Thesis (Ph.D. (Demography))--Mahidol University, 2005.

Diss. (sammanfattning) Uppsala : Sveriges lantbruksuniversitet, 2010. / Härtill 5 uppsatser.

Thesis (MNutr (Interdisciplinary Health Sciences. Human Nutrition))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: Introduction : The health status of women in peri-urban areas has been influence by the South African political transition. Despite some progress, maternal and child mortality rates are still unacceptably high. A mother’s nutritional status is one of the most important determinants of maternal and birth outcomes. The Institute of Medicine’s pre-pregnancy Body Mass Index (BMI) method is not always appropriate to use in a peri-urban setting as many women attend their first antenatal clinic later on in their pregnancy. Two alternative methods, the gestational BMI (GBMI) and the gestational risk score (GRS), have been used elsewhere to screen for at risk pregnancies, but have not been used in a South African peri-urban setting. Furthermore, examining socio-economic variables (SEV) aids in the explanation of the impact of social structures on an individual. Risk factors can then be established and pregnant women in these higher risk groups can be identified and given additional antenatal clinic appointments and priority during labour. Aim: The first aim was to investigate the strength of the GBMI and GRS methods for predicting birth outcomes and maternal morbidities. The second aim was to investigate the relationships between SEV, GBMI and maternal morbidities. Methods: This was a sub-study of the Philani Mentor Mothers Study. A sample of 103 and 205 were selected for investigating the prediction methods and SEV respectively. Maternal anthropometry, gestational weeks and SEV were obtained during interviews before birth. Information obtained was used to calculate GBMI and GRS and to assess the SEV. Birth outcomes were obtained from the infant’s clinic cards and maternal morbidities were obtained from interviews two days after the birth. Results No significant association was found between GBMI and birth outcomes and maternal morbidities. A significant positive association was found between GRS and birth head circumference percentile (r=0.22, p<0.05). The higher the GRS, the higher the risk of an infant spending longer time in the hospital (Kruskal Wallis X2 = 4, p<0.05). A significant positive association was found between GBMI and the following SEV factors; age (r=0.33, p<0.05), height (r=0.15, p<0.05), parity (r=0.23, p<0.05), income (r=0.2, p<0.05), marital status (X2 = 9.35, p<0.05), employment (U=2.9, p<0.05) and HIV status (U=2.54, p<0.05). No statistically significant relationships were found between gestational hypertension and gestational diabetes mellitus and SEV. Conclusion: From the findings of this sub-study there were some promising results, however it is still unclear as to which method is the most appropriate to predict adverse birth outcomes and maternal morbidity. It is recommended that the GBMI and GRS once-off methods be repeated in a larger population to see if there are more parameters that could be predicted. Women who were older, shorter, married, had more pregnancies, HIV negative and had a higher socioeconomic status tended to have a greater GBMI. This can lead to adverse birth outcomes and increases the risk of women developing maternal morbidities and other chronic diseases later in their life. Optimal nutrition and health promotion strategies targeting women before conception should be implemented. / AFRIKAANSE OPSOMMING: Inleiding: Die gesondheidstatus van vroue in semi-stedelike areas is beïnvloed deur die Suid-Afrikaanse politiese oorgang. Ten spyte van ’n mate van vooruitgang is die sterftesyfers vir moeders en kinders steeds onaanvaarbaar hoog. ‘n Moeder se voedingstatus is een van die mees belangrike bepalende faktore van moeder- en geboorteuitkomste. Die Instituut van Geneeskunde se voorswangerskap Liggaamsmassa Indeks (LMI) metode is nie altyd toepaslik om te gebruik in ‘n semi-stedelike opset nie aangesien baie vroue hul eerste voorgeboorte-kliniek eers later in hul swangerskap bywoon. Twee alternatiewe metodes, die swangerskap LMI (SLMI) en die swangerskap risiko telling (SRT) is al elders gebruik as sifting vir hoë risiko swangerskappe, maar is nog nie gebruik in ‘n Suid-Afrikaanse semi-stedelike opset nie. Vervolgens kan ‘n ondersoek na sosio-ekonomiese veranderlikes (SEV) help om die impak van maatskaplike strukture op ‘n individu te verduidelik. Risiko faktore kan dan vasgestel word en swanger vroue wat in hierdie hoër risiko groepe val kan geïdentifiseer word. Dié vroue kan addisionele voorgeboorte-kliniek afsprake ontvang asook voorkeurbehandeling tydens die geboorteproses. Doelstellings: Die eerste doelstelling was om die sterkte van die SLMI en SRT metodes te ondersoek as voorspellers van geboorte uitkomste en moeder-morbiditeite. Die tweede doelstelling was om die verhoudings tussen SEV, SLMI en moeder-morbiditeite te ondersoek. Metodes: Hierdie projek was ‘n sub-studie van die Philani Mentor Moeders Studie. ‘n Steekproefgrootte van 103 en 205 was geselekteer om onderskeidelik die voorspeller metodes en SEV te ondersoek. Die moeder se antropometrie, swangerskap weke en SEV was verkry gedurende onderhoude voor geboorte. Informasie ingewin was gebruik om die SLMI en SRT te bereken en om die SEV te ondersoek. Geboorteuitkomste was verkry vanaf die babas se kliniekkaarte en moeder-morbiditeite was verkry tydens onderhoude twee dae na die geboorte. Resultate: Geen betekenisvolle assosiasie was gevind tussen SLMI, geboorteuitkomste en moeder-morbiditeite nie. ‘n Betekenisvolle positiewe assosiasie was gevind tussen SRT en die geboorte kopomtrek persentiel (r=0.22, p<0.05). Hoe hoër die SRT, hoe hoër die risiko dat ‘n baba langer in die hospitaal sou bly (Kruskal Wallis X2=4, p<0.05). ‘n Betekenisvolle positiewe assosiasie was gevind tussen SLMI en die volgende SEV faktore: ouderdom (r=0.33, p<0.05), lengte (r=0.15, p<0.05), pariteit (r=0.23, p<0.05), inkomste (r=0.2, p<0.05), huwelikstatus (X2=9.35, p<0.05), besit van ‘n identiteitsdokument (U=1.75, p<0.05), werkstatus (U=2.9, p<0.05) en MIV status (U=2.54, p<0.05). Geen statisties beduidende verhoudings was gevind tussen swangerskap hipertensie, swangerskap diabetes mellitus en SEV nie. Gevolgtrekking Sommige bevindinge van hierdie sub-studie dui op belowende resultate, alhoewel dit steeds nie duidelik is watter metode die mees toepaslike is om ongewenste geboorteuitkomste en moeder-morbiditeit te voorspel nie. Dit word aanbeveel dat die SLMI en SRT eenmalige metodes herhaal word in ‘n groter populasie om te sien of daar meer parameters is wat voorspel kan word. Vroue wat ouer, korter, getroud, meer swangerskappe, MIV negatief en ‘n hoër sosio-ekonomiese status gehad het was geneig om ‘n hoër SLMI te hê. Dit kan lei tot ongewenste geboorteuitkomste en verhoogde risiko om moeder-morbiditeite en ander chroniese siektes later in hul lewe te ontwikkel. Optimale voeding en gesondheidsbevordering strategieë wat vroue teiken voor bevrugting behoort geïmplementeer te word.

