INTRODUÇÃO: Instabilidade hemodinâmica é uma condição comum nos pacientes críticos e usualmente associada aos quadros de choque. O débito cardíaco (DC) é o principal determinante da oferta de oxigênio, e uma importante variável a ser monitorizada nestes pacientes. A diferença arterio-venosa central de dióxido de carbono (△pCO2), é um índice que mostrou-se correlacionar inversamente com o DC em diversos estudos. OBJETIVOS: O objetivo deste estudo é avaliar o valor prognóstico do △pCO2 (utilizando uma amostra venosa central) em uma série heterogênea de pacientes com instabilidade hemodinâmica. MÉTODOS: Estudo prospectivo observacional conduzido em uma UTI geral de 18 leitos do Complexo Hospitalar da Santa Casa de Porto Alegre. Variáveis hemodinâmicos, coleta de gasometria arterial e venosa central, e lactato, foram realizadas na admissão e após 6, 12, 18 e 24 horas. A gravidade da doença foi avaliada pelo escore APACHE II e a a severidade da disfunção orgânica pelo escore SOFA. RESULTADOS: 60 pacientes foram analisados. A mortalidade ao fim de 28 dias foi de 43% (26/60). Não se observou diferença entre os valores médios de △pCO2 entre sobreviventes e óbitos em 28 dias. O tempo de permanência na UTI foi maior entre os pacientes com um △pCO2 “alargado”. Os valores médios de saturação venosa central de oxigênio (SvcO2) foram maiores no subgrupo de pacientes com △pCO2 normal. CONCLUSÃO: Este estudo mostrou que o △pCO2, em pacientes com instabilidade hemodinâmica, está relacionado com desfecho secundários importantes como uma maior permanência em UTI. A explicação fisiológica para isso deve-se a uma maior redução no fluxo sanguíneo (como sugerido por uma SvcO2 mais baixa) causando uma oferta inadequada de oxigênio na fase aguda da doença, resultando em maior tempo de recuperação. / BACKGROUND: Hemodynamic instability is a common condition in critically ill patients and is usually associated with shock. Cardiac output (CO) is the major determinant of oxygen supply and an important variable to be monitored in these patients. The venousarterial difference of partial pressure of carbon dioxide (△pCO2) ,is an index that have shown to be inversely correlate with the CO in several studies. OBJECTIVES: This study aims to evaluate the prognostic value of the △pCO2 (using a central venous sample) in a heterogeneous series of critically ill patients with hemodynamic instability. METHODS: A prospective observational clinical study was conducted in a 18-bed general ICU of the Santa Casa de Porto Alegre Hospital Complex. Hemodynamic measurements, collection of arterial and central venous blood gases, and lactate were performed on admission and after 6 , 12, 18 and 24 hours.The severity of patient’s disease was evaluated by the APACHE II score and the level of organic disfunction was evaluated by the SOFA score. RESULTS: A total of 60 patients were analyzed. The mortality rate at 28 days was 43% (26/60). There was no difference observed in mean △pCO2 between deaths and survivors after 28 days.The ICU stay was longer among patients with “enlarged” △pCO2.Mean central venous oxygen saturation values were higher in patients with normal △pCO2. CONCLUSION: This study showed that the △pCO2 in patients with hemodynamic instability is correlated with important secondary outcomes like higher ICU stay. The physiologic explanation is a reduction on blood flow (as suggested by lower ScvO2) causing a inadequate oxygen tissue delivery in the acute phase of the disease, resulting in a greater recovery time.

