Predicting death and readmission after intensive care discharge.

BACKGROUND: Despite initial recovery from critical illness, many patients deteriorate after discharge from the intensive care unit (ICU). We examined prospectively collected data in an attempt to identify patients at risk of readmission or death after intensive care discharge. METHODS: This was a secondary analysis of clinical audit data from patients discharged alive from a mixed medical and surgical (non-cardiac) ICU. RESULTS: Four hundred and seventy-five patients (11.2%) died in hospital after discharge from the ICU. Increasing age, time in hospital before intensive care admission, Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and discharge Therapeutic Intervention Scoring System (TISS) score were independent risk factors for death after intensive care discharge. Three hundred and eighty-five patients (8.8%) were readmitted to intensive care during the same hospital admission. Increasing age, time in hospital before intensive care, APACHE II score, and discharge to a high dependency unit were independent risk factors for readmission. One hundred and forty-three patients (3.3%) were readmitted within 48 h of intensive care discharge. APACHE II scores and discharge to a high dependency or other ICU were independent risk factors for early readmission. The overall discriminant ability of our models was moderate with only marginal benefit over the APACHE II scores alone. CONCLUSIONS: We identified risk factors associated with death and readmission to intensive care. It was not possible to produce a definitive model based on these risk factors for predicting death or readmission in an individual patient.

Polymorphonuclear leukocytes from patients with severe sepsis have lost the ability to degrade fibrin via u-PA.

Fibrin persistence in the vasculature is an important complication of sepsis that can often lead to mortality. We have previously established that polymorphonuclear leukocytes (PMN) from healthy individuals have the capacity to degrade fibrin via urokinase-type plasminogen activator (u-PA). We have also demonstrated an increase in u-PA antigen in the plasma of patients suffering from septic shock. In this study, we investigate the hypothesis that PMN from patients with sepsis have lost their fibrinolytic ability and that this might contribute to the persistence of fibrin deposits. We show here that PMN from these patients do not express any u-PA activity, despite retaining some u-PA antigen. Additionally, thrombi prepared from the whole blood of the patients exhibit reduced endogenous lysis compared with those from healthy individuals. These data indicate that loss of fibrinolytic activity from PMN may be a contributing factor in fibrin persistence in the microvasculature in sepsis.

Plasminogen activator inhibitor 2 and urokinase-type plasminogen activator in plasma and leucocytes in patients with severe sepsis.

Proteins influencing plasminogen activation to plasmin, namely plasminogen activators tissue-type plasminogen activator (t-PA) and urokinase-type plasminogen activator (u-PA) and their principal inhibitors, plasminogen activator inhibitor 1 (PAI-1) and PAI-2, were measured in the plasma, the polymorph and mononuclear cell fractions taken from patients with major sepsis who were entering a general intensive care unit. The purpose of this study was to elucidate the factors favouring the persistence of fibrin in the microvasculature and thus contributing to multiple organ failure. Levels of u-PA antigen in plasma rose in sepsis and u-PA activity, not detectable in normal plasma, appeared. Levels of u-PA antigen in the cell fractions fell concomitantly. t-PA antigen in plasma and in the mononuclear cell fraction rose in sepsis, but t-PA activity was not detectable. Plasma PAI-1 antigen levels were strikingly raised in sepsis, presumably accounting for the complete neutralization of t-PA activity. PAI-2 antigen, not normally detected in plasma, appeared in the plasma of some patients, whereas it disappeared from the cellular fractions. Appearance of PAI-2 in plasma was associated with non-survival of the patient. The observations indicate that all the agents involved in plasminogen activation are released into the plasma in major sepsis. The levels of PAI-1 reached were quantitatively sufficient to suppress all activity of the released t-PA, but the inhibitors did not prevent expression of u-PA activity in the circulation. Circulating active u-PA and PAI-2 in the plasma of patients with severe sepsis may represent material originating from leucocytes. Leucocyte release of these agents within fibrin deposits may influence the persistence of fibrin and thus the development of multiple organ failure.

Fibrinolytic changes in a patient with toxic shock syndrome; release of active u-PA.