RESUMO Pouco se conhece sobre o perfil epidemiológico da população indígena infantil, principalmente na Amazônia Ocidental. O conhecimento das morbidades e da demanda aos serviços auxilia na estruturação dos sistemas de saúde. A Casa de Saúde Indígena (CASAI) funciona como unidade de apoio, recebendo os indígenas referenciados para tratamento de saúde na rede do Sistema Único de Saúde. A CASAI de Rio Branco atende aos indígenas dos Distritos Sanitários Especiais Indígenas (Dsei) do Alto Rio Purus e do Alto Rio Juruá. Com o objetivo de traçar o perfil epidemiológico das crianças indígenas referenciadas no Estado do Acre e adjacências, foram avaliados todos os prontuários das crianças menores de 10 anos de idade que passaram pela CASAI de Rio Branco entre janeiro de 2003 e dezembro de 2007. As causas mais freqüentes de internação na CASAI foram as doenças infecciosas e parasitárias (cap. I do CID 10) com 19por cento , seguidas das doenças do aparelho respiratório (cap. X do CID 10) com 16,5por cento . A malária foi a quinta causa mais encontrada. Em 23por cento dos casos não foi encontrado registro sobre a causa da internação. Houve aumento na participação das malformações congênitas (cap. XVII do CID 10) 8 nos anos finais do estudo, sugerindo um processo de transição epidemiológica. Observou-se predominância de crianças menores de dois anos de idade (44por cento ), principalmente no primeiro ano de vida (26,5por cento ), com poucos recém nascidos (3,9por cento ). Quase a totalidade das crianças atendidas (95por cento ) era moradora de aldeias. As diferenças culturais entre as etnias e a dificuldade de acesso às aldeias foram os principais fatores determinantes da demanda e do tipo de morbidade que chega a CASAI / There is little information available about the epidemiological profile of the indigenous children population, mainly in the Western Amazon Region. Knowledge about morbidity and demand to health services assists in the structuring of health systems. The Indigenous Nursing House (CASAI) works as a support unit, receiving the indigenous people referred for health treatment in the network of the Brazilian Public Health System (SUS). The CASAI of Rio Branco serves the indigenous population from two Special Indigenous Sanitary Districts (DSEI): Alto Rio Purus and Alto Rio Jurua. Aiming to outline the epidemiological profile of referred indigenous children in Acre and nearby areas, we evaluated medical records of all children under 10 years of age who have passed the CASAI of Rio Branco from January 2003 to December 2007. The most frequent causes of hospitalization in the CASAI were infectious and parasitic diseases (Chapter I of International Classification of Diseases - 10º edition) with 19per cent, followed by respiratory diseases (chapter X of the ICD-10) with 16.5per cent. Malaria was the fifth most frequent cause. In 23per cent of cases the 10 cause of hospitalization was not found on records. There was an increased participation of congenital malformations (Chapter XVII of ICD-10) in the final years of the study, suggesting an epidemiological transition process. A high prevalence of children under two years of age (44per cent) was observed, mainly in the first year of life (26.5per cent), with few newborns (3.9per cent). Almost all the children enrolled (95per cent) were resident in villages in the forest. Cultural differences between ethnic and difficult access to villages were the main determinants of demand and the type of morbidity that comes to CASAI

abstract: I present the results of studies from two historically separate fields of research: heat related illness and human thermal comfort adaptation. My research objectives were: (a) to analyze the relationships between climate and heat related morbidity in Phoenix, Arizona and Chicago, Illinois; (b) explore possible linkages of human thermal comfort adaptation to heat-related illness; and (c) show possible benefits of collaboration between the two fields of research. Previous climate and mortality studies discovered regional patterns in summertime mortality in North America: lower in hot, southern cities compared to more temperate cities. I examined heat related emergency (911) dispatches from these two geographically and climatically different cities. I analyzed with local weather conditions with 911 dispatches identified by responders as "heat" related from 2001 to 2006 in Phoenix and 2003 through 2006 in Chicago. Both cities experienced a rapid rise in heat-related dispatches with increasing temperature and heat index, but at higher thresholds in Phoenix. Overall, Phoenix had almost two and half times more heat-related dispatches than Chicago. However, Phoenix did not experience the large spikes of heat-related dispatches that occurred in Chicago. These findings suggest a resilience to heat-related illness that may be linked to acclimatization in Phoenix. I also present results from a survey based outdoor human thermal comfort field study in Phoenix to assess levels of local acclimatization. Previous research in outdoor human thermal comfort in hot humid and temperate climates used similar survey-based methodologies and found higher levels of thermal comfort (adaptation to heat) that in warmer climates than in cooler climates. The study presented in this dissertation found outdoor thermal comfort thresholds and heat tolerance levels in Phoenix were higher than previous studies from temperate climates more similar to Chicago. These differences were then compared to the differences in weather conditions associated with heat-related dispatches. The higher comfort thresholds in Phoenix were similar in scale to the climate differences associated with the upsurge in heat-related dispatches in Phoenix and Chicago. This suggests a link between heat related illness and acclimatization, and illustrates potential for collaboration in research between the two fields. / Dissertation/Thesis / Ph.D. Geological Sciences 2012

