Neurohistological abnormalities during early porcine endotoxemia.

BACKGROUND: Brain dysfunction is common in sepsis. We aimed to assess whether cerebral perfusion, oxygenation, and/or metabolism are abnormal during early endotoxemia, and how they may relate to potential neurohistological changes. METHODS: In this prospective animal study, we included 12 pigs (weight: 42 ± 4 kg; mean ± SD) that were exposed to Escherichia coli lipopolysaccharide (E. coli LPS B0111 : B4, 0.4 µg/kg/h) or saline infusion (n = 6, each) for 10 h. Systemic hemodynamics, cerebral blood flow, intracranial pressure, and brain tissue oxygen tension were continuously measured. At the end of the experiment, formalin-fixed brains were cut in coronal sections and embedded in paraffin. Afterwards, the sections were cut at 5 microns and stained with hematoxylin and eosin. RESULTS: Stable systemic hemodynamics in both groups were associated with higher carotid arterial blood flow after 10 h of endotoxemia (9.0 ± 2.2 ml/kg/min) compared to controls (6.6 ± 1.2 ml/kg/min; time-group interaction: P = 0.014). Intracranial pressure, cerebral perfusion pressure, brain oxygen consumption, and brain tissue oxygen tension were similar in both groups. In four of the six endotoxemic animals but in none of the controls, cerebral tissue lesions were found (encephalomalacia with spongy degeneration of white matter, axonal swelling, and ischemic neuronal thalamic necrosis), including significant venous vascular alterations, predominantly in the brainstem, in three of the four animals. CONCLUSIONS: Early endotoxemia seems to be associated with histological signs of brain damage unrelated to systemic or cerebral hemodynamics or oxygenation.

Assessment of pain in sedated and mechanically ventilated patients: an observational study.

BACKGROUND: Critically ill patients often undergo unpleasant procedures. We quantified the effects of an unpleasant stimulus on physiological and behavioral parameters and evaluated how they are modified by sedation and analgesia. METHODS: A 6-month study in the 30-bed intensive care unit (ICU) of a university hospital examined 21 sedated patients from various diagnostic groups. Hemodynamic and respiratory parameters, pupil size, facial expression, muscle tone, body movement, and the Richmond Agitation-Sedation Scale (RASS) score were measured before and during intratracheal suctioning, first in sedated patients, after sedation was stopped, and after an opioid bolus. RESULTS: Before intratracheal suctioning, patients had RASS scores of -1.8 ± 1.2 (mean ± standard deviation; sedation), -0.6 ± 1.7 (sedation stop), and -0.9 ± 1.4 (analgesia) (P = 0.014). Intratracheal suctioning significantly increased RASS during both sedation (to -0.6 ± 1.7) and sedation stop (to 1.0 ± 1.5) (both P < 0.01), but not during analgesia. Systolic blood pressure increased during sedation (by 9 ± 10 mmHg), during sedation stop (by 15 ± 17 mmHg) and during analgesia (by 9 ± 4 mmHg; all P < 0.01), but diastolic pressure only during sedation and sedation stop (both P < 0.01). Facial expression, body movement, and muscle tone changed significantly during the episodes of intratracheal suctioning. Heart rate, tidal volume, and pupil size remained stable under all conditions. CONCLUSIONS: Intratracheal suctioning evoked significant changes in some physiological and behavioral parameters. Some physiological changes were suppressed by analgesia, but at our ICU's standard doses, neither analgesia nor sedation attenuated changes in behavioral parameters at the intensity tested.

Forced fluid removal in critically ill patients with acute kidney injury.

PURPOSE: The aim was to test the feasibility of protocol-driven fluid removal with continuous renal replacement therapy (CRRT) in patients in whom standard fluid balance prescription did not result in substantial negative fluid balances. MATERIALS AND METHODS: In 10 mechanically ventilated patients with sepsis or signs of inflammation and acute kidney injury [age 65 (48-78 years; median, range), simplified acute physiology score II 66 (39-116)], fluid removal was guided by mean arterial pressure (MAP), cardiac index (CI), mixed venous oxygen saturation (SvO(2)), lactate/base excess, peripheral circulation, and filling pressures, and adjusted hourly with the goal to maximize volume removal for up to 3 days. RESULTS: Fluid removal rates during the 3 days before and during the study period [66 (36-72) h] were 11 (-30 to +36) ml/kg/day and -59 (-85 to -31) ml/kg/day, respectively (P = 0.002). In 12% of a total of 594 fluid removal rate evaluations, fluid removal had to be decreased or stopped. Most frequent reasons leading to decreasing fluid removal were (n, % of all instances, median lowest value from all patients): SvO(2) (44, 28%, 59%), MAP (36, 23%, 57 mmHg), CI (26, 17%, 2.4 l/min/m(2)), low peripheral temperature (22, 14%, 'cold'). Overall, systemic hemodynamics remained stable throughout the study period. CONCLUSIONS: In these patients, protocolized fluid removal with CRRT was associated with large negative fluid balances.

