Matt Harmon

18 Chapter one Respiratory targets such as PaCO 2 can influence the risk of secondary brain-injury post cardiac arrest. 57 Both hypercapnia and hypocapnia have been associated with adverse neurologic outcome. 58 Moreover, ventilator settings may affect lung-brain interactions and impact neurologic outcome in cardiac arrest. 59 Therefore, it is important to examine mechanical ventilation practices and their association with outcome in patients with cardiac arrest. This can aid in establishing respiratory targets in order to optimize ventilatory settings post-cardiac arrest aimed at mitigating hypoxic-ischemic brain injury. Patients with hypoxic-ischemic brain injury are also at risk for nosocomial infections resulting in increased duration of mechanical ventilation and increased ICU stay. 60 There are several interesting aspects relating to the relationship between TTM and these nosocomial infections. For one, cooling to hypothermia can increase the risk for nosocomial infections. 61 Previous TTM research has even suggested that the microbiological make up of these infections could depend on the chosen target temperature, with relatively more gram-negative organisms in patients treated with hypothermia compared to patients not treated with TTM. 60 Also, the difficulties in diagnosing infection during TTM may result in delay in initiation of antibiotic treatment with subsequent increased duration of ICU- and hospital-stay. 62 This may prompt the question whether the use of prophylactic antibiotics may reduce infectious complications in cardiac arrest patients. Taken together, body temperature can have a profound impact on clinical outcome but our understanding of spontaneous and induced temperature alterations in the ICU is incomplete and warrants further exploration. Induced normothermia is a cheap treatment modality that is easily applicable in ICUs and could potentially benefit septic patients, but needs to be better understood before clinical application in patients. Part I of this thesis will examine patient characteristics and treatment of patients with hypothermic sepsis, as well as explore mechanisms of the hypothermic response in sepsis. Part II of this thesis will focus on induced normothermia in sepsis. Using a model of human endotoxemia, which mimics the hyperinflammatory state in sepsis at the physiological and biological level 63 , we studied the effects of induced normothermia on physiology, inflammation and coagulation. Part III of this thesis will explore the effects of different target temperatures in cardiac arrest on mechanical ventilation, gas exchange and nosocomial infections using data collected during and after the TTM-trial.