ix, 88 p. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / This dissertation addresses the high disease burden in developing countries today by examining the role of disease in economic development through its impact on productivity, fertility and human capital investment. In the second chapter of this dissertation, I model the impact on labor productivity of a change in disease susceptibility that results from intellectual property rights (IPR) reform. I develop a North-South model in which the disease environments differ between the rich and poor countries, and individuals consume innovated health goods to avoid the cost (labor time lost) of getting a disease. Southern welfare is shown to increase with the imposition of IPR protection when health needs in the South differ sufficiently from those in the North, and when health goods are accessible (in terms of adequate health care infrastructure) and effective (in counteracting disease). In the third chapter of this dissertation, I model the impact of child disease burden on fertility and human capital investment. The fertility response to a decline in child mortality depends on the morbidity effect of the disease, the level of disease burden, and whether prevalence rates or case fatalities decline. Fertility rates follow mortality and morbidity, but since mortality and morbidity do not always move in the same direction, the fertility response may be dampened or non-monotonic. Using a 20-year panel data set on malaria prevalence for 44 countries in sub-Saharan Africa, I find empirical support for the cases defined by the model; changes in malaria prevalence affect fertility more in non-endemic areas, where cases are more severe and more fatal relative to endemic areas. Historical and biological evidence suggest a link between (infectious) diseases early in life and (non-infectious) diseases later in life. In Chapter IV I model this link using a three-period overlapping generations model in which childhood disease outcomes affect longevity. Simulations in a general equilibrium framework duplicate the defining characteristics of the epidemiological-demographic transition as it occurred in many industrialized countries: as disease declines parents engage in a quantity-quality tradeoff for children, longevity rises and population declines after an initial jump. This dissertation includes unpublished co-authored material. / Committee in charge: Shankha Chakraborty, Chairperson, Economics; Bruce Blonigen, Member, Economics; Peter Lambert, Member, Economics; Laura Leete, Member, Planriing Public Policy & Mgmt; Jean Stockard, Outside Member, Planning Public Policy & Mgmt

The thesis contains basic information regarding Viral hepatitis E, particularly the causative agents, human and animal strains, the clinical picture and occurrence in the Czech Republic as well as abroad. The practical part is focused on analysis of trends of occurrence of the viral hepatitis E in the Czech Republic from the very earliest, up until the present time. The research further looks at morbidity in the different regions of the Czech Republic, which is compared to the morbidity in Plzen region. The practical part further focuses on the specific source, route of transmission and possible preventive measures against viral hepatitis E. In this work, both quantitative and qualitative types of research were used. In quantitative research, secondary analysis of data was used from various sources like EPIDAT. In the qualitative research, semi standardised interviews with doc. MUDr. Petr Pazdiora CSc, and with MVDR. Vaclav Polacek. Based on the interviews, an educational material in form of a leaflet was created. From the year 1996 to the year 2012, a total of 894 cases of HEV where recorded. In 1996, there was only one case of Viral hepatitis E which was recorded. From 1997 to 2005, the occurrence was sporadic and negligible. From 2006, there was a linear increase in viral hepatitis E. The incidence significantly increased in the year 2009 and 2011. In the Czech Republic, Viral Hepatitis E has an upward trend of occurrence. Viral hepatitis E cannot be prevented by active immunization, possibility of a vaccine is still being investigated. Basic preventive measures include maintaining principles of hygiene like avoiding drinking water from unknown sources, food should be prepared using drinking water and fruit should be properly washed and then peeled before consumption. It is important to find out about the occurrence of viral hepatitis E, before travelling to any foreign country. During meat processing, it is important to avoid cross contamination between clean and unclean operations. Before consuming foods like black pudding, sausages, steaks and pies, it is important to ensure that this food is properly cooked using adequate heat and for a period of time. From the research, it can be concluded that, there is need to devote adequate attention to viral hepatitis E. It may seem that the incidence in the Czech Republic is not serious, but it is important to focus on public awareness of this disease, especially the preventive measures so as to stop it from spreading.

