ABSTRACT
INTRODUCTION: Legionellosis is a multisystem bacterial disease, which causes pneumonia with high mortality in patients with comorbidity and admitted in intensive care units (ICU). OBJECTIVE: Determine predictors of mortality or ICU admission. METHODS: Retrospective follow-up of patients diagnosed with Legionella pneumophila pneumonia in Complexo Hospitalario Universitario de A Coruña. Period 2000-2013 (n=240). Analysis of multivariate logistic regression was performed. RESULTS: Mean age was 57.2±15.4 years old, 88.3% were male. Average score of comorbidity (Charlson score) was 2.3±2.3. There was a clear seasonal variation. Predominant symptoms were fever (92.5%), dry cough (38.1%) and dyspnea (33.9%). Creatinine clearance was lower than 60 mL/min/1.73 m² in 29.7% and sodium<135 mEq/l in 58.3%. Admission to ICU rate was 16.3% and 10.8% needs mechanical ventilation. Inhospital mortality rate was 4.6%, rising to 23.1% in patients admitted to ICU. Variables associated to predict ICU admission were age (OR=0.96), liver disease (OR=7.13), dyspnea (OR=4.33), delirium (OR=5.86) and high levels of lactatedehydrogenase (OR=1.002). Variables associated with inhospital mortality were Charlson index (OR=1.70), mechanical ventilation (OR=31.44) and high levels of lactatedehydrogenase (OR=1.002). DISCUSSION: Younger patients with liver disease, dyspnea and confusion are more likely to be admitted to ICU. Comorbidity, mechanical ventilation and elevated LDH levels are associated with higher mortality rate.
Subject(s)
Hospitalization , Legionella pneumophila , Legionnaires' Disease/diagnosis , Pneumonia, Bacterial/microbiology , Adult , Age Factors , Aged , Comorbidity , Creatinine/metabolism , Delirium/epidemiology , Dyspnea/epidemiology , Female , Hospital Mortality , Hospitalization/statistics & numerical data , Humans , Intensive Care Units/statistics & numerical data , L-Lactate Dehydrogenase/blood , Legionnaires' Disease/mortality , Liver Diseases/epidemiology , Logistic Models , Male , Middle Aged , Pneumonia, Bacterial/mortality , Prognosis , Retrospective Studies , Seasons , Spain/epidemiologyABSTRACT
Introduction: Legionellosis is a multisystem bacterial disease, which causes pneumonia with high mortality in patients with comorbidity and admitted in intensive care units (ICU). Objective: Determine predictors of mortality or ICU admission. Methods: Retrospective follow-up of patients diagnosed with Legionella pneumophila pneumonia in Complexo Hospitalario Universitario de A Coruña. Period 2000-2013 (n = 240). Analysis of multivariate logistic regression was performed. Results: Mean age was 57.2 ± 15.4 years old, 88.3% were male. Average score of comorbidity (Charlson score) was 2.3 ± 2.3. There was a clear seasonal variation. Predominant symptoms were fever (92.5%), dry cough (38.1%) and dyspnea (33.9%). Creatinine clearance was lower than 60 mL/min/1.73 m² in 29.7% and sodium < 135 mEq/l in 58.3%. Admission to ICU rate was 16.3% and 10.8% needs mechanical ventilation. Inhospital mortality rate was 4.6%, rising to 23.1% in patients admitted to ICU. Variables associated to predict ICU admission were age (OR = 0.96), liver disease (OR = 7.13), dyspnea (OR = 4.33), delirium (OR = 5.86) and high levels of lactatedehydrogenase (OR = 1.002). Variables associated with inhospital mortality were Charlson index (OR = 1.70), mechanical ventilation (OR = 31.44) and high levels of lactatedehydrogenase (OR = 1.002). Discussion: Younger patients with liver disease, dyspnea and confusion are more likely to be admitted to ICU. Comorbidity, mechanical ventilation and elevated LDH levels are associated with higher mortality rate.
