, 2021.
Article in English | MEDLINE | ID: covidwho-1749439


We develop various AI models to predict hospitalization on a large (over 110k) cohort of COVID-19 positive-tested US patients, sourced from March 2020 to February 2021. Models range from Random Forest to Neural Network (NN) and Time Convolutional NN, where combination of the data modalities (tabular and time dependent) are performed at different stages (early vs. model fusion). Despite high data unbalance, the models reach average precision 0.96-0.98 (0.75-0.85), recall 0.96-0.98 (0.74-0.85), and F1-score 0.97-0.98 (0.79-0.83) on the non-hospitalized (or hospitalized) class. Performances do not significantly drop even when selected lists of features are removed to study model adaptability to different scenarios. However, a systematic study of the SHAP feature importance values for the developed models in the different scenarios shows a large variability across models and use cases. This calls for even more complete studies on several explainability methods before their adoption in high-stakes scenarios.