As taxas de infecção hospitalar (IH) entre centros neonatais variam consideravelmente, sugerindo que fatores de risco possam ser modificados pela qualidade da assistência, as características do recém-nascido (RN) e o controle das infecções. O objetivo deste estudo foi analisar a epidemiologia da infecção e da mortalidade hospitalar na Unidade de Terapia Intensiva Neonatal do Hospital Geral de Itapecerica da Serra SECONCI SP OSS de 1º de janeiro de 2002 a 31 de dezembro de 2003. O estudo foi desenvolvido em modelo de coorte e a análise dos dados referentes às IH precoces e tardias foi retrospectiva, mas eles foram coletados prospectivamente, seguindo os métodos do NNIS (National Nosocomial Infection Surveillance System). Os RN foram classificados pelo Neonatal Therapeutic Intervention Scoring System (NTISS) para avaliar sua gravidade, dentro das primeiras 24 horas após a admissão. Foram incluídos no estudo 486 RN: 426 de origem interna (87,7%) e 60 de origem externa (12,3%). A incidência acumulada de IH foi de 30,6% e a densidade de 25,1 por 1.000 pacientes-dia (7,9 para infecção precoce e 17,2 para a tardia). A sepse foi o tipo de infecção mais freqüente (54,0%) seguida pela pneumonia (20,0%). Dos agentes microbianos isolados, 54,3% foram gram-positivos, sendo o mais encontrado o Staphylococcus coagulase negativo. A maioria dos RN teve um escore de gravidade menor ou igual a 19 (88,1%), sendo a pontuação máxima encontrada de 39, e os RN externos obtiveram uma pontuação significantemente maior. A aquisição de IH, tanto precoce como tardia, mostrou-se associada com a gravidade do RN à admissão. A taxa de mortalidade hospitalar foi de 8,6% e mostrou-se mais elevada entre os RN de origem externa. As IH foram a causa ou contribuíram para o óbito em 26 (61,9%) dos RN que faleceram. Não houve associação estatística entre o local de nascimento e a ocorrência de infecção hospitalar precoce e tardia. A análise univariada mostrou os seguintes fatores de risco para infecção tardia: prematuridade, baixo peso, pequeno para a idade gestacional, número de consultas de prénatal, reanimação na sala de parto, uso de respirador, cateter central, nutrição parenteral, tempo de permanência e escore de gravidade à admissão. Na análise múltipla, o modelo final incluiu as variáveis: peso de nascimento, escore terapêutico nas primeiras 24 horas após a admissão e uso de nutrição parenteral. A epidemiologia da infecção hospitalar da UTIN do HGIS está de acordo com o observado na literatura médica. Ela está sujeita às características dos RN assistidos, às práticas assistenciais e de controle de infecção hospitalar implementadas pelo serviço de terapia intensiva neonatal, independente do local de nascimento do RN admitido na UTIN. / Nosocomial infections rates varies widely among Neonatal Centers suggesting that risk factors can be modify by assistance quality, newborn characteristics and infection control practices. The aim of this study was to analyze nosocomial infections epidemiology and mortality rate among neonates admitted to a Neonatal Intensive Care Unit of Hospital Geral de Itapecerica da Serra SECONCI SP OSS from January 1, 2002 to December 31, 2003. The study was carried out in a cohort model, with data analyze retrospectively but collected by active surveillance following the NNIS (National Nosocomial Infection Surveillance System) methodology. Neonates were classified according to Neonatal Therapeutic Intervention Scoring System (NTISS) to assess illness severity within the first 24 hours of admission. 486 newborn infants were included in the study: 426 (87.7%) inborn infants and 60 (12.3%) out born infants. Nosocomial infection incidence rate was 30.6% and the incidence density was 25.1 per 1000 patients-day (7.9 for early infections and 17.2 for late infections). Sepsis was the most frequent infection (54.0%), followed by pneumonia (20.0%). Among microbial agents isolated 54.3% were Gram-positive organisms, and coagulase-negative staphylococci were the most frequent. Most neonates have shown a severity score lower or equal to 19 (88.1%), and the maximum score was 39. Outborn neonates have shown a significant higher severity