Fundamentos: A insuficiência cardíaca tem uma grande importância como preditor de morbimortalidade em pacientes com síndrome coronariana aguda. Objetivo: Avaliar a ocorrência de insuficiência cardíaca e outros preditores de morbimortalidade na síndrome coronariana aguda em longo prazo. Métodos: Foi estudada uma coorte de 403 pacientes consecutivos e prospectivos, com queixas de dor torácica. Descreveram-se os dados demográficos e as características clínicas e laboratoriais. Comparou-se a estratificação de risco invasiva versus não invasiva, e as diferenças entre o tratamento medicamentoso com o intervencionista avaliando a evolução durante a internação e no período de até oito anos após a alta, em relação aos eventos cardiovasculares, não cardiovasculares e óbitos. Os dados numéricos serão apresentados em médias e desvios-padrão ou mediana e distância interquartílica, os dados categóricos através da porcentagem. Foram utilizados testes de t de Student, teste de Mann-Whitney, teste de qui-quadrado e teste exato de Fisher segundo sua indicação. Para a construção do modelo de sobrevida foram utilizados o teste de Kaplan-Meier e o teste de log-rank; o modelo multivariado foi ajustado utilizando o modelo de Cox. Após realizar a análise de sobrevida de Cox, para garantir o pressuposto do risco proporcional foi utilizado o modelo de Cox estratificado. Todas as análises foram realizadas utilizando o programa R versão 2.9.1. Resultados: População amostral constituída por 403 pacientes com queixas de dor torácica, sendo 65,8% com diagnóstico de SCA sem supra de ST, 27,8% SCA com supra de ST e 6,5% sem SCA. Da população amostral, foram avaliados 377 pacientes, sendo 37,93% do sexo feminino e a média de idade foi de 62,211,6 anos. A creatinina merece destaque como fator prognóstico, sendo o ponto de corte de 1,4mg/dL, com acurácia de 62,1%. Foram ainda observadas diferenças estatisticamente significativas quanto à idade na presença de insuficiência cardíaca; e quanto às terapias utilizadas antes e depois de 2002 em relação à mortalidade. Incluiu-se uma variável adicional no modelo multivariado, que se chamou de FC>PAS, para qualquer frequência cardíaca maior que a pressão arterial sistólica na admissão. Conclusões: A presença de IC na admissão, creatinina inicial >1,4mg/dL, idade e FC dos pacientes internados com SCA são preditores independentes de mortalidade. Observou-se que os pacientes com IC atendidos antes de 2002 apresentaram pior sobrevida do que os pacientes atendidos a partir de 2002 e que a mudança na terapia foi a responsável por isso. Mesmo com a diferença da sobrevida relacionada com a época da internação, o impacto dessas variáveis clínicas e laboratoriais na mortalidade foi semelhante. Outros estudos devem ser realizados a fim de avaliar se condutas diferenciadas para os pacientes com IC na admissão da internação de SCA podem reduzir esta mortalidade. / Background: Heart failure (HF) is extremely important as a morbidity and mortality predictor in patients with acute coronary syndrome. Objective: To assess the occurrence of heart failure and other morbidity and mortality predictors in acute coronary syndrome over the long term. Methods: A cohort of 403 prospective consecutive patients was studied, complaining of chest pain. The demographic data were described, together with the clinical and laboratory characteristics. Invasive versus non-invasive risk stratification was compared, as well as differences between surgical and percutaneous intervention and drug treatment, assessing progression during hospitalisation and for periods of up to eight years after discharge, in terms of cardiovascular and non-cardiovascular events and deaths. The numerical data will be presented as means and standard deviations or as medians and interquartíle distances, with the categorical data shown as percentages. The Students t, Mann-Whitney, chi-square and Fishers exact tests were used as indicated. The Kaplan-Meier and log-rank tests were used to construct the survival model; the multivariate model was adjusted through the Cox model. After conducting the Cox survival analysis in order to underpin the proportional risk assumption, the stratified Cox model was used. All the analyses were conducted