Oxygen transport and mitochondrial function in porcine septic shock, cardiogenic shock, and hypoxaemia.

INTRODUCTION: The relevance of tissue oxygenation in the pathogenesis of organ dysfunction during sepsis is controversial. We compared oxygen transport, lactate metabolism, and mitochondrial function in pigs with septic shock, cardiogenic shock, or hypoxic hypoxia. METHODS: Thirty-two anaesthetized, ventilated pigs were randomized to faecal peritonitis (P), cardiac tamponade (CT), hypoxic hypoxia (HH) or controls. Systemic and regional blood flows, lactate, mitochondrial respiration, and tissue hypoxia-inducible factor 1 alpha (HIF-1α) were measured for 24 h. RESULTS: Mortality was 50% in each intervention group. While systemic oxygen consumption (VO(2) ) was maintained in all groups, hepatic VO(2) tended to decrease in CT [0.84 (0.5-1.3) vs. 0.42 (0.06-0.8)/ml/min/kg; P = 0.06]. In P, fractional hepatic, celiac trunk, and portal vein blood flows, and especially renal blood flow [by 46 (14-91)%; P = 0.001] decreased. In CT, renal blood flow [by 50.4 (23-81)%; P = 0.004] and in HH, superior mesenteric blood flow decreased [by 38.9 (16-100)%, P = 0.009]. Hepatic lactate influx increased > 100% in P and HH, and > 200% in CT (all P < 0.02). Hepatic lactate uptake remained unchanged in P and HH and converted to release in CT. Mitochondrial respiration remained normal. Muscle adenosine triphosphate (ATP) concentrations decreased in P (5.9 ± 1.4 µmol/g wt vs. 2.8 ± 2.7 µmol/g wt, P = 0.04). HIF-1α expression was not detectable in any group. CONCLUSION: We conclude that despite shock and renal hypoperfusion, tissue hypoxia is not a major pathophysiological issue in early and established faecal peritonitis. The reasons for reduced skeletal muscle tissue ATP levels in the presence of well-preserved in-vitro muscle mitochondrial respiration should be further investigated.

Effect of site of lactate infusion on regional lactate exchange in pigs.

BACKGROUND: The rate of extra-hepatic lactate production and the route of influx of lactate to the liver may influence both hepatic and extra-hepatic lactate exchange. We assessed the dose-response of hepatic and extra-hepatic lactate exchange during portal and central venous lactate infusion. METHODS: Eighteen pigs randomly received either portal (n=5) or central venous (n=7) lactate infusion or saline (n=6). Sodium lactate was infused at 33, 66, 99, and 133 µmol kg⁻¹ min⁻¹ for 20 min each. Systemic and regional abdominal blood flows and plasma lactate were measured at 20 min intervals until 1 h post-infusion, and regional lactate exchange was calculated (area under lactate uptake-time curve). RESULTS: Total hepatic lactate uptake [median (95% confidence interval)] during the experimental protocol (140 min) was higher during portal [8198 (5487-12 798) µmol kg(-1)] than during central venous lactate infusion [4530 (3903-5514) µmol kg⁻¹, P<0.05]. At a similar hepatic lactate delivery (∼400 µmol kg⁻¹ min⁻¹), hepatic lactate uptake [mean and standard deviation (sd)] was higher during portal [118 (sd 55) µmol kg⁻¹ min⁻¹] than during central venous lactate infusion [44 (12) µmol kg⁻¹ min⁻¹, P < 0.05]. Time courses of arterial lactate concentrations and lactate uptake at other measured regions were similar in both groups. CONCLUSIONS: Higher hepatic lactate uptake during portal compared with central venous lactate infusion at a similar total hepatic lactate influx underlines the role of portal vein lactate concentration in total hepatic lactate uptake capacity. Arterial lactate concentration does not depend on the site of lactate infusion. At higher arterial lactate concentrations, all regions participated in lactate uptake.