score. Nosocomial infections were associated with newborn severity index at admission. Nosocomial mortality rate was 8.6% and higher among out born neonates. Hospital infections were classified as cause or contributed for death in 26 (61.9%) neonates. No statistic association was seen between the neonate birth place and nosocomial infections. Univariate analyzes showed the following risk factors for late infections: prematurity, low birth weigh, low weight for gestational age, prenatal visits number, resuscitation following birth, respirator, central catheter and parenteral nutrition use, length of stay and severity score at admission. Multivariate logistic regression model included the following variables: birth weigh, therapeutic score within 24 hours of admission and parenteral nutrition use. Nosocomial infection epidemiology at HGIS´s UTIN is similar with what was observed in medical literature. It is dependent of newborn characteristics, assistance and infection control practices within the neonatal intensive care therapy, and is independent of newborn place of birth

This study presents an analysis of policies addressing child mortality in Pakistan focusing on problem structuring, using a comparison with Bangladesh. Pakistan's progress in addressing child mortality rate has been much slower than that of Bangladesh despite the fact that Pakistan has excelled in economic growth, and the two countries have comparable populations and share political history. This study analyzes and provides an explanation for differential outcomes in terms of problem structuring in the two countries. A comparative analysis of policy documents reviewed for the two countries illustrates the fact that Bangladesh, in formulating its child health policy, has emphasized the input factors such as nutrition and environmental aspects, besides health services. Pakistan, on the other hand, maintains a general problem formulation strategy focusing mainly on health service and ignoring the social, environmental, and other factors causing morbidity and mortality in children. Another comparison between policy formulation in each country and the extensive literature available on child mortality suggest that neither country pays as much attention to structural factors as the literature does. The analysis provides some insight into differentials in policy formulation associated with child mortality in the two countries, but more importantly, it provides an understanding of the underlying elements for inadequate policy outcomes in case of Pakistan.

No description available.

Submitted by ÉRICK RENATO FOGAR FACIOLI null (reporterfacioli@gmail.com) on 2017-09-19T23:20:35Z No. of bitstreams: 1 ANÁLISE DA COBERTURA DO JORNAL FOLHA DE SÃO PAULO SOBRE O TEMA MORTALIDADE INFANTIL ENTRE 1990 E 2015.pdf: 2597994 bytes, checksum: eb81788785fd7baedd24fb844568747f (MD5) / Approved for entry into archive by Monique Sasaki (sayumi_sasaki@hotmail.com) on 2017-09-20T17:50:03Z (GMT) No. of bitstreams: 1 facioli_erf_me_bot.pdf: 2650643 bytes, checksum: 4c0c2a3bb3b676d9dcdaa859426e94dd (MD5) / Made available in DSpace on 2017-09-20T17:50:03Z (GMT). No. of bitstreams: 1 facioli_erf_me_bot.pdf: 2650643 bytes, checksum: 4c0c2a3bb3b676d9dcdaa859426e94dd (MD5) Previous issue date: 2017-08-21 / A taxa de mortalidade infantil é indicador sensível para refletir o grau de desenvolvimento de um país. Nas últimas décadas, o Brasil registrou grande variação negativa na taxa desse desfecho, resultado da centralidade nas políticas públicas de saúde materno-infantis nacionais. O conhecimento público sobre os fatores que contribuíram para a melhora desse indicador é importante por fazer da população parte integrante do processo, permitindo maior controle social. Possibilita, ainda, que o cidadão tenha a oportunidade de fazer a sua parte no que diz respeito à prevenção. O objetivo do presente estudo foi avaliar a cobertura jornalística realizada sobre o tema mortalidade infantil na versão impressa do jornal Folha de São Paulo entre 1990 e 2015, período definido pela Declaração do Milênio para o cumprimento da meta de redução em dois terços na taxa de mortalidade infantil global. Trata-se