OBJECTIVE: Definition of the changes in the activators of plasminogen, u-PA and t-PA, and examination of the possible generation of plasmin in the circulation in overwhelming sepsis. DESIGN: Serial blood analysis starting 4 h after development of symptoms of toxic shock syndrome. SETTING: Intensive care unit. PATIENT: A previously healthy woman underwent endometrial ablation and rapidly thereafter developed staphylococcal toxic shock syndrome with organ failure. MEASUREMENT AND RESULT: t-PA, PAI-1, t-PA-PAI-1 complexes, plasminogen, fibrinogen and plasmin-alpha 2-antiplasmin complexes were measured serially by ELISA and free u-PA by SDS-PAGE with zymography. The onset of symptoms was accompanied by a rise of t-PA antigen-followed rapidly by PAI-1 antigen. Plasmin-alpha 2-antiplasmin complex was generated in large amounts but disappeared within hours. From day 3, free u-PA was detectable in the circulation without plasmin generation. CONCLUSION: Despite the sustained presence of active u-PA in the circulation and of t-PA antigen at the onset of symptoms, plasmin-alpha 2-antiplasmin generation was largely suppressed by high levels of PAI-1. The suppression of plasmin generation by u-PA and t-PA may ensure the persistence of fibrin in the microcirculation and so contribute to organ failure.

Evidence for an active fibrinolytic system in normal human bone marrow.

Normal human bone marrow from patients undergoing heart surgery was analysed quantitatively for components of the fibrinolytic system, using functional and immunological assays. Marrow was found to contain considerable fibrinolytic activity, reflecting high levels of t-PA (tissue-type plasminogen activator). The t-PA was in an active form, despite the presence of the inhibitors PAI-1 and PAI-2. Plasminogen and alpha2-antiplasmin (alpha2-AP) were also present in marrow. The balance of proteases and inhibitors differed dramatically from that observed in plasma, with higher levels of t-PA, PAI-1 and PAI-2, and lower levels of u-PA (urokinase), plasminogen, alpha2-AP and t-PA-PAI-1 complex in bone marrow, and resulted in favourable conditions for fibrinolysis. The presence of plasmin-alpha2-AP complex at concentrations of the same order of magnitude as total plasminogen and alpha2-AP demonstrated that active generation of plasmin was indeed occurring. A role for the active fibrinolytic system in normal human bone marrow may be the removal of unnecessary fibrin deposits formed in the cavities of the marrow, in order to maintain flow through this tissue.

Inhibitors of plasminogen activator in neutrophils and mononuclear cells from septic patients.

Leucocytes, both polymorphs and mononuclear cells, play a variety of roles in the evolution of human response to sepsis, both local and generalised. In this study, inhibitors of plasminogen activator were measured in leucocytes from normal and septic patients. Plasminogen activator inhibitor-1 (PAI-1) was identified in polymorphs from normal individuals and levels rose significantly in polymorphs from septic patients: neutrophils from normal subjects did not contain PAI-2 but this protein was detectable in significant quantities in polymorph preparations from septic patients. In contrast, mononuclear cells from normal and septic patients contained no detectable quantities of PAI-1. Significant amounts of PAI-2 were present in normal mononuclear cells, and the levels rose significantly in monocytes from septic patients. PAI-2 is thus here identified in human subjects, distinct from those with pregnancy or malignancy, as playing a role in a pathological process. The increased levels of both inhibitors produced by leucocytes may clearly contribute directly to the persistence of fibrin, a characteristic feature of the response to infection, local or general; they may thus participate in successful localisation of infections (abscess formation etc.) and in the evolution of the major systemic complications of disseminated sepsis characterised by microvascular occlusion by fibrin such as renal failure, shock lung or digital ischaemia.

Influence of white blood cells on the fibrinolytic response to sepsis: studies of septic patients with or without severe leucopenia.

In septic patients capable of normal white cell responses, high plasma levels of PAI-I, t-PA antigen and t-PA-PAI-I complex were observed. The ratios of t-PA and PAI-I were such that free PA activity was almost never observed. In patients severely leucopenic prior to becoming septic the changes were significantly less marked, so presence of leucocytes enhances the fibrinolytic inhibition occurring in sepsis. The non-leucopenic septic group showed greater evidence of thrombin generation in that FPA levels were higher but fibrinogen levels were only slightly less and antithrombin levels not different from those in the leucopenic group. A greater tendency to fibrin deposition and the striking fibrinolytic inhibition noted in patients with normal white cell responses may contribute to the development of some of the complications of sepsis in which fibrin deposition participates and may explain their relative rarity in leucopenic patients. When shock supervened, levels of PAI-I were high in both leucopenic and non-leucopenic groups, indicating that a source of PAI-I outwith the leucocytes themselves contributes to the phenomena observed.