The association between early hemodynamic variables and outcome in normothermic comatose patients following cardiac arrest.

BACKGROUND: Currently, few data exist on the association between post-cardiac arrest hemodynamic function and outcome. In this explorative, retrospective analysis, the association between hemodynamic variables during the first 24 h after intensive care unit admission and functional outcome at day 28 was evaluated in 153 normothermic comatose patients following a cardiac arrest. METHODS: Medical records of a multidisciplinary intensive care unit were reviewed for comatose patients (Glasgow Coma Scale < or = 9) admitted to the intensive care unit after successful resuscitation from an in- or an out-of-hospital cardiac arrest. The hourly variable time integral of hemodynamic variables during the first 24 h after admission was calculated. At day 28, outcome was assessed as favorable or adverse based on a Cerebral Performance Category of 1-2 and 3-5, respectively. Bi- and multivariate regression models adjusted for relevant confounding variables were used to evaluate the association between hemodynamic variables and functional outcome. RESULTS: One hundred and fifty-three normothermic comatose patients were admitted after a cardiac arrest, of whom 64 (42%) experienced a favorable outcome. Neither in the adjusted bivariate models (r(2), 0.61-0.78) nor in the adjusted multivariate model (r(2), 0.62-0.73) was the hourly variable time integral of any hemodynamic variable during the first 24 h after intensive care unit admission associated with functional patient outcome at day 28 in all patients as well as in patients after an in- or an out-of-hospital cardiac arrest. CONCLUSION: Commonly measured hemodynamic variables during the first 24 h following intensive care unit admission due to a cardiac arrest do not appear to be associated with the functional outcome at day 28.

Correlation of optic nerve and optic nerve sheath diameter with intracranial pressure in pigs.

OBJECTIVE: Several studies have shown an association between intracranial pressure and the diameter of the optic nerve sheath measured by transbulbar ultrasonography. To understand the pathophysiology of this phenomenon, we aimed to measure the changes of the optic nerve, optic nerve sheath and perineural space separately with increasing intracranial pressure in a porcine model. METHODS: An external ventricular drain was placed into the third ventricle through a right paramedian burrhole in eight anesthesized pigs. The diameters of the optic nerve and the optic nerve sheath were measured while the intracranial pressure (ICP) was increased in steps of 10mmHg from baseline up to 60 mmHg. RESULTS: The median diameters of the optic nerve (ON) increased from 0.36 cm (baseline- 95% confidence interval (CI) 0.33 cm to 0.45 cm) to 0.68 cm (95% CI 0.57 cm to 0.82 cm) at ICP of 60 mmHg (p<0.0001) and optic nerve sheath (ONS) from 0.88 cm (95% CI 0.79 cm to 0.98 cm) to 1.24 cm (95% CI 1.02 cm to 1.38 cm) (p< 0.002) while the median diameter of the perineural space (PNS) (baseline diameter 95% CI 0.40 cm to 0.59 cm to diameters at ICP 60 95% CI 0.38 cm to 0.62 cm) did not change significantly (p = 0.399). Multiple comparisons allowed differentiation between baseline and values ≥40 mmHg for ON (p = 0.017) and between baseline and values ≥ 50mmHg for ONS (p = 0.006). A linear correlation between ON (R2 = 0.513, p<0.0001) and ONS (R2 = 0.364, p<0.0001) with ICP was found. The median coefficient of variation for intra- and inter-investigator variability was 8% respectively 2.3%. CONCLUSION: Unexpectedly, the increase in ONS diameter with increasing ICP is exclusively related to the increase of the diameter of the ON. Further studies should explore the reasons for this behaviour.

Time- and dose-related regional fluxes of tissue-type plasminogen activator in anesthetized endotoxemic pigs.