INTRODUÇÃO: Instabilidade hemodinâmica é uma condição comum nos pacientes críticos e usualmente associada aos quadros de choque. O débito cardíaco (DC) é o principal determinante da oferta de oxigênio, e uma importante variável a ser monitorizada nestes pacientes. A diferença arterio-venosa central de dióxido de carbono (△pCO2), é um índice que mostrou-se correlacionar inversamente com o DC em diversos estudos. OBJETIVOS: O objetivo deste estudo é avaliar o valor prognóstico do △pCO2 (utilizando uma amostra venosa central) em uma série heterogênea de pacientes com instabilidade hemodinâmica. MÉTODOS: Estudo prospectivo observacional conduzido em uma UTI geral de 18 leitos do Complexo Hospitalar da Santa Casa de Porto Alegre. Variáveis hemodinâmicos, coleta de gasometria arterial e venosa central, e lactato, foram realizadas na admissão e após 6, 12, 18 e 24 horas. A gravidade da doença foi avaliada pelo escore APACHE II e a a severidade da disfunção orgânica pelo escore SOFA. RESULTADOS: 60 pacientes foram analisados. A mortalidade ao fim de 28 dias foi de 43% (26/60). Não se observou diferença entre os valores médios de △pCO2 entre sobreviventes e óbitos em 28 dias. O tempo de permanência na UTI foi maior entre os pacientes com um △pCO2 “alargado”. Os valores médios de saturação venosa central de oxigênio (SvcO2) foram maiores no subgrupo de pacientes com △pCO2 normal. CONCLUSÃO: Este estudo mostrou que o △pCO2, em pacientes com instabilidade hemodinâmica, está relacionado com desfecho secundários importantes como uma maior permanência em UTI. A explicação fisiológica para isso deve-se a uma maior redução no fluxo sanguíneo (como sugerido por uma SvcO2 mais baixa) causando uma oferta inadequada de oxigênio na fase aguda da doença, resultando em maior tempo de recuperação. / BACKGROUND: Hemodynamic instability is a common condition in critically ill patients and is usually associated with shock. Cardiac output (CO) is the major determinant of oxygen supply and an important variable to be monitored in these patients. The venousarterial difference of partial pressure of carbon dioxide (△pCO2) ,is an index that have shown to be inversely correlate with the CO in several studies. OBJECTIVES: This study aims to evaluate the prognostic value of the △pCO2 (using a central venous sample) in a heterogeneous series of critically ill patients with hemodynamic instability. METHODS: A prospective observational clinical study was conducted in a 18-bed general ICU of the Santa Casa de Porto Alegre Hospital Complex. Hemodynamic measurements, collection of arterial and central venous blood gases, and lactate were performed on admission and after 6 , 12, 18 and 24 hours.The severity of patient’s disease was evaluated by the APACHE II score and the level of organic disfunction was evaluated by the SOFA score. RESULTS: A total of 60 patients were analyzed. The mortality rate at 28 days was 43% (26/60). There was no difference observed in mean △pCO2 between deaths and survivors after 28 days.The ICU stay was longer among patients with “enlarged” △pCO2.Mean central venous oxygen saturation values were higher in patients with normal △pCO2. CONCLUSION: This study showed that the △pCO2 in patients with hemodynamic instability is correlated with important secondary outcomes like higher ICU stay. The physiologic explanation is a reduction on blood flow (as suggested by lower ScvO2) causing a inadequate oxygen tissue delivery in the acute phase of the disease, resulting in a greater recovery time.

INTRODUÇÃO: Instabilidade hemodinâmica é uma condição comum nos pacientes críticos e usualmente associada aos quadros de choque. O débito cardíaco (DC) é o principal determinante da oferta de oxigênio, e uma importante variável a ser monitorizada nestes pacientes. A diferença arterio-venosa central de dióxido de carbono (△pCO2), é um índice que mostrou-se correlacionar inversamente com o DC em diversos estudos. OBJETIVOS: O objetivo deste estudo é avaliar o valor prognóstico do △pCO2 (utilizando uma amostra venosa central) em uma série heterogênea de pacientes com instabilidade hemodinâmica. MÉTODOS: Estudo prospectivo observacional conduzido em uma UTI geral de 18 leitos do Complexo Hospitalar da Santa Casa de Porto Alegre. Variáveis hemodinâmicos, coleta de gasometria arterial e venosa central, e lactato, foram realizadas na admissão e após 6, 12, 18 e 24 horas. A gravidade da doença foi avaliada pelo escore APACHE II e a a severidade da disfunção orgânica pelo escore SOFA. RESULTADOS: 60 pacientes foram analisados. A mortalidade ao fim de 28 dias foi de 43% (26/60). Não se observou diferença entre os valores médios de △pCO2 entre sobreviventes e óbitos em 28 dias. O tempo de permanência na UTI foi maior entre os pacientes com um △pCO2 “alargado”. Os valores médios de saturação venosa central de oxigênio (SvcO2) foram maiores no subgrupo de pacientes com △pCO2 normal. CONCLUSÃO: Este estudo mostrou que o △pCO2, em pacientes com instabilidade hemodinâmica, está relacionado com desfecho secundários importantes como uma maior permanência em UTI. A explicação