Plasma nitrate concentrations in neutropenic and non-neutropenic patients with suspected septicaemia.

Nitric oxide (NO) is an important physiological mediator of vascular tone and is thought to be involved in the pathogenesis of septic shock. Plasma nitrate is the stable end product of NO oxidation and in part reflects endogenous NO production. We measured plasma nitrate levels in 47 episodes of suspected septicaemia in 43 in-patients (16 male and 27 female, age 15-63 years). Nitrate concentrations were significantly higher (P < 0.01) compared to healthy controls. Further analysis revealed that significantly elevated levels occurred only in the septic patients who had normal or elevated numbers of neutrophils in the peripheral blood and were hypotensive on presentation. Failure of plasma nitrate concentrations to rise significantly in patients with neutropenia suggests that this cell type may be important in the activation of the arginine-NO system in severe sepsis in man.

Sublethal percussion trauma in vitro causes a persisting derangement in the nonthrombogenic properties of brain endothelial cells.

The delivery of a blow to the head represents a transfer of energy, part of which manifests itself as a short-lived pressure change within the skull. An in vitro model was developed to test whether cerebral endothelial cell hemostatic function is altered with exposure to this type of pressure event. Human cerebral microvascular endothelium (HCME) cells were subjected to rapid (2-5 msec) changes in pressure (delta atmosphere = 1.2-10), the sublethal range defined (delta atmosphere < or = 6.5), and the nonthrombogenic status of sublethally percussed HCME cells assessed using the adherence of alpha-thrombin activated platelets as an indicator. The HCME cells had lost their normal capacity to suppress adherence of activated platelets when evaluated 1 hour or 24 hours after percussion. Adherence of activated platelets to percussed HCME cells was blocked by the addition of PGI2, an inhibitor of platelet adherence, when evaluated at 1 hour but not 24 hours after percussion, indicating that percussed HCME cells were undergoing further derangement of their nonthrombogenic mechanisms. Percussed HCME cells cultured for 24 hours in medium containing scavengers of oxygen free radicals recovered their capacity to block platelet adherence. We conclude that sublethal percussion immediately compromises the nonthrombogenic character of HCME cells and initiates the development of a persisting prothrombotic state in HCME cells. This derangement appears linked to increased production of reactive oxygen species by percussed HCME cells.

Chronic or recurrent meningitis. Neurosurgical perspectives.

Chronic meningitis is not a commonly encountered clinical problem in the day-to-day patient care responsibilities of neurosurgeons. It is, however, important for the neurosurgeon to be familiar with the clinical problems, diagnostic challenges, and therapeutic options required to establish a definitive diagnosis. This article concentrates on specific entities causing chronic or recurrent meningitis that may require neurosurgical participation.

Increased cerebral blood flow in idiopathic pseudotumour cerebri.

Cerebral blood flow was studied in nine patients with idiopathic pseudotumour and one patient with cortical vein thrombosis in Denver, Colorado using the 133Xe inhalation method. Globally elevated blood flows were found in all of the idiopathic pseudotumour patients averaging 149% of control values generated in the same setting. The patient with the cortical vein thrombosis demonstrated normal global flows. Possible pathophysiological mechanisms for these findings are discussed.

Colloidal volume expansion during acute cerebral ischaemia: assessed by local cerebral blood flow and computerized power ratio index.