BACKGROUND: Acute endotoxinemia elicits an early fibrinolytic response. This study analyzes the effects of the dose and duration of endotoxin infusion on arterial levels of tissue-type plasminogen activator (tPA) and pulmonary, mesenteric and hepatic plasma tPA fluxes. METHODS: Pigs were randomized to receive an acute, high-dose (for 6 h, n=13, high ETX) or a prolonged, low-dose (for 18 h, n=18, low ETX) infusion of endotoxin or saline vehicle alone (for 18 h, n=14, control). All animals were fluid resuscitated to maintain a normodynamic circulation. Systemic and regional blood flows were measured and arterial, pulmonary arterial, portal and hepatic venous blood samples were analyzed to calculate regional net fluxes of tPA. Plasma tumor necrosis factor (TNF-alpha) levels were analyzed. RESULTS: Mesenteric tPA release and hepatic uptake increased maximally at 1.5 h in ETX groups related to dose. Maximal mesenteric tPA release [high ETX 612 (138-1185) microg/min/kg, low ETX 72 (32-94) microg/min/kg, median+/-interquartile range] and hepatic tPA uptake [high ETX -1549 (-1134 to -2194) microg/min/kg, low ETX -153 (-105 to -307) microg/min/kg] correlated to TNF-alpha levels. Regional tPA fluxes returned to baseline levels at 6 h in both ETX groups and also remained low during sustained low ETX. No changes were observed in control animals. CONCLUSIONS: Endotoxemia induces an early increase in mesenteric tPA release and hepatic tPA uptake related to the severity of endotoxemia. The time patterns of changes in mesenteric and hepatic tPA fluxes are similar in acute high-dose endotoxemia and sustained low-dose endotoxemia.

Prevention of systemic hyperlactatemia during splanchnic ischemia.

Arterial blood lactate increases as a result of poor tissue perfusion. In splanchnic hypoperfusion, increased hepatic lactate uptake may delay increases in arterial blood lactate. We hypothesized that during isolated reduction of mesenteric blood flow, maintaining systemic blood volume and flow by fluid resuscitation may prevent systemic hyperlactatemia and therefore mask splanchnic ischemia. In a randomized study, 7 pigs were subjected to 4 h of splanchnic hypoperfusion by reducing the superior mesenteric artery blood flow to 4 +/- 0.8 mL/kg min [mean +/- standard deviation (SD)]. Seven pigs served as controls. Fluid was administered in order to keep the pulmonary artery occlusion pressure at 5 to 8 mm Hg. Cardiac output, portal vein, superior mesenteric, and hepatic arterial blood flow were measured every 30 min. Arterial, mixed venous, hepatic, portal, and mesenteric venous blood lactate, and jejunal mucosal pCO2 were measured at baseline and thereafter at 30-min intervals. The initial decrease in portal venous blood flow in the ischemic animals was subsequently counterbalanced by increasing hepatic arterial blood flow from 2 +/- 1 mL x kg(-1) x min(-1) at baseline to 11 +/- 4 mL x kg(-1) x min(-1) [after 4 h of ischemia; mean +/- standard deviation (SD), P = 0.02]. Jejunal mucosal- and mesenteric vein-arterial pCO2 gradients increased in the ischemic group from 11 +/- 8 mm Hg to 73 +/- 5 mm Hg (P = 0.02), and from 10 +/- 4 mm Hg to 44 +/- 8 mm Hg, respectively (P = 0.02). Mesenteric and portal venous lactate increased in the ischemic animals from 1.1 +/- 0.3 mmol/L to 4.2 +/- 1.0 mmol/L (P = 0.02), and from 1.0 +/- 0.2 mmol/L to 1.6 +/- 0.3 mmol/L, respectively (P = 0,03). While mesenteric lactate production and hepatic lactate uptake increased in parallel in the ischemic animals from 5 +/- 6 micromol x kg(-1) x min(-1) to 14 +/- 5 micromol x kg(-1) x min(-1) (P = 0.04), and from 14 +/- 7 micromol x kg(-1) x min(-1) to 24 +/- 6 micromol x kg(-1) x min(-1), respectively (P = 0.02), hepatic venous and arterial lactate, and apparent splanchnic lactate uptake and extraction did not change. We conclude that the hepatic lactate uptake increases in response to hepatic lactate influx. Systemic hyperlactatemia and increased hepatic venous lactate concentrations are late consequences of mesenteric hypoperfusion if hypovolemia is prevented. The net exchange of lactate across the splanchnic region does not reflect hepato-portal lactate kinetics in this animal model of intestinal hypoperfusion.