Introducción: La legionelosis es una enfermedad bacteriana multisistémica, causante de neumonías con mortalidad elevada en pacientes con comorbilidad e ingresos en Unidad de Cuidados Intensivos (UCI). Objetivo: Determinar factores pronósticos de mortalidad o ingreso en UCI. Material y Métodos: Estudio de seguimiento retrospectivo de pacientes diagnosticados de neumonía por Legionella pneumophila en Complexo Hospitalario Universitario de A Coruña (España). Período 2000-2013 (n = 240), con análisis de regresión logística multivariada. Resultados: La edad media fue 57,2 ± 15,4 años, 88,3% fueron hombres. La puntuación media de comorbilidad (score Charlson) fue 2,3 ± 2,3. Existe clara estacionalidad. La clínica predominante fue fiebre (92,5%), tos seca (38,1%) y disnea (33,9%). El 29,7% presentó aclaramiento de creatinina < 60 mL/min/1,73 m² y el 58,3% sodio < 135 mEq/l. Un 16,3% ingresó en UCI, precisando ventilación mecánica invasiva el 10,8%. La mortalidad global fue 4,6% y de 23,1% en ingresados en UCI. Variables asociadas para predecir ingreso en UCI fueron menor edad (OR = 0,96), hepatopatía (OR = 7,13), disnea (OR = 4,33), síndrome confusional (OR = 5,86) y lactato deshidrogenasa elevada (OR = 1,002). Las variables asociadas a mortalidad intrahospitalaria fueron índice de Charlson (OR = 1,70), ventilación mecánica invasiva (OR = 31,44) y cifras elevadas de lactato deshidrogenasa (OR = 1,002). Discusión: Pacientes jóvenes, con hepatopatía, disnea o confusión tienen más probabilidad de ingresar en UCI. Comorbilidad, ventilación mecánica y lactato deshidrogenasa elevada se asocian a mortalidad.
Subject(s)
Adult , Aged , Female , Humans , Male , Middle Aged , Hospitalization , Legionella pneumophila , Legionnaires' Disease/diagnosis , Pneumonia, Bacterial/microbiology , Age Factors , Comorbidity , Creatinine/metabolism , Delirium/epidemiology , Dyspnea/epidemiology , Hospital Mortality , Hospitalization/statistics & numerical data , Intensive Care Units/statistics & numerical data , L-Lactate Dehydrogenase/blood , Logistic Models , Legionnaires' Disease/mortality , Liver Diseases/epidemiology , Prognosis , Pneumonia, Bacterial/mortality , Retrospective Studies , Seasons , Spain/epidemiologyABSTRACT
OBJECTIVE: The aim of this study was to describe 5 microepidemics of tuberculosis occurring in schools, establish the risk factors associated with the outbreaks, assess how well a concentric circles strategy for contact tracing predicts infection, and assess the usefulness of genotyping strains in the analysis of the outbreaks. MATERIAL AND METHODS: The study assessed 5 epidemic outbreaks of tuberculosis using a standard contact tracing procedure. The outbreaks occurred in 2 day nurseries and 2 high schools between 1998 and 2005. Contacts were stratified using a concentric circle system based on level of exposure. DNA fingerprints of the available strains were determined based on the restriction fragment length polymorphism (RFLP) IS6110 and compared with the contact study to interpret the transmission of the infection. RESULTS: We analyzed 5 outbreaks. Eighty-five contacts were analyzed in the first outbreak, 519 in the second, 116 in the third, 655 in the fourth, and 102 in the fifth. The rate of infection was 31%, 29%, 66%, 37.6%, and 32%, respectively. Secondary cases of active disease were detected: 9 in the first outbreak, 16 in the second, 5 in the third, 6 in the fourth, and 13 in the fifth. RFLP analysis revealed that a single strain was involved in 3 of the outbreaks, and in a fourth, at least 2 strains were involved. In outbreaks 2, 3, and 5, there was a significant association between the degree of contact and the probability of infection (P< .05). In all of the outbreaks, the relative risk of developing the disease was associated with the level of exposure. CONCLUSIONS: Analysis of contacts based on concentric circles of risk predicts the likelihood of infection. RFLP facilitates analysis of complex transmission routes that are not detected using traditional methods of contact screening.