de estudo de abordagem quantiqualitativa, de caráter descritivo e retrospectivo. A escolha do jornal Folha de São Paulo decorreu do fato de constituir o veículo jornalístico de maior circulação no país. Para seleção das matérias, utilizaram-se as palavras-chave “mortalidade” and “infantil” no acervo digital do próprio jornal. A coleta de dados foi realizada entre setembro de 2015 e julho de 2016. No total, 2436 matérias foram inicialmente selecionadas para leitura dinâmica. Aplicaram-se, em seguida, os critérios de exclusão: citações isoladas, comentários dos leitores, publicidade, informes publicitários, gráficos isolados, matérias idênticas publicadas e erratas, restando 429 textos. Como era critério de inclusão analisar as publicações que fizessem referência aos determinantes da mortalidade infantil, nova seleção foi necessária e, nessa etapa, foram obtidas 267 matérias, às quais foram aplicadas o Protocolo de Análise de Cobertura Jornalística, com a função de quantificar e organizar o material. Para a análise qualitativa, foram utilizados os textos com chamada de capa, o que resultou em 82 matérias, avaliadas a partir dos referenciais teórico-metodológicos do enquadramento, agenda-setting e Cenário de Representação Política (CR-P). Os resultados revelam, em geral, cobertura jornalística superficial e sensacionalista do tema, carecendo de análises mais amplas, como as razões históricas, políticas e sociais a ele relacionadas. Houve privilégio a fontes oficiais, com pouco espaço para opiniões de especialistas independentes e do cidadão comum. Por vezes, o Estado foi apresentado como ineficiente para cuidar da saúde da população. / The child mortality rate is a sensitive indicator to reflect a country’s development level. In the last few decades, Brazil has shown a decrease great variation in such rate, which has been the result of centrality in the national public policies on maternal-infant health. Public knowledge about the factors that have contributed to the improvement of this indicator is important as it makes the population an integral part of the process, thus allowing for greater social control. It also provides citizens with the opportunity to play their role as regards prevention. This study aimed at evaluating the journalistic coverage of the topic child mortality in the printed version of the newspaper Folha de São Paulo from 1990 to 2015, which is the period defined by the Millennium Declaration for achieving the goal of reducing global mortality rates by twothirds. It is a study with a quanti-qualitative, descriptive and retrospective approach. The choice for the newspaper Folha de São Paulo was due to the fact that it is the journalistic vehicle with the largest circulation in the country. In order to select the articles, the key words “mortality” and “child” were used in the newspaper’s digital collection. Data were collected from September 2015 to July 2016. In total, 2,436 articles were initially selected for dynamic reading. Isolated citations, readers’ comments, publicity, advertising reports, isolated graphs, identical published articles and errata, were not included in the study. Thus, 429 texts remained. Since the inclusion criterion was the analysis of publications that referred to the determinants of child mortality, a new selection was necessary and, in that phase, 267 articles were obtained to which the Protocol for Analysis of Media Coverage was applied with the purpose to quantify and organize the material. For qualitative analysis, the texts referred to on the first page of the newspaper were used, which resulted in 82 articles that were evaluated based on the theoretical and methodological frameworks of framing, agenda-setting and Political Representation Setting (RS-P). In general, the results showed superficial and sensationalist coverage of the topic that lacked more comprehensive analyses, such as those of the historical, political and social reasons related to it. Official sources were privileged, and little space was opened for opinions of independent experts and for those of common citizens. At times, the State was presented as inefficient to care for the population’s health.