Intensive care 1980-1995: change in patient characteristics, nursing workload and outcome.

OBJECTIVE: To assess temporal changes in patient characteristics, nursing workload and outcome of the patients and to compare the actual amount of available nursing staff with the estimated needs in a medical-surgical ICU. DESIGN: Retrospective analysis of prospectively collected data. SETTING: A medical-surgical adult intensive care unit (ICU) in a Swiss university hospital. PATIENTS: Data of all patients staying in the ICU between January 1980 and December 1995 were included. INTERVENTIONS: None. MEASUREMENTS AND RESULTS: The estimated number of nurses needed was defined according to the Swiss Society of Intensive Care Medicine (SGI) grading system: category I = one nurse/patient/shift (= 8 h), category II = one nurse/two patients/shift, category III = one nurse/three patients/shift. An intervention score (IS) was obtained, based on a number of specific activities in the ICU. There was a total of 35,327 patients (32% medical and 68% postoperative/trauma patients). Over time, the number of patients per year increased (1980/1995: 1,825/2,305, p < 0.001) and the length of ICU stay (LOS) decreased (4.1/3.8 days, p < 0.013). There was an increase in the number of patients aged > 70 years (19%/28%, p < 0.001), and a decrease in the number of patients < 60 years (58%/41%, p < 0.001). During the same time period, the IS increased two-fold. Measurement of nursing workload showed an increase over time. The number of nursing days per year increased (1980/1995: 7454/8681, p < 0.019), as did the relative amount of patients in category I (49%/71%, p < 0.001), whereas the portion of patients in category II (41%/28%, p < 0.019) and category III (10%/0%) decreased. During the same time period, mortality at ICU discharge decreased (9.0%/7.0%, p < 0.002). CONCLUSIONS: During the last 16 years, there has been a marked increase in workload at this medical-surgical ICU. Despite an increase in the number of severely sick patients (as defined by the nursing grading system) and patient age, ICU mortality and LOS declined from 1980 to 1995. This may be ascribed to improved patient treatment or care. Whether an increasingly liberal discharge policy (transfer to newly opened intermediate care units, transfer of patients expected to die to the ward) or a more rigorous triage (denying admission to patients with a very poor prognosis) are confounding factors cannot be answered by this investigation. The present data provide support for the tenet that there is a trend toward more complex therapies in increasingly older patients in tertiary care ICUs. Calculations for the number or nurses needed in an ICU should take into acount the increased turnover of patients and the changing patient characteristics.

Gut luminal lactate release during gradual intestinal ischemia.

OBJECTIVES: Gut ischemia induced by occlusion of the superior mesenteric artery (SMA) results in release of lactate into the gut lumen. We studied the threshold of SMA flow reduction that leads to increased gut luminal lactate during stepwise reduction in SMA blood flow. DESIGN AND SETTING: A randomized, controlled animal experiment in a university experimental research laboratory. INTERVENTIONS: Anesthetized, normoventilated, normovolemic domestic pigs were randomized to ischemia ( n=7) and sham groups ( n=7). SMA blood flow was reduced stepwise at 15-min intervals by 22%, 44%, 66%, and 88% and kept constant thereafter for 60 min. MEASUREMENTS AND RESULTS: Jejunal luminal microdialysate lactate and mucosal pCO(2) were measured every 15 min. The luminal lactate increased over the upper normal limit of 0.2 mmol/l at a median SMA blood flow of 9.6 ml kg(-1) min(-1) (range 7.5-23.7). In five of seven animals the increase in luminal lactate was preceded by or accompanied by an increase in the mucosal-arterial pCO(2) gradient. CONCLUSIONS: There is a threshold of SMA blood flow below which gut luminal lactate increases, indicating mucosal anaerobic metabolism. Measurement of gut luminal lactate by microdialysis can be used to assess the adequacy of gut perfusion and the onset of anaerobic metabolism.

Prevention of acute renal failure--fluid repletion and colloids.

Hypovolemia alone or in conjunction with other factors is a main reason for acute renal failure in critically ill patients. Various crystalloid and colloid solutions are available to correct hypovolemia. Some of them have been implicated in impairment of renal function. Infusion of large amounts of sodium chloride is associated with increased incidence of nausea, vomiting and hyperchloremic metabolic acidosis. While gelatins and HES are preferred colloids in patients with normal kidney function, there is some evidence that the latter are associated with impaired renal function in patients with pre-existing kidney disease. Any hyperoncotic colloid given in large amounts may decrease glomerular filtration, and should therefore be combined with crystalloids.

