[Nosocomial Candida infections and thrombocytopenia in very low birth weight newborns]. / Infecciones nosocomiales por Candida y trombocitopenia en recién nacidos de muy bajo peso.

OBJECTIVE: To analyze thrombocytopenia as an early marker of late-onset neonatal Candida sepsis and to determine whether there are significant differences in platelet count between neonates infected by different infectious agents. MATERIAL AND METHODS: We retrospectively reviewed the medical charts of 42 neonates with late-onset sepsis with positive blood culture in our neonatology service between January 1, 2003 and December 31, 2004. Only the first episode of sepsis in any single patient was included. The principal parameter examined was the incidence of thrombocytopenia according to the microorganisms causing the sepsis (Gram-positive bacteria, Gram negative bacteria and fungi). The statistical analysis was performed using SPSS version 11.5. The data are expressed as mean+/-standard deviation. The chi-squared test was used to compare qualitative variables. For quantitative variables, the Kolmogorov-Smirnov was used; ANOVA was used for parametric variables and the Kruskal-Wallis test for non-parametric variables. A value of p<0.5 was considered significant. RESULTS: Late-onset sepsis was diagnosed in 42 (2.3%) of 1,757 newborn admissions. Gestational age at birth was 31+/-4.9 weeks (24-41.5 weeks) with a mean birth weight of 1,618+/-911 g (750-4,070 g). There were 34 boys (81%) and eight girls (19%). When patients infected with the three groups of microorganisms were compared, no significant differences were found except for birth weight, days of stay in the neonatal intensive care unit, thoracotomy, days of mechanical ventilation, antibiotic therapy before sepsis, and thrombocytopenia. Gram-positive sepsis was found in 47.6%, Gram-negative sepsis in 33 % and fungal sepsis in 19%. The only cause of fungal sepsis was Candida, mainly C. glabrata (50%), followed by C. albicans (37.5%) and C. parapsilosis (12.5%). The most reliable marker of Candida sepsis was thrombocytopenia, which was found in 17.7% of the episodes of late-onset sepsis. The incidence of thrombocytopenia was significantly higher in Candida sepsis than in bacterial sepsis (100% vs 5.9%) (p<0.001). Only two patients with bacterial sepsis had thrombocytopenia and both cases were caused by Gram-negative bacteria. CONCLUSION: Thrombocytopenia is a highly specific marker of neonatal Candida sepsis. Analysis of platelet counts is a simple laboratory test that helps to guide diagnosis and the use of early empirical therapy.

Infecciones nosocomiales por Candida y trombocitopenia en recién nacidos de muy bajo peso / Nosocomial Candida infections and thrombocytopenia in very low birth weight newborns

Objetivo Analizar la trombocitopenia como marcador precoz de las sepsis tardías por Candida. Determinar la existencia en los recién nacidos sépticos de diferencias en el recuento plaquetario según el agente etiológico responsable de la infección. Material y métodos: Revisión retrospectiva de 42 recién nacidos con sepsis tardías confirmadas por hemocultivo en nuestro servicio desde el 1 de enero de 2003 hasta el 31 de diciembre de 2004, incluyéndose únicamente los primeros episodios infecciosos. El principal parámetro analizado fue la incidencia de trombocitopenia en función de los gérmenes causantes de sepsis (bacterias grampositivas, gramnegativas y hongos). El análisis estadístico se realizó con el paquete estadístico SPSS versión 11.5. Los datos se expresan como media desviación estándar. Para comparar los grupos se utilizaron en el caso de variables cualitativas la prueba de chi al cuadrado y en el caso de las variables cuantitativas tras realizar la prueba de Kolmogorov-Smirnov se usaron para las variables paramétricas la prueba de ANOVA y para las variables no paramétricas la prueba de Kruskal-Wallis. Se consideró significativo un valor de p ≤ 0,05. Resultados: De un total de 1.757 recién nacidos se diagnosticaron 42 sepsis tardías (2,3 %). La edad de gestación y peso de los recién nacidos sépticos fueron de 31 4,9 semanas (24-41,5) y 1.618 911 g (750-4.070 g). Había 34 varones (81 %) y 8 mujeres (19 %). El estudio comparativo no mostró diferencias estadísticamente significativas entre los tres grupos de gérmenes, excepto en el peso al nacer, días de estancia en la unidad de cuidados intensivos (UCI), días de tubo torácico, días de ventilación mecánica, antibioterapia previa a la sepsis y trombocitopenia. Fueron sepsis grampositivas el 47,6 %; gramnegativas, el 33%, y fúngicas, el 19 %. La única responsable de las sepsis fúngicas fue Candida, principalmente C. glabrata (50 %), seguida de C. albicans (37,5 %) y C. parapsilosis (12,5 %). El marcador de infección de sepsis por Candida más importante fue la trombocitopenia, que se constató en el 17,7 % de todas las sepsis tardías. Fue significativamente mayor en las fúngicas que en las no fúngicas (100 % frente a 5,9 %) (p < 0,001), aunque sólo se pudo comprobar en dos casos de sepsis bacterianas por gramnegativos. Conclusión La trombocitopenia es un marcador muy específico de las sepsis neonatal por Candida. El recuento plaquetario es una prueba sencilla que facilita la orientación diagnóstica y la instauración de un tratamiento empírico precoz

Objective To analyze thrombocytopenia as an early marker of late-onset neonatal Candida sepsis and to determine whether there are significant differences in platelet count between neonates infected by different infectious agents. Material and methods We retrospectively reviewed the medical charts of 42 neonates with late-onset sepsis with positive blood culture in our neonatology service between January 1, 2003 and December 31, 2004. Only the first episode of sepsis in any single patient was included. The principal parameter examined was the incidence of thrombocytopenia according to the microorganisms causing the sepsis (Gram-positive bacteria, Gram negative bacteria and fungi). The statistical analysis was performed using SPSS version 11.5. The data are expressed as mean standard deviation. The chi-squared test was used to compare qualitative variables. For quantitative variables, the Kolmogorov-Smirnov was used; ANOVA was used for parametric variables and the Kruskal-Wallis test for non-parametric variables. A value of p ≤ 0.5 was considered significant. Results Late-onset sepsis was diagnosed in 42 (2.3 %) of 1,757 newborn admissions. Gestational age at birth was 31 4.9 weeks (24-41.5 weeks) with a mean birth weight of 1,618 911 g (750-4,070 g). There were 34 boys (81%) and eight girls (19 %). When patients infected with the three groups of microorganisms were compared, no significant differences were found except for birth weight, days of stay in the neonatal intensive care unit, thoracotomy, days of mechanical ventilation, antibiotic therapy before sepsis, and thrombocytopenia. Gram-positive sepsis was found in 47.6 %, Gram-negative sepsis in 33 % and fungal sepsis in 19 %. The only cause of fungal sepsis was Candida, mainly C. glabrata (50 %), followed by C. albicans (37.5 %) and C. parapsilosis (12.5 %). The most reliable marker of Candida sepsis was thrombocytopenia, which was found in 17.7 % of the episodes of late-onset sepsis. The incidence of thrombocytopenia was significantly higher in Candida sepsis than in bacterial sepsis (100% vs 5.9 %) (p < 0.001). Only two patients with bacterial sepsis had thrombocytopenia and both cases were caused by Gram-negative bacteria. Conclusion Thrombocytopenia is a highly specific marker of neonatal Candida sepsis. Analysis of platelet counts is a simple laboratory test that helps to guide diagnosis and the use of early empirical therapy