Esta dissertação de mestrado tem por objetivo principal estabelecer os perfis de ocorrência de estresse e mortalidade de dois diferentes grupos pré-históricos de São Paulo - os construtores do sambaqui fluvial Moraes e do sambaqui litorâneo Piaçaguera - relacionando-os com o estilo de vida e subsistência de cada um desses grupos. O período foco de interesse foi o infantil mesmo que os indivíduos adultos estivessem sendo analisados. Foram considerados para tanto 88 indivíduos de Piaçaguera e 55 de Moraes, todos eles analisados para o estabelecimento do perfil de mortalidade. Já para a análise de hipoplasias foram considerados, respectivamente 28 e 21 indivíduos, que cumpriram os critérios de inclusão definidos para a análise. Como resultado observou-se uma alta mortalidade na infância em ambas as séries, no entanto com diferenças entre os períodos mais afetados. A série de Piaçaguera apresentou maior risco de morrer em períodos mais tardios da infância enquanto que Moraes apresentou maior risco ainda no período neonatal. Já ao se considerar as hipoplasias lineares de esmalte, há uma sinalização no sentido da ocorrência de estresse relativamente contínuo durante toda a fase intermediária da infância em Moraes, que não apresentou picos de ocorrência de defeitos hipoplásicos. Já para Piaçaguera, os dados sugerem a ocorrência de um período de maior ocorrência/susceptibilidade ao estresse, o qual corresponde a um pico de prevalência de defeitos hipoplásicos entre 3 e 4 anos. / This dissertation has as main objective to establish the profiles of occurrence of stress and mortality on two different prehistoric groups from São Paulo State, Brazil - the builders of the riverine shellmound Moraes and the builders of the coastal shellmound Piaçaguera - relating them to the lifestyle and subsistency of each of these groups. The period focus of interest was the childhood, even if the adults were being analyzed. In order to establish the mortality profile were analized 88 individuals from Piaçaguera and 55 individuals from Moraes. For the analysis of hypoplasias were considered, respectively, 28 and 21 individuals which met the inclusion criteria defined for the analysis. There is possible to stabilish that if the child mortality is very high in both series, there are strong differences between the ages periods more affecteds. The Piaçaguera\'s skeletal serie showed a higher risk of dying in later periods of childhood while Moraes had a higher risk in the neonatal period. Considering the data obteined for linear enamel hypoplasia, there is a signal towards the occurrence of stress relatively continuous throughout the intermediate phase of childhood in Moraes. For this sample there are no peaks of occurrence of hypoplastic defects. In contrast, the data produced for Piaçaguera\'s sample suggest the occurrence of a period of higher incidence / susceptibility to stress, which corresponds to a peak prevalence of hypoplastic defects between 3 and 4 years old .

Magister Public Health - MPH / Aim: The aim of this study was to assess the prescribing practices and availability of antimalarial drugs to children under five years of age in primary health care facilities in Bondo district.

INTRODUÇÃO: As infecções hospitalares (IH) representam uma causa importante de morbidade e mortalidade em crianças criticamente enfermas. Há poucos trabalhos na faixa etária pediátrica e a maioria deles demonstra que a infecção de corrente sangüínea (ICS) é a causa mais importante de IH em pacientes graves. O OBJETIVO deste estudo foi analisar a epidemiologia das infecções de corrente sangüínea de origem hospitalar em crianças internadas no Instituto da Criança do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo hospital de ensino e de assistência terciária, no período de janeiro de 1996 a agosto de 2003. MÉTODOS: O estudo foi feito no modelo de coorte, com análise retrospectiva de dados referentes às infecções de corrente sangüínea, coletados por método de vigilância ativa, seguindo os métodos validados pelo NNIS (National Nosocomial Infections Surveillance System). RESULTADOS: As infecções hospitalares (n = 4233) foram uma causa significativa de morbidade intra-hospitalar no local e período estudados. O risco de um paciente desenvolver uma ou mais infecções hospitalares