Elsevier

Journal of Critical Care

Volume 36, December 2016, Pages 234-239
Journal of Critical Care

Outcomes/Predictions
Prognostication of critically ill patients with acute-on-chronic liver failure using the Chronic Liver Failure–Sequential Organ Failure Assessment: A Canadian retrospective study,☆☆,,★★

https://doi.org/10.1016/j.jcrc.2016.08.003Get rights and content

Abstract

Purpose

We evaluated the Chronic Liver Failure–Sequential Organ Failure Assessment (CLIF-SOFA) score to predict survival in a Canadian critically ill cohort with acute-on-chronic liver failure.

Methods

We retrospectively examined 274 acute-on-chronic liver failure patients admitted to a quaternary level intensive care unit (ICU) between April 1, 2000, and April 30, 2011. We evaluated severity of illness scores, including the Acute Physiology and Chronic Health Evaluation (APACHE) II, model for end-stage liver disease (MELD), Child-Turcotte-Pugh (CTP), SOFA, and CLIF-SOFA.

Results

On ICU admission, patients had the following median (interquartile range): APACHE II, 23 (19-28); MELD, 26 (19-35); CTP, 12 (10-13); SOFA, 15 (11-18); and CLIF-SOFA, 17 (13-21). In-hospital survival was 40%. There were no significant differences in survival for cirrhosis etiology, reason, or year of admission. The CLIF-SOFA score had the greatest area under receiver operating curve of 0.865 (95% confidence interval, 0.820-0.909) and outperformed the CTP, MELD, SOFA, and APACHE II scores. Sequential Organ Failure Assessment score performance improved on the third day of ICU admission (area under receiver operating curve, 0.935; 95% confidence interval, 0.895-0.975).

Conclusions

The CLIF-SOFA and SOFA scores during the first 3 days of ICU admission appear to be highly predictive of in-hospital mortality.

Introduction

Acute-on-chronic liver failure (ACLF) occurs in patients with previously compensated chronic liver disease and is distinct from chronic decompensation of cirrhosis [1], [2], [3]. Patients with ACLF present with worsening jaundice, coagulopathy, ascites, encephalopathy, and/or multisystem organ failure in the setting of an acute precipitant, such as sepsis, drug ingestion, viral reactivation, recent surgery, or gastrointestinal hemorrhage [4], [5]. Patients with ACLF have high mortality, ranging from 29% to 77% [2], [6], [7], [8].

Several risk factors are associated with higher mortality in critically ill cirrhotic patients. Patients with greater than 3 organ failures, higher fraction of inspired oxygen, higher serum lactate levels, higher serum bilirubin levels, older age, need for vasopressors, lower serum sodium levels, and renal failure have higher associated mortality on multivariate analyses [9], [10], [11], [12]. Many liver-specific and general systems scores have been used to predict mortality of critically ill cirrhotic patients [9], [12], [13], [14], [15], [16], [17]. The Sequential Organ Failure Assessment (SOFA) score has been validated to predict mortality better than other scoring systems, with potentially better prediction at 48 hours after admission [12], [13].

In 2013, the European Association for the Study of the Liver–Chronic Liver Failure Consortium adapted the SOFA score into the Chronic Liver Failure–SOFA (CLIF-SOFA), and graded ACLF into 4 grades [7]. The CLIF-SOFA score evaluates organs over 6 domains similar to the SOFA score, but replaces the platelet count with the international normalized ratio (INR) as the hematologic parameter and replaces the Glasgow Coma Scale (GCS) score with hepatic encephalopathy grade as the neurologic parameter. It has been validated to predict mortality in several populations of patients with ACLF in Europe, Brazil, India, and Southeast Asia [6], [7], [8], [16], [18], [19]. In critically ill patients, the CLIF-SOFA score has been compared with other scoring systems, with good discriminatory ability in cohorts from the United Kingdom and Taiwan [19], [20], [21].

The primary aim of our study was to evaluate the ability of the CLIF-SOFA score to predict mortality in North American critically ill patients with cirrhosis with ACLF. The secondary aim of our study was to identify risk factors or precipitants associated with higher mortality. We hypothesized that the CLIF-SOFA score will be better at discriminating mortality than other liver-specific or general scores in critically ill patients with ACLF.

Section snippets

Materials and methods

The reporting of this study followed the Strengthening of the Reporting of Observational Studies in Epidemiology statement [22]. The local health research and university research ethics boards approved the study and waived the requirement for individual informed consent.

Descriptive characteristics of all patients in the cohort

We identified 1046 potential cases in ICU database. Three hundred four cases met the inclusion criteria and had data collected. After excluding repeated visits to ICU, there were 274 unique admissions between April 1, 2000, and April 30, 2011. A summary of their descriptive characteristics, clinical parameters, and outcomes is presented in Table 1, Table 2.

Cirrhosis was biopsy proven in 20% of cases. The most common risk factors for cirrhosis were alcohol (50%), hepatitis C infection (40%), and

Discussion

We report one of the first studies in North America, examining the CLIF-SOFA score to predict mortality in critically ill cirrhotic patients. We found that both CLIF-SOFA and SOFA scores predicted in-hospital mortality well on the first day of ICU admission. Score performance improved on the third day of ICU admission for both MELD and SOFA scores, with the SOFA outperforming the MELD score. On multivariate analysis, we identified higher serum lactate levels on admission to be associated with

Conclusions

In conclusion, the SOFA and CLIF-SOFA scores in the first 3 days of ICU admission appear to predict in-hospital mortality well in critically ill cirrhotic patients with ACLF. Patients with increasing number of organ failures as defined by the CLIF-SOFA score have significantly increased in-hospital mortality.

The following are the supplementary data related to this article.

. In-hospital and ICU mortality when patients classified by ACLF grade on admission (n = 268 patients, 6 missing data). The

Acknowledgments

We thank Denise Foster, Shelly Fleck, and Sharmin Gani for their support. We also thank Sharon Power and Elfriede Janzen from the Vancouver General Hospital Health Records department. We also thank Dr Donald Griesdale for his assistance.

References (30)

  • PESe Silva et al.

    Single-centre validation of the EASL-CLIF consortium definition of acute-on-chronic liver failure and CLIF-SOFA for prediction of mortality in cirrhosis

    Liver Int

    (2015)
  • E Cholongitas et al.

    Risk factors, sequential organ failure assessment and model for end-stage liver disease scores for predicting short term mortality in cirrhotic patients admitted to intensive care unit

    Aliment Pharmacol Ther

    (2006)
  • D du Cheyron et al.

    The attributable mortality of acute renal failure in critically ill patients with liver cirrhosis

    Intensive Care Med

    (2005)
  • V Das et al.

    Cirrhotic patients in the medical intensive care unit: early prognosis and long-term survival

    Crit Care Med

    (2010)
  • E Cholongitas et al.

    Prognostic models in cirrhotics admitted to intensive care units better predict outcome when assessed at 48 h after admission

    J Gastroenterol Hepatol

    (2008)
  • Cited by (19)

    View all citing articles on Scopus

    Author contributions: E.S. contributed to the study by planning design, collecting data, analyzing data, and drafting the manuscript. J.J.R. contributed to the study by planning design and reviewing the manuscript. R.S. assisted with collection of data and reviewing the manuscript. C.J.K. contributed to the study by planning design, collecting data, analyzing data, and drafting the manuscript. All authors reviewed the final manuscript and contributed to its content.

    ☆☆

    Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors

    Disclosures: The authors declare that there are no conflicts of interest.

    ★★

    Ethics approval: University of British Columbia/Vancouver Coastal Health Research Ethics Board.

    View full text