fisiológica para isso deve-se a uma maior redução no fluxo sanguíneo (como sugerido por uma SvcO2 mais baixa) causando uma oferta inadequada de oxigênio na fase aguda da doença, resultando em maior tempo de recuperação. / BACKGROUND: Hemodynamic instability is a common condition in critically ill patients and is usually associated with shock. Cardiac output (CO) is the major determinant of oxygen supply and an important variable to be monitored in these patients. The venousarterial difference of partial pressure of carbon dioxide (△pCO2) ,is an index that have shown to be inversely correlate with the CO in several studies. OBJECTIVES: This study aims to evaluate the prognostic value of the △pCO2 (using a central venous sample) in a heterogeneous series of critically ill patients with hemodynamic instability. METHODS: A prospective observational clinical study was conducted in a 18-bed general ICU of the Santa Casa de Porto Alegre Hospital Complex. Hemodynamic measurements, collection of arterial and central venous blood gases, and lactate were performed on admission and after 6 , 12, 18 and 24 hours.The severity of patient’s disease was evaluated by the APACHE II score and the level of organic disfunction was evaluated by the SOFA score. RESULTS: A total of 60 patients were analyzed. The mortality rate at 28 days was 43% (26/60). There was no difference observed in mean △pCO2 between deaths and survivors after 28 days.The ICU stay was longer among patients with “enlarged” △pCO2.Mean central venous oxygen saturation values were higher in patients with normal △pCO2. CONCLUSION: This study showed that the △pCO2 in patients with hemodynamic instability is correlated with important secondary outcomes like higher ICU stay. The physiologic explanation is a reduction on blood flow (as suggested by lower ScvO2) causing a inadequate oxygen tissue delivery in the acute phase of the disease, resulting in a greater recovery time.

Contexte: la survenue d’IRA chez les patients ayant subi un traumatisme est une problématique qui a été peu étudiée jusqu’à ce jour. La présence de cette atteinte rénale a été démontrée comme étant associée à un risque accru de morbidités et de mortalité chez les sujets atteints. Objectifs: identifier les facteurs prédictifs d’insuffisance rénale ou plus récemment appelée atteinte rénale dans cette population particulière et tenter de trouver des facteurs qui peuvent être mesurés dans les premières heures de la prise en charge du patient. Aussi, nous avons cherché à savoir si l’injection de produit de contraste est associée à un risque accru d’insuffisance rénale aiguë dans cette population. Méthodes et résultats: la recherche a eu lieu à l’Hôpital du Sacré-Coeur de Montréal, un centre de traumatologie tertiaire en milieu urbain. Nous avons utilisé le registre des patients hospitalisés en traumatologie dans notre centre hospitalier entre 2002 et mars 2007 de même que les banques de données de laboratoire et de radiologie pour obtenir les données sur la créatinine et les examens avec produits de contraste. Finalement, une revue de dossiers structurée fut conduite pour recueillir le reste de l’information requise. L’incidence d’IRA dans la population étudiée est estimée à environ 5 %. Une analyse cas témoins fut conduite pour identifier les facteurs prédictifs d’IRA. Quarante-neuf cas d’IRA diagnostiqués par le médecin traitant et 101 témoins sélectionnés au hasard ont été analysés. Les facteurs prédictifs suivants ont été identifiés à l’analyse univariée : la première valeur de créatinine obtenue (p<0,001), l’instabilité hémodynamique (p<0,001), les antécédents d’insuffisance rénale chronique tels que notés dans le dossier par le médecin traitant (p=0,009), une maladie cardiaque (p=0,007), une chirurgie dans les 48 premières heures suivant le traumatisme (p=0,053), le niveau de gravité du traumatisme (Injury Severity Score) (p=0,046) et l’injection de produit de contraste au cours des 48 heures suivant le trauma (p=0,077). Parmi ces facteurs, deux ont été identifiés comme prédicteurs indépendants d’IRA à l’analyse multivariée. Une des valeurs était la première valeur de créatinine obtenue RC = 6,17 (p<0,001, IC95 % 2,81 – 13,53) pour chaque augmentation de 0.5mg/dL de créatinine. L’autre facteur était la présence d’instabilité hémodynamique RC 11,61 (p<0,001, IC95 % 3,71 – 36,29). Conclusion: des informations obtenues tôt dans la prise en charge du patient permettent de prédire le risque d’IRA chez ces patients. L’administration de contraste (intraveineuse ou intra-artérielle) ne s’est pas avérée un facteur indépendant de prédiction d’insuffisance rénale aiguë dans cette population dans le modèle multivarié. / Background: acute kidney injury (AKI) has important mortality and morbidity complications. Few studies have looked at predictors of acute renal failure in a trauma patient population. Objectives: we sought to identify factors associated with AKI that can be assessed in the early hospital stay of trauma patients. We also specifically assessed if the administration of radiological contrast was a predictor of