| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 626.08 KB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
As infeções durante o tratamento em doentes com neoplasias hematológicas constitutem um fator de morbilidade e mortalidade importante neste grupo de doentes. O
conhecimento epidemiológico das intercorrências infeciosas é uma ferramenta
importante na sua abordagem terapêutica. Os métodos de regressão são uma componente fundamental da análise de dados que pretende estabelecer a relação entre um resultado e uma ou mais variáveis preditivas. Este estudo teve como objetivo identificar, recorrendo a métodos de regressão logística, os fatores de risco associados ao internamento por infeção em doentes com doença hematológica tratados num centro não académico. As causas mais frequentes de internamento foram neutropenia febril sem foco (28%), infeção do trato urinário (16%) e pneumonia (8%), sendo os agentes microbiológicos mais frequentemente isolados bactérias gram-negativas. Na análise multivariada foi identificado como fator protetor o valor de hemoglobina (OR=0,6, IC 95% 0,42-0,89) e como fator de risco o tratamento antineoplásico prévio (OR=4,8, IC 95% 3,35-6,98).
Infections during treatment of patients with hematological neoplasms are an important morbidity and mortality factor. The knowledge in the epidemiology of infections in this group of patients is an important tool in its therapeutic approach. Regression methods are a fundamental component of data analysis for estimating relationships between an outcome and one or more predictive variables. This study aimed to identify, using logistic regression methods, the risk factors linked with hospitalization for infection in patients with hematological neoplasms treated in a non-academic centre. The most frequent causes of hospitalization were febrile neutropenia (28%), urinary tract infections (16%) and pneumonia (8%) and the most frequently isolated microbiological agents were gram negative bacteria. In the multivariate analysis, the hemoglobin value on diagnosis had a protective effect (OR 0.6, IC 95% 0,42-0,89) while previous antineoplastic treatment (OR 4.8, IC 95% 3,35-6,98) was identified as a risk factor.
Infections during treatment of patients with hematological neoplasms are an important morbidity and mortality factor. The knowledge in the epidemiology of infections in this group of patients is an important tool in its therapeutic approach. Regression methods are a fundamental component of data analysis for estimating relationships between an outcome and one or more predictive variables. This study aimed to identify, using logistic regression methods, the risk factors linked with hospitalization for infection in patients with hematological neoplasms treated in a non-academic centre. The most frequent causes of hospitalization were febrile neutropenia (28%), urinary tract infections (16%) and pneumonia (8%) and the most frequently isolated microbiological agents were gram negative bacteria. In the multivariate analysis, the hemoglobin value on diagnosis had a protective effect (OR 0.6, IC 95% 0,42-0,89) while previous antineoplastic treatment (OR 4.8, IC 95% 3,35-6,98) was identified as a risk factor.
Descrição
Palavras-chave
Saúde pública Estatística Tratamento de neoplasias hematológicas Morbilidade Regressão logística Factores de risco de infeções