foi de 11,5 para 100 saídas. As ICS representaram mais de um terço das IH nos oito anos analisados, com uma densidade de incidência que variou de 20,4 na Oncologia, 7,7 na UTI Neonatal, 7,3 na Pediátrica até 1,9 por 1000 pacientes-dia na Cirurgia. Ocorreram com maior freqüência em crianças com idade 5 anos (70,0%), com cateter venoso central (66,7%), e com doenças de base graves (80,4%). Pelo menos um agente infeccioso foi isolado em 78,9% dos episódios de ICS, sendo 41,5% gram-positivos e 44,8% gram-negativos. O microrganismo mais freqüente foi o Staphylococcus coagulase negativo (22,7%). A resistência do S. aureus e dos Staphylococcus coagulase negativos à oxacilina atingiu 58,9 e 80,3%, respectivamente. As cepas dos principais gram-negativos isolados (Klebsiella spp, Enterobacter spp, Pseudomonas spp e E.coli) mostraram-se amplamente resistentes à ceftriaxona, ao aztreonam e, em cerca de 35 a 57%, aos aminoglicosídeos. As ICS foram a causa ou contribuinte para o óbito em 21,9% dos pacientes, mas durante o período do estudo houve um decréscimo significante na mortalidade dos pacientes com ICS. CONCLUSÕES: A ICS foi uma causa importante de morbidade e mortalidade em pacientes pediátricos, predominando em crianças jovens e com doenças de base graves. Os principais fatores associados a ICS incluíram o uso de cateter vascular central e doença de base grave. Patógenos gram-negativos predominaram em todos os anos. O diagnóstico e terapêutica precoce são essenciais para a prevenção da morbidade e mortalidade e com a caracterização das ICS de origem hospitalar, pode-se auxiliar o programa de prevenção destas infecções e suas repercussões / INTRODUCTION: The hospital infections (HI) are main causes of morbidity and mortality in critically ill children. There are only few studies in pediatric age groups, and most of them demonstrated that the bloodstream infection (BSI) is the most important cause of HI in critically ill children. The OBJECTIVE of this study was to analyze the epidemiology of the nosocomial bloodstream infections (BSI) in children admitted to Instituto da Criança of Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo a teaching hospital with complex care attendance, from January 1996 to August 2003. METHODS: The study was carried out in a cohort model, with retrospective data analyses regarding bloodstream infections collected through active surveillance, following the methods validated by the National Nosocomial Infection Surveillance System NNIS. RESULTS: The HI represented a significant cause of morbidity in patients admitted to the hospital within period and local studied. The patients risk for developing one or more HI was 11,5 per 100 exits. The BSI represented more than one third of the HI in the eight analyzed years, with a incidence density varying from 20,4 at Oncology, 7,7 on Neonatal Intensive Care Unit, 7,3 at Pediatric Intensive Care Unit up to 1.9 per 1000 patient-days at Surgery. It occurred more frequently on children on the age of < 5 years old (70,0%), with central vein catheter (66,7%), and critically ill (80,4%). At least one infection agent was isolated in 78,9% of the BSI episodes, 41,5% gram-positive and 44,8% gram-negative. The most frequent pathogen was coagulase-negative staphylococci (22,7%). The proportion of S. aureus and coagulase-negative staphylococci methicillin resistant reached 58,9 and 80,3%, respectively. The main isolated gram-negatives (Klebsiella spp, Enterobacter spp, Pseudomonas spp and E.coli) showed thoroughly resistance to ceftriaxone, to aztreonam and on 35 to 57% to aminoglycosides. The BSIs were the cause or contribution for death in 21,9% of the patients, but during the period of this study, there was a significant lowering on mortality rate of patients with BSI. CONCLUSIONS: The BSI was important cause of morbidity and mortality in pediatric patients, predominantly in young and critically ill children. The main factors associated to BSI included the use of central vascular catheter and severe disease. Gram-negative pathogens predominated in every one all the years. The diagnoses and precocious therapy are essential on the prevention of morbidity, mortality and the characterization of nosocomial BSI, and would help on prevention programs of these infections and its repercussions