AKI. Methods: we conducted a nested case-control study from the trauma registry of an urban Level I trauma center which includes data on more than 6 000 subjects. The cases consisted of 49 patients with a diagnosis of AKI by their treating physician in the first 7 days following their trauma between 2002 and 2007 (March 2007). The controls were randomly selected for a 1:2 case to control ratio. Data were retrieved from the prospective trauma registry database. Additional data were also obtained via the hospital laboratory and radiology databases. Finally, a structured chart review was conducted to obtain the remaining information. Univariate analyses were conducted. Elements with a significance level of <0.1 were included in a multivariate logistic regression model. Results: predictors identified in the univariate analysis were: the first creatinine value obtained (p<0,001), hemodynamic instability (p<0,001), history of coronary artery disease (p=0,007), history of chronic renal insufficiency as per physician’s diagnosis in the chart (p=0,009), surgery in the 48 hours following the trauma (p=0,053), and, injection of contrast in the 48 hours following the trauma (p=0,077). In the final multivariate model, two factors were statistically significant. One factor was the first creatinine value p<0,001, OR 6,17 CI95 % (2,81 – 13,53) for each increase of creatinine by 0,5mg/dL. The other factor was the presence of hemodynamic instability p<0,001 OR 11,61 CI95 % (3,71 – 36,29). Conclusion: easily obtained information in the emergency department can aid in predicting the risk of AKI in a trauma population. Early administration of radiological contrast was not an independant predictor of AKI in this population.

Contexte: la survenue d’IRA chez les patients ayant subi un traumatisme est une problématique qui a été peu étudiée jusqu’à ce jour. La présence de cette atteinte rénale a été démontrée comme étant associée à un risque accru de morbidités et de mortalité chez les sujets atteints. Objectifs: identifier les facteurs prédictifs d’insuffisance rénale ou plus récemment appelée atteinte rénale dans cette population particulière et tenter de trouver des facteurs qui peuvent être mesurés dans les premières heures de la prise en charge du patient. Aussi, nous avons cherché à savoir si l’injection de produit de contraste est associée à un risque accru d’insuffisance rénale aiguë dans cette population. Méthodes et résultats: la recherche a eu lieu à l’Hôpital du Sacré-Coeur de Montréal, un centre de traumatologie tertiaire en milieu urbain. Nous avons utilisé le registre des patients hospitalisés en traumatologie dans notre centre hospitalier entre 2002 et mars 2007 de même que les banques de données de laboratoire et de radiologie pour obtenir les données sur la créatinine et les examens avec produits de contraste. Finalement, une revue de dossiers structurée fut conduite pour recueillir le reste de l’information requise. L’incidence d’IRA dans la population étudiée est estimée à environ 5 %. Une analyse cas témoins fut conduite pour identifier les facteurs prédictifs d’IRA. Quarante-neuf cas d’IRA diagnostiqués par le médecin traitant et 101 témoins sélectionnés au hasard ont été analysés. Les facteurs prédictifs suivants ont été identifiés à l’analyse univariée : la première valeur de créatinine obtenue (p<0,001), l’instabilité hémodynamique (p<0,001), les antécédents d’insuffisance rénale chronique tels que notés dans le dossier par le médecin traitant (p=0,009), une maladie cardiaque (p=0,007), une chirurgie dans les 48 premières heures suivant le traumatisme (p=0,053), le niveau de gravité du traumatisme (Injury Severity Score) (p=0,046) et l’injection de produit de contraste au cours des 48 heures suivant le trauma (p=0,077). Parmi ces facteurs, deux ont été identifiés comme prédicteurs indépendants d’IRA à l’analyse multivariée. Une des valeurs était la première valeur de créatinine obtenue RC = 6,17 (p<0,001, IC95 % 2,81 – 13,53) pour chaque augmentation de 0.5mg/dL de créatinine. L’autre facteur était la présence d’instabilité hémodynamique RC 11,61 (p<0,001, IC95 % 3,71 – 36,29). Conclusion: des informations obtenues tôt dans la prise en charge du patient permettent de prédire le risque d’IRA chez ces patients. L’administration de contraste (intraveineuse ou intra-artérielle) ne s’est pas avérée un facteur indépendant de prédiction d’insuffisance rénale aiguë dans cette population dans le modèle multivarié. / Background: acute kidney injury (AKI) has important mortality and morbidity complications. Few studies have looked at predictors of acute renal failure in a trauma patient population. Objectives: we sought to identify factors associated with AKI that can be assessed in the early hospital stay of trauma patients. We also specifically assessed if the administration of radiological contrast was a predictor of AKI. Methods: we conducted a nested case-control study from the trauma registry of an urban Level I trauma center which includes data on more than 6 000 subjects. The cases consisted of 49 patients with a diagnosis of AKI by