Women with cancer and COVID-19: an analysis of lethality and clinical aspects in Pernambuco / Mulheres com câncer e COVID-19: uma análise da letalidade e aspectos clínicos em Pernambuco

Abstract Objectives: to analyze the lethality and clinical characteristics in Pernambuco women with neoplasia that were infected by SARS-CoV-2. Methods: a cross-sectional, retrospective study with female patients with neoplasm sin the state of Pernambuco registered and made available by the Secretariat of Planning and Management of the State of Pernambuco (SEPLAG PE). Secondary data from public domain notifications and the independent factors associated with death were analyzed through logistic regression. The value ofp<0.25 was considered significant in the bivariate analysis and for a multivariate analysis, the value ofp<0.05 was considered significant. Results: forty-nine women died. The mean age and standard deviation were 58.75 ± 20.93 years. 55.86% of the patients were 60 years old or more. The overall lethality rate was 72.06% (CI95%=59.8 - 82.2). The most prevalent symptoms were fever (70.59%), cough (58.82%), dyspnea (57.35%) and O2 saturation less than 95% (48.53%). Conclusions: female patients, with cancer and infected by SARS-CoV-2 are particularly susceptible to death, regardless of the presence of comorbidities or age, with peripheral O2 saturation <95% being the only independent factor associated with death in this group.

Resumo Objetivos: analisar a letalidade e características clínicas em mulheres pernambucanas portadoras de neoplasia que apresentaram infecção por SARS-CoV-2. Métodos: estudo de corte transversal, retrospectivo com pacientes do sexo feminino, portadoras de neoplasias no estado de Pernambuco com registros disponibilizados pela Secretaria de Planejamento e Gestão do Estado de Pernambuco. Analisou-se dados secundários de notificações de domínio público e os fatores independentes associados ao óbito através de regressão logística. Foi considerado significativo o valor de p<0,25 na análise bivariada e para a análise multivariada foi considerado significativo o valor de p<0,05. Resultados: quarenta e nove mulheres vieram a óbito. A média da idade e desvio padrão foram 58, 75 ± 20,93 anos. 55,86% das pacientes tinham 60 anos ou mais. A taxa de letalidade global foi de 72,06% (IC95%= 59,8 - 82,2). Os sintomas mais prevalentes foram febre (70,59%), tosse (58,82%), dispneia (57,35%) e saturação de O2 <95% (48,53%). Conclusão: pacientes do sexo feminino, com câncer e infectadas pelo SARS-CoV-2 são particularmente suscetíveis a óbito, independentemente da presença de comorbidades ou da idade, sendo a saturação periférica de O2 <95% o único fator independente associado ao óbito nesse grupo.

Death risk and the importance of clinical features in elderly people with COVID-19 using the Random Forest Algorithm / Previsão de óbito e importância de características clínicas em idosos com COVID-19 utilizando o Algoritmo Random Forest

Abstract Objectives: train a Random Forest (RF) classifier to estimate death risk in elderly people (over 60 years old) diagnosed with COVID-19 in Pernambuco. A "feature" of this classifier, called feature importance, was used to identify the attributes (main risk factors) related to the outcome (cure or death) through gaining information. Methods: data from confirmed cases of COVID-19 was obtained between February 13 and June 19, 2020, in Pernambuco, Brazil. The K-fold Cross Validation algorithm (K=10) assessed RF performance and the importance of clinical features. Results: the RF algorithm correctly classified 78.33% of the elderly people, with AUC of 0.839. Advanced age was the factor representing the highest risk of death. The main comorbidity and symptom were cardiovascular disease and oxygen saturation ≤ 95%, respectively. Conclusion: this study applied the RF classifier to predict risk of death and identified the main clinical features related to this outcome in elderly people with COVID-19 in the state of Pernambuco.

Resumo Objetivos: treinar um classificador do tipo Random Forest (RF) para estimar o risco de óbito em idosos (com mais de 60 anos) diagnosticados com COVID-19 em Pernambuco. Uma "feature" deste classificador, chamada feature_importance, foi usada para identificar os atributos (principais fatores de risco) relacionados com o desfecho final (cura ou óbito) através do ganho de informação. Métodos: dados dos casos confirmados de COVID-19foram obtidos entre os dias 13 de fevereiro e 19 de junho de 2020, em Pernambuco, Brasil. O algoritmo K-fold Cross Validation, com K=10, foi usado para avaliar tanto o desempenho do RF quanto a importância das características clínicas. Resultados: o algoritmo RF classificou corretamente 78,33% dos idosos, com AUC de 0,839. A idade avançada é o fator que representa maior risco de evolução para óbito. Além disso, a principal comorbidade e sintoma também identificados, foram, respectivamente, doença cardiovascular e saturação de oxigênio ≤95%. Conclusão: este trabalho se dedicou à aplicação do classificador RF para previsão de óbito e identificou as principais características clínicas relacionadas com este desfecho em idosos com COVID-19 no estado de Pernambuco.

Developing Machine Learning Algorithms for the Prediction of Early Death in Elderly Cancer Patients: Usability Study.

BACKGROUND: The importance of classifying cancer patients into high- or low-risk groups has led many research teams, from the biomedical and bioinformatics fields, to study the application of machine learning (ML) algorithms. The International Society of Geriatric Oncology recommends the use of the comprehensive geriatric assessment (CGA), a multidisciplinary tool to evaluate health domains, for the follow-up of elderly cancer patients. However, no applications of ML have been proposed using CGA to classify elderly cancer patients. OBJECTIVE: The aim of this study was to propose and develop predictive models, using ML and CGA, to estimate the risk of early death in elderly cancer patients. METHODS: The ability of ML algorithms to predict early mortality in a cohort involving 608 elderly cancer patients was evaluated. The CGA was conducted during admission by a multidisciplinary team and included the following questionnaires: mini-mental state examination (MMSE), geriatric depression scale-short form, international physical activity questionnaire-short form, timed up and go, Katz index of independence in activities of daily living, Charlson comorbidity index, Karnofsky performance scale (KPS), polypharmacy, and mini nutritional assessment-short form (MNA-SF). The 10-fold cross-validation algorithm was used to evaluate all possible combinations of these questionnaires to estimate the risk of early death, considered when occurring within 6 months of diagnosis, in a variety of ML classifiers, including Naive Bayes (NB), decision tree algorithm J48 (J48), and multilayer perceptron (MLP). On each fold of evaluation, tiebreaking is handled by choosing the smallest set of questionnaires. RESULTS: It was possible to select CGA questionnaire subsets with high predictive capacity for early death, which were either statistically similar (NB) or higher (J48 and MLP) when compared with the use of all questionnaires investigated. These results show that CGA questionnaire selection can improve accuracy rates and decrease the time spent to evaluate elderly cancer patients. CONCLUSIONS: A simplified predictive model aiming to estimate the risk of early death in elderly cancer patients is proposed herein, minimally composed by the MNA-SF and KPS. We strongly recommend that these questionnaires be incorporated into regular geriatric assessment of older patients with cancer.