Pulse oximeters are low-cost, easy to use, and effective at detecting hypoxemia (low blood oxygen levels), a common complication of bronchiolitis, asthma, and pneumonia, the leading infectious cause of death in children worldwide. However, pulse oximeters are often unavailable in lowincome settings, and if available, often underused, yet little research investigates why. In this thesis, I examine pulse oximeter implementation in low-income settings, focusing on Kenyan hospitals as a case study, and using a mixed-methods approach. I conducted a systematic literature review, examining how pulse oximeter use with children at admission to hospital impacts health outcomes; I then conducted quantitative analyses of 28,000 children admitted to seven Kenyan hospitals to determine with which children pulse oximeters are used, and pulse oximetry's impact on treatment provision; these analyses informed the qualitative research component, for which I conducted interviews with 30 healthcare workers (HCWs) and staff in 14 Kenyan hospitals and employed theoretical frameworks to determine how HCWs decide whether to use pulse oximeters, and the barriers to pulse oximetry. I found that pulse oximeter use varies substantially between and within Kenyan hospitals over time. After adjusting for case-mix and signs of illness severity, HCWs were most likely to use pulse oximeters with children with a very high respiratory rate, indrawing and/or who were not alert; children who obtained a pulse oximeter reading were more likely to be prescribed oxygen than if a pulse oximeter was not used; and children with a reading below 90% were more likely to be prescribed oxygen than those with higher readings, suggesting that HCW decision-making is influenced by international and national guidelines. However, HCWs sometimes cannot use pulse oximeters when they intend to, because of insufficient pulse oximeter availability, largely due to inefficient and confusing procurement processes and repair delays. Furthermore, HCWs sometimes use pulse oximeters incorrectly or misinterpret their results, because of insufficient training. Pulse oximeter promotion programme planners can use the recommendations I provide to effectively target barriers to pulse oximeter uptake in low-income settings. Increased pulse oximetry implementation could enable early detection of hypoxemia, improving accurate diagnosis, and supporting prompt, effective treatment, which could help reduce mortality in children needing oxygen, in line with Sustainable Development Goal 3.

Au cours des vingt dernières années, la santé de la reproduction s’est imposée comme une priorité d’action pour le développement des pays du Sud. Au Sénégal, et dans le département de Saint-Louis, d’indiscutables progrès ont été réalisés mais certaines statistiques restent accablantes et les inégalités de santé, loin de se réduire, demeurent et se creusent. Lorsque ces disparités sont appréhendées au prisme du territoire, des configurations particulières se dessinent et font apparaître des fractures spatiales qui dépassent l’opposition classique entre la ville et la campagne. Désormais, certains quartiers et villages partagent de mêmes préoccupations sanitaires, tandis que des localités rurales acquièrent progressivement des équipements et des comportements que l’on réservait autrefois en ville. Comprendre et mesurer les inégalités territoriales de santé de la reproduction pour pouvoir travailler à leur réduction, tel est le sens de cette étude. / For the past twenty years, reproductive health has been a top priority for the developing countries. In Senegal and in The Saint-Louis department, undeniable progress has been achieved but some statistics remain bleak and health inequalities, far from decreasing tend to grow. When we look closer at these disparities over the studied region, one can observe particular configurations with layouts highlighting spatial splits which go beyond the known urban/rural divide. From now on urban areas and villages share the same sanitary concerns, while rural communities gradually acquire equipment and new habits previously reserved for towns. To understand and estimate the territorial reproductive health inequalities in order to get them to decrease, such is the goal of this study.