through the R Programme, version 2.9.1. Results: In the sample population of 403 patients complaining of chest pain, 65.8% were diagnosed with non-ST elevation ACS, 27.8% with ST elevation ACS, and 6.5% without ACS. In the sample population, 377 patients were assessed, with 37.93% of them women and an average age of 62.211.6 years. Creatinine was particularly noteworthy as a prognostic factor, with a cut-off point of 1.4mg/dL and accuracy of 62.1%. Statistically significant differences were also observed for age with HF and the treatments used before and after 2002 in terms of mortality. An additional variable was included in the multivariate model called the HR>SBP, for any heart rate higher than systolic blood pressure on admission. Conclusion: The presence of HF on admission, with initial creatinine of >1.4mg/dL, age and HR of patients hospitalized with ACS are independent mortality predictors. It was noted that patients with HF treated prior to 2002 presented shorter survival times than HF patients treated after 2002, due to alterations in treatment. Even with the difference in survival times related to the period of hospitalization, the impact of the clinical variables remains significant regardless of the time of hospitalization. Other studies must be conducted in order to assess whether different types of conduct for patients with HF on admission for hospitalization with ACS might lower this mortality rate.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Population aging is a global reality. Initially observed in developed countries, today presents intensively in developing countries. Brings with it many concerns to managers, as is becoming a public health problem, especially when considering the external causes of morbidity and mortality, especially falls. This research aims to evaluate the morbidity and mortality due to falls in elderly Brazilians and analyze the temporal trend of hospital admissions for people aged sixty or older, residing in the State of Rio Grande do Sul (RS), from 1998 to 2012 and were drawn their projections relative and absolute, by the year 2020. This study consists of three articles. The former is a reflection article about the human aging process, the second, an integrative review of literature on the morbidity and mortality due to falls in elderly Brazilians and the third refers to the temporal trend and projection of hospital admissions by falls for elderly residents in the RS. This study did not require submission to the Research Ethics Committee (CEP) of the Universidade Federal de Santa Maria (UFSM) because it is a search using secondary data in the public domain, available online at DATASUS (www.datasus.gov.br) and thus there is no variables that allow the identification of the research subjects. / O envelhecimento populacional é uma realidade mundial. Inicialmente observado em países desenvolvidos, hoje se apresenta de forma intensa nos países em desenvolvimento. Trás consigo inúmeras preocupações aos gestores, uma vez que vem se tornando um problema de saúde pública, principalmente quando se analisam as causas externas de morbimortalidade, com destaque para as quedas Esta pesquisa buscou verificar a morbimortalidade por quedas em idosos brasileiros e analisar a tendência temporal de internações hospitalares para pessoas com sessenta anos ou mais de idade, residentes no Estado do Rio Grande do Sul (RS), no período de 1998 a 2012 e foram traçadas suas projeções relativas e absolutas, até o ano de 2020. O presente estudo é constituído de três artigos. O primeiro trata-se de um artigo de reflexão a cerca do processo de envelhecimento humano, o segundo, uma revisão integrativa da literatura a respeito da morbimortalidade por quedas em idosos brasileiros e o terceiro, refere-se à tendência temporal e projeção de internações hospitalares por quedas para idosos residentes no RS. O presente estudo não necessitou de submissão ao Comitê de Ética em Pesquisa (CEP) da Universidade Federal de Santa Maria (UFSM) por se tratar de uma pesquisa utilizando banco de dados secundários de domínio público, disponível online pelo DATASUS (www.datasus.gov.br) e, desta forma, não havendo variáveis que possibilitem a identificação dos sujeitos da pesquisa.