Colloidal volume expansion during acute cerebral ischaemia was assessed by local cerebral blood flow (CBF) and the power ratio index (PRI) in 8 anaesthetized Macaque monkeys. Focal cerebral ischaemia was produced by right middle cerebral artery occlusion. The animals were then volume expanded (to maximum cardiac output) with 6% hetastarch and then exsanguinated to baseline cardiac output. During volume expansion, local CBF in the ischaemic hemisphere increased from 25 +/- 12 to 39 +/- 23 cc/100 g/min (p less than 0.01) and during exsanguination decreased to 32 +/- 18 cc/100 g/min. Local CBF did not change significantly in the nonischaemic hemisphere. EEG power data, as assessed by PRI [(delta + theta power/alpha + beta power) x 100] changed significantly during blood volume manipulation. The mean PRI value in the right hemisphere deteriorated by increasing from 65 +/- 22 to 94 +/- 25 (p less than 0.01) following vessel occlusion but improved by decreasing to 81 +/- 23 (p less than 0.05) following volume expansion. Following exsanguination, the PRI value increased to 87 +/- 21. These data demonstrate the benefits of volume expansion during acute cerebral ischaemia. Changes in local CBF were consistently associated with changes in the PRI maps and values.

Pulsatile versus nonpulsatile blood flow in the treatment of acute cerebral ischemia.

The effects of pulsatile and nonpulsatile perfusion on local cerebral blood flow (CBF) and on computerized mapping (CME) of electroencephalograms (EEG) in nonischemic and ischemic brain were studied using a canine stroke model. Nine anesthetized mongrel dogs were placed on normothermic right atrial-femoral artery cardiopulmonary bypass at a flow of 100 ml/kg/minute. Local CBF measurements and CME data were collected during nonpulsatile perfusion and maximal pulsatile perfusion. The stroke model was then produced, and local CBF measurements and CME data were again collected during nonpulsatile and pulsatile perfusion. In the nonischemic brain, local CBF increased 19%, from 32 +/- 10 to 38 +/- 11 ml/100 g/minute (P less than 0.01), when perfusion was changed from nonpulsatile flow (pulse pressure less than 4mm Hg) to pulsatile flow (pulse pressure 39 +/- 11 mm Hg). In the ischemic brain, local CBF increased 55%, from 11 +/- 5 to 17 +/- 7 ml/100 g/minute (P less than 0.01), when perfusion was changed from nonpulsatile (pulse pressure less than 3 mm Hg) to pulsatile (pulse pressure 36 +/- 7) flow. EEG power data, expressed as a power ratio index (PRI = low frequency power/high frequency power), improved significantly, from 110 +/- 33 to 101 +/- 41 (P less than 0.01) with pulsatile perfusion. These data demonstrate the importance of pulsatile blood flow in ischemic brain.

Blood volume expansion with hetastarch in acute ischaemic stroke: the effects on local cerebral blood flow and computer mapped EEG.

Focal cerebral ischaemia was induced in seven anaesthetized monkeys by unilateral middle cerebral artery (MCA) occlusion. Local cerebral blood flow (CBF) and computer mapped EEG (CME) changes were then studied as blood volume and cardiac output (CO) were varied. CO was increased by colloidal volume expansion and decreased by exsanguination. Local CBF fell to 24 +/- 9% of control values in ischaemic areas following MCA occlusion and increased to 43 +/- 19% of control values (p less than 0.01) when CO was increased by 130 +/- 70% with colloid infusion. Local CBF to nonischaemic regions was not altered significantly by blood volume expansion. Exsanguination led to return of CO to control levels and was associated with reduction of local CBF in ischaemic regions to 24 +/- 12% of control values (p less than 0.05). CME showed bifrontal or ipsilateral slow wave foci (delta) following MCA occlusion. Blood volume expansion brought about a marked reduction of this slow wave activity and exsanguination led to recurrence of the slow wave foci. This data demonstrated that colloidal blood volume expansion induced an increase in local CBF and improved neuroelectric status of ischaemic brain following MCA occlusion. It was also shown that a reduction of blood volume and cardiac output resulted in a reduction in local CBF and a worsening of neuroelectric status in ischaemic areas. This study supports the contention that blood volume and cardiac output maintenance is extremely important in the management of acute ischaemic stroke.

Topographic electroencephalographic study with power ratio index mapping in patients with malignant brain tumors.

A variant of electroencephalogram (EEG) power spectral mapping called power ratio index (PRI) mapping was used to monitor 15 patients with malignant brain tumors. This index is generated by dividing the low frequency (delta, theta) power by the high frequency (alpha, beta) power. Because the nonparoxysmal effect of a brain tumor on the EEG is reflected as a relative loss of high frequency power and a gain in low frequency power, utilization of the PRI has the effect of placing the epicenter of the "power dysfunction" coincident with the epicenter of the tumor.