Cerebral perfusion in sepsis-associated delirium.

INTRODUCTION: The pathophysiology of sepsis-associated delirium is not completely understood and the data on cerebral perfusion in sepsis are conflicting. We tested the hypothesis that cerebral perfusion and selected serum markers of inflammation and delirium differ in septic patients with and without sepsis-associated delirium. METHODS: We investigated 23 adult patients with sepsis, severe sepsis, or septic shock with an extracranial focus of infection and no history of intracranial pathology. Patients were investigated after stabilisation within 48 hours after admission to the intensive care unit. Sepsis-associated delirium was diagnosed using the confusion assessment method for the intensive care unit. Mean arterial pressure (MAP), blood flow velocity (FV) in the middle cerebral artery using transcranial Doppler, and cerebral tissue oxygenation using near-infrared spectroscopy were monitored for 1 hour. An index of cerebrovascular autoregulation was calculated from MAP and FV data. C-reactive protein (CRP), interleukin-6 (IL-6), S-100beta, and cortisol were measured during each data acquisition. RESULTS: Data from 16 patients, of whom 12 had sepsis-associated delirium, were analysed. There were no significant correlations or associations between MAP, cerebral blood FV, or tissue oxygenation and sepsis-associated delirium. However, we found a significant association between sepsis-associated delirium and disturbed autoregulation (P = 0.015). IL-6 did not differ between patients with and without sepsis-associated delirium, but we found a significant association between elevated CRP (P = 0.008), S-100beta (P = 0.029), and cortisol (P = 0.011) and sepsis-associated delirium. Elevated CRP was significantly correlated with disturbed autoregulation (Spearman rho = 0.62, P = 0.010). CONCLUSION: In this small group of patients, cerebral perfusion assessed with transcranial Doppler and near-infrared spectroscopy did not differ between patients with and without sepsis-associated delirium. However, the state of autoregulation differed between the two groups. This may be due to inflammation impeding cerebrovascular endothelial function. Further investigations defining the role of S-100beta and cortisol in the diagnosis of sepsis-associated delirium are warranted. TRIAL REGISTRATION: ClinicalTrials.gov NCT00410111.

Human factors affect the quality of cardiopulmonary resuscitation in simulated cardiac arrests.

AIM: Cardiopulmonary resuscitation is a team endeavour. There are only limited data on whether team performance during cardiopulmonary resuscitation is influenced by behavioural issues. The aim of the study was to determine whether and how human factors affect the quality of cardiopulmonary resuscitation. METHODS: 16 teams, each consisting of three health-care workers, were studied in a patient simulator. A scenario of witnessed cardiac arrest due to ventricular fibrillation was used. Ventricular fibrillation could be converted into sinus rhythm by two countershocks administered during the first 2 min or by two countershocks administered during the first 5 min provided that uninterrupted basic life support was started in under 60 s. Teams were rated to be successful if ventricular fibrillation was converted into sinus rhythm. Behavioural rating included leadership, task distribution, information transfer, and conflicts. RESULTS: Only six out of 16 teams were successful. Compared with successful teams, teams that failed exhibited significantly less leadership behaviour (P=0.033) and explicit task distribution (P=0.035). All teams shared among them sufficient theoretical knowledge to successfully treat the simulated cardiac arrest. CONCLUSIONS: In a scenario of simulated witnessed cardiac arrest almost two thirds of teams composed of qualified health-care workers failed to provide basic life support and/or defibrillation within an appropriate time window. Absence of leadership behaviour and absence of explicit task distribution were associated with poor team performance. Failure to translate theoretical knowledge into effective team activity appears to be a major problem.

Activities during interruptions in cardiopulmonary resuscitation: a simulator study.

OBJECTIVE: Studies investigating the quality of cardiopulmonary resuscitation (CPR) have revealed frequent unnecessary interruptions of life support. The primary objective of the study is to analyze what happens during interruptions. We investigated (a) whether interruptions are filled with "secondary activities", i.e., activities only indirectly related to the primary task of providing life support (e.g., preparatory and diagnostic activities), and (b) whether all group members focus on the same secondary activity during interruptions, thus impeding group coordination, and detracting from the primary task of providing life support. DESIGN: Prospective observational study. SETTING: Twenty teams of general practitioners were videotaped during a simulated cardiac arrest. OUTCOME MEASURES: Resuscitation performance was assessed as hands-on time according to resuscitation guidelines. Unnecessary interruptions were defined as periods the patient received no hands-on support. RESULTS: Teams of general practitioners achieved hands-on time in accordance with the resuscitation guidelines (chest compression/ventilation/defibrillation) during 62% of the time the patient had no pulse. Unnecessary interruptions consumed 32% of the available time. During most of the unnecessary interruption time, team members engaged in secondary medical activities, particularly observing the monitor (47%) and dealing with the defibrillator (47%). During 56% of the unnecessary interruption time, all team members focussed their attention on the same secondary activity, thus neglecting the need for task distribution among team members. CONCLUSIONS: Unnecessary interruptions of CPR occur frequently and consume approximately one-third of the time patients should receive continuous life support. Unnecessary interruptions are mainly characterized by secondary medical activities that may be perceived as meaningful. During the majority of unnecessary interruptions, all team members focus on the same secondary activity, indicating shortcomings in task distribution in the resuscitation team. The findings emphasize the importance of team training with particular emphasis on situational awareness and task distribution.

Performance of first responders in simulated cardiac arrests.

OBJECTIVE: Survival of in-hospital cardiac arrests depends more on first responders than on cardiac arrest teams. The objective of this study was to determine the adherence to algorithms of cardiopulmonary resuscitation of first responders in simulated cardiac arrests in intensive care. A second objective was to assess the effect of the early vs. late availability of a physician on the performance of nurse-based teams acting as first responders. DESIGN: Prospective study. SETTING: Patient simulator in a tertiary level intensive care unit. PARTICIPANTS: A total of 20 teams consisting of three registered nurses and one resident each. INTERVENTIONS: A simulated witnessed cardiac arrest due to ventricular fibrillation occurred in the presence of one nurse while the remaining two nurses could be called to help. Depending on the time of the residents' arrival, teams were classified as "early" (median arrival 50 secs after the onset of the arrest) or "late" (median arrival 150 secs after the onset of the arrest). MEASUREMENTS AND MAIN RESULTS: In all teams, the recognition of the arrest and the calling for help occurred in a timely fashion. However, a median of 85 secs (interquartile range [IQ], 130 secs) elapsed until the start of cardiac massage and 100 secs (IQ, 45 secs) to the first defibrillation. Once commenced, cardiac massage and mask ventilation were carried out during 61% (IQ, 33%) and 77% (IQ, 23%) of the possible time only. Delays and interruptions were generally not recalled by the participants. Compared with teams with late arriving residents, teams with early arriving residents administered more countershocks: 4.5 (IQ, 2) vs. 3.5 (IQ, 1.5; p = .026). CONCLUSIONS: First responders in intensive care often failed to build a team structure that ensured timely, effective, monitored, and ongoing team activity. The early availability of a physician increased the number of countershocks administered. Self-reporting is unsuitable to reliably assess the quality of cardiopulmonary resuscitation.