Incidence and Outcomes for Patients With Cirrhosis Admitted to the United Kingdom Critical Care Units.

OBJECTIVE: To assess the epidemiology and outcome of patients with cirrhosis following critical care unit admission. DESIGN: Retrospective cohort study. SETTING: Critical care units in England, Wales, and Northern Ireland participating in the U.K. Intensive Care National Audit and Research Centre Case Mix Programme. PATIENTS: Thirty-one thousand three hundred sixty-three patients with cirrhosis identified of 1,168,650 total critical care unit admissions (2.7%) admitted to U.K. critical care units between 1998 and 2012. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Ten thousand nine hundred thirty-six patients had alcohol-related liver disease (35%). In total, 1.6% of critical care unit admissions in 1998 had cirrhosis rising to 3.1% in 2012. The crude critical care unit mortality of patients with cirrhosis was 41% in 1998 falling to 31% in 2012 (p < 0.001). Crude hospital mortality fell from 58% to 46% over the study period (p < 0.001). Mean(SD) Acute Physiology and Chronic Health Evaluation II score in 1998 was 20.3 (8.5) and 19.5 (7.1) in 2012. Mean Acute Physiology and Chronic Health Evaluation II score for patients with alcohol-related liver disease in 2012 was 20.6 (7.0) and 19.0 (7.2) for non-alcohol-related liver disease (p < 0.001). In adjusted analysis, alcohol-related liver disease was associated with increased risk of death (odds ratio, 1.51 [95% CI, 1.42-1.62; p < 0.001]) with a year-on-year reduction in hospital mortality (adjusted odds ratio, 0.95/yr, [0.94-0.96, p < 0.001]). CONCLUSIONS: More patients with cirrhosis are being admitted to critical care units but with increasing survival rates. Patients with alcohol-related liver disease have reduced survival rates partly explained by higher levels of organ failure at admission. Patients with cirrhosis and organ failure warrant a trial of organ support and universal prognostic pessimism is not justified.

A multicentre randomized controlled trial of moderate hypothermia to prevent intracranial hypertension in acute liver failure.

BACKGROUND & AIMS: Animal models and human case series of acute liver failure (ALF) suggest moderate hypothermia (MH) to have protective effects against cerebral oedema (CO) development and intracranial hypertension (ICH). However, the optimum temperature for patient management is unknown. In a prospective randomized controlled trial we investigated if maintenance of MH prevented development of ICH in ALF patients at high risk of the complication. METHODS: Patients with ALF, high-grade encephalopathy and intracranial pressure (ICP) monitoring in specialist intensive care units were randomized by sealed envelope to targeted temperature management (TTM) groups of 34°C (MH) or 36°C (control) for a period of 72h. Investigators were not blinded to group assignment. The primary outcome was a sustained elevation in ICP >25mmHg, with secondary outcomes the occurrence of predefined serious adverse effects, magnitude of ICP elevations and cerebral and all-cause hospital mortality (with or without transplantation). RESULTS: Forty-six patients were randomized, of whom forty-three were studied. There was no significant difference between the TTM groups in the primary outcome during the study period (35% vs. 27%, p=0.56), for the MH (n=17) or control (n=26) groups respectively, relative risk 1.31 (95% CI 0.53-3.2). Groups had similar incidence of adverse events and overall mortality (41% vs. 46%, p=0.75). CONCLUSIONS: In patients with ALF at high risk of ICH, MH at 33-34°C did not confer a benefit above management at 36°C in prevention of ICH or in overall survival. This study did not confirm advantage of its prophylactic use. (ISRCTN registration number 74268282; no funding.) LAY SUMMARY: Studies in animals with acute liver failure (ALF) have suggested that cooling (hypothermia) could prevent or limit the development of brain swelling, a dangerous complication of the condition. There is limited data on its effects in humans. In a randomized controlled trial in severely ill patients with ALF we compared the effects of different temperatures and found no benefit on improving survival or preventing brain swelling by controlling temperature at 33-34°C against 36°C.

The impact of organ dysfunction in cirrhosis: survival at a cost?

BACKGROUND & AIMS: The incidence of cirrhosis and subsequent development of organ dysfunction (OD) requiring intensive care unit (ICU) support is rising. Historically, critically ill cirrhotics are perceived as having poor prognosis and substantial cost of care. METHODS: The aim was to prospectively analyse resource utilisation and cost of a large cohort of patients (n=660) admitted to a Liver ICU from 2000 to 2007 with cirrhosis and OD. Child Pugh, MELD, SOFA, APACHE II, and organ support requirements were collected. The Therapeutic Intervention Scoring System (TISS) score, a validated tool for estimating cost in ICU, was calculated daily. Logistic regression was used to determine independent predictors of increased cost. RESULTS: Alcohol was the most common etiology (47%) and variceal bleeding (VB) the most common reason for admission (35%). Invasive ventilatory support was required in 74% of cases, vasopressors in 49%, and 50% required renal replacement therapy. Forty-nine per cent of non-transplanted patients survived to ICU discharge. Median TISS score and ICU cost per patient were 261 and €14,139, respectively. VB patients had the highest survival rates (53% vs. 24%; p<0.001) and lower associated cost. A combination of VB (OR 0.48), need for ventilation (OR 2.81), low PO(2)/FiO(2) on admission (OR 0.97), and lactate (OR 0.93) improved cost prediction on multivariate analysis (AUROC 0.7; p<0.001) but organ failure scores per se were poor predictors of cost. CONCLUSIONS: Patients with cirrhosis and OD result in considerable resource expenditure but have acceptable hospital survival. Further health economic assessment and outcome prediction tools are required to appropriately target resource utilisation.