Quantification of haemodynamic response to auditory stimulus in intensive care.

Measuring effects of sensory stimuli on haemodynamics could provide information about the interplay between central and autonomous nervous system (ANS). However, ANS response to sensory stimulus has received little attention. In this paper we present a signal processing scheme to extract the responses of heart rate and systemic arterial pressure on auditory stimulus in intensive care patients (N=5). In short, the effect of mechanical ventilation is rejected by optimal linear modelling. Other disturbances are attenuated by filtering and efficient rejection of outlying sweeps of data. The results show identifiable responses on three out of five cases. The response characteristics may be explained by synchronisation of spontaneous variability in systemic arterial pressure to auditory stimulus.

IBIS data library: clinical description of the Finnish database. Improved Monitoring for Brain Dysfunction during Intensive Care and Surgery.

Improved monitoring of brain function in intensive care and surgery is a project aiming to develop methods of biosignal processing and interpretation, in order to characterise critical events during anaesthesia and cardiac surgery, effect of anaesthesia regimen, neurophysiological findings of different sedation levels, arousal from post-anaesthesia sedation, post-cardiopulmonary bypass brain dysfunction and early brain dysfunction in patients with multiple organ failure. A data library was collected in the three participating hospitals. This is a description of the data library from the University Hospital of Kuopio, Finland, which includes data from 40 patients after cardiac surgery and from seven patients with multiple organ dysfunction. This project demonstrates that active neuromonitoring can be performed in the intensive care unit without interference with the normal treatment and care. The presented database may serve other scientific workers as a reference for a typical spectrum of perioperative data with respect to severity of disease, length of cardiopulmonary bypass, postoperative levels of sedation and length of hospital stay in cardiac surgery patients, and for types of diseases and outcome in patients with multiple organ dysfunction.

The influence of local and systemic preconditioning on oxygenation, metabolism and survival in critically ischaemic skin flaps in pigs.

Stress proteins represent a group of highly conserved intracellular proteins that provide adaptation against cellular stress. The present study aims to elucidate the stress protein-mediated effects of local hyperthermia and systemic administration of monophosphoryl lipid A (MPL) on oxygenation, metabolism and survival in bilateral porcine random pattern buttock flaps. Preconditioning was achieved 24h prior to surgery by applying a heating blanket on the operative site (n = 5), by intravenous administration of MPL at a dosage of 35 microg/kg body weight (n = 5) or by combining the two (n = 5). The flaps were monitored with laser Doppler flowmetry, polarographic microprobes and microdialysis until 5h postoperatively. Semiquantitative immunohistochemistry was performed for heat shock protein 70 (HSP70), heat shock protein 32 (also termed haem oxygenase-1, HO-1), and inducible nitrc oxide synthase (iNOS). The administration of MPL increased the impaired microcirculatory blood flow in the proximal part of the flap and partial oxygen tension in the the distal part by approximately 100% each (both P<0.05), whereas both variables remained virtually unaffected by local heat preconditioning. Lactate/pyruvate (L/P) ratio and glycerol concentration (representing cell membrane disintegration) in the distal part of the flap gradually increased to values of approximately 500 mmol/l and approximately 350 micromol/l, respectively (both P<0.01), which was substantially attenuated by heat application (P<0.01 for L/P ratio and P<0.05 for glycerol) and combined preconditioning (P<0.01 for both variables), whereas the effect of MPL was less marked (not significant). Flap survival was increased from 56% (untreated animals) to 65% after MPL (not significant), 71% after heat application (P<0.05) and 78% after both methods of preconditioning (P<0.01). iNOS and HO-1 were upregulated after each method of preconditioning (P<0.05), whereas augmented HSP70 staining was only observed after heat application (P<0.05). We conclude that local hyperthermia is more effective in preventing flap necrosis than systemic MPL administration because of enhancing the cellular tolerance to hypoxic stress, which is possibly mediated by HSP70, whereas some benefit may be obtained with MPL due to iNOS and HO-1-mediated improvement in tissue oxygenation.