Les infections nosocomiales à bactéries multi-résistantes sont en constante augmentation en réanimation. Elles ont des conséquences individuelles et collectives majeures. La mortalité en réanimation et les prolongations des durées de séjour sont les deux principales conséquences individuelles connues à ce jour. Plusieurs facteurs confondants rendent l'interprétation des études difficiles, dont l'état sous jacent du patient, la virulence de la bactérie et l'adéquation thérapeutique. Mesurer la part de chacun de ces facteurs et préciser leur responsabilité respective est indispensable pour mobiliser les différents acteurs et améliorer le pronostic des patients en réanimation. Dans cette thèse nous avons souhaité approcher la réponse quant aux conséquences individuelles. A partir d'une base de données incluant des patients de réanimation, nous avons utilisé les méthodes statistiques les plus récentes et avons tenté de prendre en compte les différents facteurs confondants , pour répondre à trois questions précises que sont : la mortalité liée à une espèce bactérienne donnée, les facteurs associés à la mortalité des patients présentant un sepsis sévère ou choc septique en réanimation et les conséquences liées à l'isolement des patients infectés ou colonisés avec une bactérie multi-résistante. Nous montrons que (1) par l'intermédiaire d'une prolongation de la durée de séjour en réanimation, l'infection à Clostridium difficile augmente la pression de colonisation, sans pour autant avoir d'impact direct sur le décès. (2) que le pronostic des sepsis sévères et des chocs septiques dépend de l'adéquation de l'antibiothérapie et que les bactéries résistantes sont plus souvent traitées de manière inadéquate. (3) que l'isolement contact est associé non seulement à une augmentation attendue du risque de pneumonie nosocomiale a germe multi-résistants mais aussi à une augmentation du risque d'erreurs thérapeutiques et d'événements indésirables non infectieux. Cet impact délétère suggéré par des études en dehors de la réanimation doit être pris en compte lors de la mise en place des précautions contact en réanimation. / Nosocomial infections with multidrug-resistant bacteria are increasing in ICU. They have major individual and collective consequences. Mortality in the ICU and prolongation of length of stay are the two main individual consequences known to date. Several confounding factors make it difficult to interpret studies, including the patient's underlying condition, the virulence of bacteria and the adequacy of therapy. It is essential to measure the share of each of these factors and to clarify their respective responsibilities to mobilize the different actors and improve the prognosis of patients in intensive care. In this thesis, and drawing upon a database including ICU patients, we used the latest statistical methods and tried to take into account the various confounding factors to evaluate the individual consequences of multidrug-resistant bacteria in ICU. We sought to address three specific questions: mortality linked to specific bacterial species, factors associated with mortality in patients with severe sepsis or septic shock in intensive care unit, and the consequences of the isolation of patients infected or colonized with multidrug-resistant bacteria. We demonstrated that (1) because they are associated with a longer stay in intensive care unit, Clostridium difficile infections increase the pressure of colonization although they have no direct consequence on mortality; (2) the prognosis of sever sepsis or septic shock depends on the adequacy of the antibiotic therapy and that resistant bacteria are often inadequately treated; and (3) that isolation is not only associated with an expected increase in the risk of nosocomial pneumonia with multi-drug resistant pathogens strains but also with an increase in non-infectious adverse events.