their treating physician in the first 7 days following their trauma between 2002 and 2007 (March 2007). The controls were randomly selected for a 1:2 case to control ratio. Data were retrieved from the prospective trauma registry database. Additional data were also obtained via the hospital laboratory and radiology databases. Finally, a structured chart review was conducted to obtain the remaining information. Univariate analyses were conducted. Elements with a significance level of <0.1 were included in a multivariate logistic regression model. Results: predictors identified in the univariate analysis were: the first creatinine value obtained (p<0,001), hemodynamic instability (p<0,001), history of coronary artery disease (p=0,007), history of chronic renal insufficiency as per physician’s diagnosis in the chart (p=0,009), surgery in the 48 hours following the trauma (p=0,053), and, injection of contrast in the 48 hours following the trauma (p=0,077). In the final multivariate model, two factors were statistically significant. One factor was the first creatinine value p<0,001, OR 6,17 CI95 % (2,81 – 13,53) for each increase of creatinine by 0,5mg/dL. The other factor was the presence of hemodynamic instability p<0,001 OR 11,61 CI95 % (3,71 – 36,29). Conclusion: easily obtained information in the emergency department can aid in predicting the risk of AKI in a trauma population. Early administration of radiological contrast was not an independant predictor of AKI in this population.

Chaque année, environ 1 à 1,25 million d’individus subiront une chirurgie cardiaque. [1] Environ 36 000 chirurgies cardiaques sont effectuées au Canada et 8000 procédures au Québec (http://www.ccs.ca). Le vieillissement de la population aura pour conséquence que la chirurgie cardiaque sera offerte à des patients de plus en plus à risque de complications, principalement en raison d’une co-morbidité plus importante, d’un risque de maladie coronarienne plus élevée, [2] d’une réserve physiologique réduite et par conséquent un risque plus élevé de mortalité à la suite d’une chirurgie cardiaque. L’une des complications significatives à la suite d’une chirurgie cardiaque est le sevrage difficile de la circulation extracorporelle. Ce dernier inclut la période au début du sevrage de la circulation extracorporelle et s’étend jusqu’au départ du patient de la salle d’opération. Lorsque le sevrage de la circulation extracorporelle est associé à une défaillance ventriculaire droite, la mortalité sera de 44 % à 86 %. [3-7] Par conséquent le diagnostic, l’identification des facteurs de risque, la compréhension du mécanisme, la prévention et le traitement du sevrage difficile de la circulation extracorporelle seront d’une importance majeure dans la sélection et la prise en charge des patients devant subir une chirurgie cardiaque. Les hypothèses de cette thèse sont les suivantes : 1) le sevrage difficile de la circulation extracorporelle est un facteur indépendant de mortalité et de morbidité, 2) le mécanisme du sevrage difficile de la circulation extracorporelle peut être approché d’une façon systématique, 3) la milrinone administrée par inhalation représente une alternative préventive et thérapeutique chez le patient à risque d’un sevrage difficile de la circulation extracorporelle après la chirurgie cardiaque. / Every year, 1 million to 1.25 million patients worldwide undergo cardiac surgery. [1] Up to 36,000 cardiac surgeries are performed each year in Canada and close to 8000 in Quebec (http://www.ccs.ca). Because of the aging of the population, cardiac surgery will increasingly be offered to patients at a higher risk of complications. Indeed, elderly patients have increased co-morbidities, and aging is also a significant risk factor in the prevalence of coronary artery disease. [2] The consequence is a reduced physiologic reserve, hence an increased risk of mortality. These issues will have a significant impact on future healthcare costs, because our population undergoing cardiac surgery will be older and more likely to develop postoperative complications. One of the most dreaded complications in cardiac surgery is difficult separation from cardiopulmonary bypass (CPB). The definition of difficult separation from CPB includes the time period from when CPB is initiated and until the patient leaves the operating room. When separation from CPB is associated with right ventricular failure, the mortality rate will range from 44% to 86%. [3-7] Therefore the diagnosis, the preoperative prediction, the mechanism, prevention and treatment of difficult separation from CPB will be crucial in order to improve the selection and care of patients and to prevent complications for this high-risk patient population. The hypotheses of this thesis are the following: 1) difficult separation from CPB is an independent factor of morbidity and mortality, 2) the mechanism of difficult separation from CPB can be understood through a systematic approach, 3) inhaled milrinone is a preventive and therapeutic approach in the patient at risk for difficult weaning from CPB after cardiac surgery.