INTRODUÇÃO: As infecções hospitalares (IH) representam uma causa importante de morbidade e mortalidade em crianças criticamente enfermas. Há poucos trabalhos na faixa etária pediátrica e a maioria deles demonstra que a infecção de corrente sangüínea (ICS) é a causa mais importante de IH em pacientes graves. O OBJETIVO deste estudo foi analisar a epidemiologia das infecções de corrente sangüínea de origem hospitalar em crianças internadas no Instituto da Criança do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo hospital de ensino e de assistência terciária, no período de janeiro de 1996 a agosto de 2003. MÉTODOS: O estudo foi feito no modelo de coorte, com análise retrospectiva de dados referentes às infecções de corrente sangüínea, coletados por método de vigilância ativa, seguindo os métodos validados pelo NNIS (National Nosocomial Infections Surveillance System). RESULTADOS: As infecções hospitalares (n = 4233) foram uma causa significativa de morbidade intra-hospitalar no local e período estudados. O risco de um paciente desenvolver uma ou mais infecções hospitalares foi de 11,5 para 100 saídas. As ICS representaram mais de um terço das IH nos oito anos analisados, com uma densidade de incidência que variou de 20,4 na Oncologia, 7,7 na UTI Neonatal, 7,3 na Pediátrica até 1,9 por 1000 pacientes-dia na Cirurgia. Ocorreram com maior freqüência em crianças com idade 5 anos (70,0%), com cateter venoso central (66,7%), e com doenças de base graves (80,4%). Pelo menos um agente infeccioso foi isolado em 78,9% dos episódios de ICS, sendo 41,5% gram-positivos e 44,8% gram-negativos. O microrganismo mais freqüente foi o Staphylococcus coagulase negativo (22,7%). A resistência do S. aureus e dos Staphylococcus coagulase negativos à oxacilina atingiu 58,9 e 80,3%, respectivamente. As cepas dos principais gram-negativos isolados (Klebsiella spp, Enterobacter spp, Pseudomonas spp e E.coli) mostraram-se amplamente resistentes à ceftriaxona, ao aztreonam e, em cerca de 35 a 57%, aos aminoglicosídeos. As ICS foram a causa ou contribuinte para o óbito em 21,9% dos pacientes, mas durante o período do estudo houve um decréscimo significante na mortalidade dos pacientes com ICS. CONCLUSÕES: A ICS foi uma causa importante de morbidade e mortalidade em pacientes pediátricos, predominando em crianças jovens e com doenças de base graves. Os principais fatores associados a ICS incluíram o uso de cateter vascular central e doença de base grave. Patógenos gram-negativos predominaram em todos os anos. O diagnóstico e terapêutica precoce são essenciais para a prevenção da morbidade e mortalidade e com a caracterização das ICS de origem hospitalar, pode-se auxiliar o programa de prevenção destas infecções e suas repercussões / INTRODUCTION: The hospital infections (HI) are main causes of morbidity and mortality in critically ill children. There are only few studies in pediatric age groups, and most of them demonstrated that the bloodstream infection (BSI) is the most important cause of HI in critically ill children. The OBJECTIVE of this study was to analyze the epidemiology of the nosocomial bloodstream infections (BSI) in children admitted to Instituto da Criança of Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo a teaching hospital with complex care attendance, from January 1996 to August 2003. METHODS: The study was carried out in a cohort model, with retrospective data analyses regarding bloodstream infections collected through active surveillance, following the methods validated by the National Nosocomial Infection Surveillance System NNIS. RESULTS: The HI represented a significant cause of morbidity in patients admitted to the hospital within period and local studied. The patients risk for developing one or more HI was 11,5 per 100 exits. The BSI represented more than one third of the HI in the eight analyzed years, with a incidence density varying from 20,4 at Oncology, 7,7 on Neonatal Intensive Care Unit, 7,3 at Pediatric Intensive Care Unit up to 1.9 per 1000 patient-days at Surgery. It occurred more frequently on children on the age of < 5 years old (70,0%), with central vein catheter (66,7%), and critically ill (80,4%). At least one infection agent was isolated in 78,9% of the BSI episodes, 41,5% gram-positive and 44,8% gram-negative. The most frequent pathogen was coagulase-negative staphylococci (22,7%). The proportion of S. aureus and coagulase-negative staphylococci methicillin resistant reached 58,9 and 80,3%, respectively. The main isolated gram-negatives (Klebsiella spp, Enterobacter spp, Pseudomonas spp and E.coli) showed thoroughly resistance to ceftriaxone, to aztreonam and on 35 to 57% to aminoglycosides. The BSIs were the cause or contribution for death in 21,9% of the patients, but during the period of this study, there was a significant lowering on mortality rate of patients with BSI. CONCLUSIONS: The BSI was important cause of morbidity and mortality in pediatric patients, predominantly in young and critically ill children. The main factors associated to BSI included the use of central vascular catheter and severe disease. Gram-negative pathogens predominated in every one all the years. The diagnoses and precocious therapy are essential on the prevention of morbidity, mortality and the characterization of nosocomial BSI, and would help on prevention programs of these infections and its repercussions