Subject(s)
Data Interpretation, Statistical , Disease Outbreaks , Genotype , Tuberculosis, Pulmonary , Adolescent , Female , Humans , Incidence , Male , Polymorphism, Restriction Fragment Length/genetics , Prevalence , Prospective Studies , Schools , Students/statistics & numerical data , Tuberculosis, Pulmonary/epidemiology , Tuberculosis, Pulmonary/genetics , Tuberculosis, Pulmonary/transmissionABSTRACT
Objetivo: El propósito del presente estudio ha sido describir 5 microepidemias de tuberculosis en centros escolares, establecer los factores de riesgo relacionados con su aparición, valorar la predicción de infección mediante la estrategia de círculos concéntricos y analizar la utilidad de la tipificación genética de cepa en su interpretación. Material y métodos: Se presenta el estudio de 5 brotes de tuberculosis, y su estudio convencional de contactos, ocurridos en 2 guarderías y 2 institutos de enseñanza media entre 1998 y 2005. Los contactos se estratificaron según el grado de convivencia siguiendo el sistema de círculos concéntricos. Se identificaron los perfiles genéticos (RFLP polimorfismo de longitud de los fragmentos de restricción IS6110) de las cepas disponibles y se cotejaron los resultados con el estudio de contactos para interpretar la transmisión de la infección. Resultados: Analizamos 5 brotes. En el primero estudiamos a 85 contactos; en el segundo, a 519; en el tercero, a 116; en el cuarto, a 655, y 102 en el quinto. La prevalencia de infección fue del 31, el 29, el 66, el 37,6 y el 32%, respectivamente. Se detectaron casos secundarios de enfermedad activa: 9 en el primero, 16 en el segundo, 5 en el tercero, 6 en el cuarto y 13 en el quinto. El análisis mediante RFLP identificó la coincidencia genética de todas las cepas en 3 brotes, y en el cuarto evidenció la existencia de al menos 2 cepas implicadas en su desarrollo. En los brotes 2, 3 y 5 encontramos una asociación importante entre el grado de convivencia y las probabilidades de infectarse (p < 0,05). En todos los brotes el riesgo relativo de infectarse y enfermar se asoció con la intensidad de la exposición. Conclusiones: El estudio de contactos basado en círculos de riesgo predice la probabilidad de infección. La RFLP permite aclarar vías de transmisión complejas que no son detectables mediante el estudio convencional de contactos
Objective: The aim of this study was to describe 5 microepidemics of tuberculosis occurring in schools, establish the risk factors associated with the outbreaks, assess how well a concentric circles strategy for contact tracing predicts infection, and assess the usefulness of genotyping strains in the analysis of the outbreaks. Material and methods: The study assessed 5 epidemic outbreaks of tuberculosis using a standard contact tracing procedure. The outbreaks occurred in 2 day nurseries and 2 high schools between 1998 and 2005. Contacts were stratified using a concentric circle system based on level of exposure. DNA fingerprints of the available strains were determined based on the restriction fragment length polymorphism (RFLP) IS6110 and compared with the contact study to interpret the transmission of the infection. Results: We analyzed 5 outbreaks. Eighty-five contacts were analyzed in the first outbreak, 519 in the second, 116 in the third, 655 in the fourth, and 102 in the fifth. The rate of infection was 31%, 29%, 66%, 37.6%, and 32%, respectively. Secondary cases of active disease were detected: 9 in the first outbreak, 16 in the second, 5 in the third, 6 in the fourth, and 13 in the fifth. RFLP analysis revealed that a single strain was involved in 3 of the outbreaks, and in a fourth, at least 2 strains were involved. In outbreaks 2, 3, and 5, there was a significant association between the degree of contact and the probability of infection (P<.05). In all of the outbreaks, the relative risk of developing the disease was associated with the level of exposure. Conclusions: Analysis of contacts based on concentric circles of risk predicts the likelihood of infection. RFLP facilitates analysis of complex transmission routes that are not detected using traditional methods of contact screening