Chaque année, environ 1 à 1,25 million d’individus subiront une chirurgie cardiaque. [1] Environ 36 000 chirurgies cardiaques sont effectuées au Canada et 8000 procédures au Québec (http://www.ccs.ca). Le vieillissement de la population aura pour conséquence que la chirurgie cardiaque sera offerte à des patients de plus en plus à risque de complications, principalement en raison d’une co-morbidité plus importante, d’un risque de maladie coronarienne plus élevée, [2] d’une réserve physiologique réduite et par conséquent un risque plus élevé de mortalité à la suite d’une chirurgie cardiaque. L’une des complications significatives à la suite d’une chirurgie cardiaque est le sevrage difficile de la circulation extracorporelle. Ce dernier inclut la période au début du sevrage de la circulation extracorporelle et s’étend jusqu’au départ du patient de la salle d’opération. Lorsque le sevrage de la circulation extracorporelle est associé à une défaillance ventriculaire droite, la mortalité sera de 44 % à 86 %. [3-7] Par conséquent le diagnostic, l’identification des facteurs de risque, la compréhension du mécanisme, la prévention et le traitement du sevrage difficile de la circulation extracorporelle seront d’une importance majeure dans la sélection et la prise en charge des patients devant subir une chirurgie cardiaque. Les hypothèses de cette thèse sont les suivantes : 1) le sevrage difficile de la circulation extracorporelle est un facteur indépendant de mortalité et de morbidité, 2) le mécanisme du sevrage difficile de la circulation extracorporelle peut être approché d’une façon systématique, 3) la milrinone administrée par inhalation représente une alternative préventive et thérapeutique chez le patient à risque d’un sevrage difficile de la circulation extracorporelle après la chirurgie cardiaque. / Every year, 1 million to 1.25 million patients worldwide undergo cardiac surgery. [1] Up to 36,000 cardiac surgeries are performed each year in Canada and close to 8000 in Quebec (http://www.ccs.ca). Because of the aging of the population, cardiac surgery will increasingly be offered to patients at a higher risk of complications. Indeed, elderly patients have increased co-morbidities, and aging is also a significant risk factor in the prevalence of coronary artery disease. [2] The consequence is a reduced physiologic reserve, hence an increased risk of mortality. These issues will have a significant impact on future healthcare costs, because our population undergoing cardiac surgery will be older and more likely to develop postoperative complications. One of the most dreaded complications in cardiac surgery is difficult separation from cardiopulmonary bypass (CPB). The definition of difficult separation from CPB includes the time period from when CPB is initiated and until the patient leaves the operating room. When separation from CPB is associated with right ventricular failure, the mortality rate will range from 44% to 86%. [3-7] Therefore the diagnosis, the preoperative prediction, the mechanism, prevention and treatment of difficult separation from CPB will be crucial in order to improve the selection and care of patients and to prevent complications for this high-risk patient population. The hypotheses of this thesis are the following: 1) difficult separation from CPB is an independent factor of morbidity and mortality, 2) the mechanism of difficult separation from CPB can be understood through a systematic approach, 3) inhaled milrinone is a preventive and therapeutic approach in the patient at risk for difficult weaning from CPB after cardiac surgery.