Increased model for end-stage liver disease score at the time of liver transplant results in prolonged hospitalization and overall intensive care unit costs.

Organ allocation based on Model for End-Stage Liver Disease (MELD) resulted in decreased waiting list mortality in the United States. However, reports suggest an increase in resource utilization as a consequence of this. The aim of this study is to assess the correlation of MELD at transplant with post-liver transplant (LT) intensive care unit (ICU) costs. We assessed clinical and demographic variables of 402 adult patients who underwent LT at King's College Hospital, London, UK, between January 2000 and December 2003. ICU cost calculations were based on the therapeutic intervention scoring system (TISS). Graft quality was assessed using the donor risk index (DRI). Patients with a MELD score > 24 had significantly longer post-LT ICU stay (P < 0.0001) and total post-LT hospital stay (P = 0.008). In addition, they had significantly increased TISS scores, ICU cost, and need for renal replacement therapy (RRT) (P < 0.001). MELD score (by point) and MELD > 24 was associated with prolonged ICU stay (P = 0.004 and P = 0.005, respectively). On univariate analysis, etiology of alcohol-related liver disease (ALD), repeat LT, Budd-Chiari syndrome, and refractory ascites were associated with prolonged ICU stay. Using multivariate analysis, MELD > 24, refractory ascites, ALD and Budd-Chiari syndrome were associated with prolonged ICU stay. There was no association between using grafts with higher DRI and longer ICU stay, need for RRT, increased cost, or hospital survival on univariate analyses (P = not significant). Use of MELD as a method of organ allocation results in significant increase in ICU cost after LT. Using TISS as surrogate marker for ICU costs reveals that the cost implications are related to the need for RRT and prolonged ICU stay.

Hepatic dysfunction in sickle cell disease: a new system of classification based on global assessment.

BACKGROUND & AIMS: Hepatic dysfunction in adults with sickle cell disease varies in character and severity from self-limited cholestasis to life-threatening acute liver failure and cirrhosis. Because previous attempts to describe patterns of liver disease have not reflected clinical experience, we aimed to characterize the presentation, clinicopathologic findings, and natural history of such patients. METHODS: We reviewed the clinical, laboratory, radiographic, and histologic features with the natural history of 38 patients (mean age, 33 years) with Hb SS, SC, or S-beta thalassemia referred to a tertiary liver center for assessment. RESULTS: Distinct disease patterns were identified that comprised massive hepatocellular necrosis (5%), acute severe sequestration and cholestasis in the context of sepsis (18%), cirrhosis (18%), chronic, fluctuating sequestration without cholestasis (21%), mechanical biliary obstruction (8%), siderosis without cirrhosis (8%), generalized cholangiopathy (8%), venous outflow obstruction (3%), and miscellaneous (11%). Of the 20 who required emergency admission, 8 did not survive their index admission, and 3 patients died during follow-up admissions (4 months-4 years later). There were 3 instances of hemorrhage related to liver biopsy. One patient underwent transplantation but died. Hematologic and biochemical markers did not discriminate well between survivors and nonsurvivors. The incidence of a second hepatic pathology (ie, viral hepatitis, autoimmune disease, transfusional siderosis) was 37% and was associated with the finding of more advanced histologic fibrosis. CONCLUSIONS: Patterns of hepatic dysfunction in sickle cell disease are diverse and demand clear characterization for each individual; however, groups with a poor prognosis can be identified after collation of clinical, laboratory, and radiologic data. Findings at biopsy (which is associated with higher risk of bleeding in this group) might be anticipated by noninvasive test results.

Cell-free artificial liver support: design of appropriate clinical studies.

The clinical use of cell-free liver support devices is dependent upon proof of safety and efficacy in clinical trials. The current published data to support their use is limited in both quantity and quality. In most studies, the methodology is such that the effects of these devices are difficult to establish with certainty. Bias might be introduced by the use of uncontrolled studies, and in randomized controlled trials limited size, poorly matched control groups or medical therapy or the use of clinically inappropriate endpoints might be important. Guidelines are now available to provide a framework for the design and execution of such trials, specifically to minimize the systematic errors that are frequently present and that might result in biased estimates of treatment effects. In the present review, the limitations of current studies are discussed and suggestions made as to the design and conduct of future clinical trials.