Article Text

Download PDFPDF
P64 CyberLiver Animal Recognition Test (CL-ART): a novel remote monitoring tool to assess hepatic encephalopathy
  1. Kohilan Gananandan1,
  2. Ahmed El Shabrawi1,
  3. Mapi Pilar Ballester1,
  4. Konstantin Kazankov1,2,
  5. Karen Louise Thomsen1,2,
  6. Raj Mookerjee1,2,
  7. Anu Balaji3,
  8. Ravan Boddu3,
  9. Ravi Kumar3
  1. 1University College London, Institute for Liver and Digestive Health, UK
  2. 2Aarhus University Hospital, Department of Hepatology and Gastroenterology, Aarhus, Denmark
  3. 3CyberLiver Limited, UK


Introduction The CyberLiver-Animal-Recognition-Test (CL-ART) is a novel smartphone application (app) that was developed as part of a program for remote monitoring of cirrhosis, and specifically, remote diagnosis of hepatic encephalopathy (HE) and to guide need for intervention. This study compared CL-ART with clinical standard tests for assessing early HE, to determine its diagnostic performance, and to establish an optimal CL-ART threshold for diagnosis of HE with remote monitoring.

Methods We conducted a prospective study in cirrhosis patients and in healthy controls, applying clinical assessment and three different cognitive function tests to determine utility of CL-ART (Ethics reference code – 20/HRA/3843). The CL-ART involved an easy, timed recognition and appropriate naming of animals using a smartphone app, with time to test completion and number(s) of failed attempts logged. EncephalApp Stroop Test and Psychometric Hepatic Encephalopathy Score (PHES) were chosen as test comparisons. Pearson rank correlation and area under the receiver operator curves (AUROC) were calculated.

Results 45 cirrhosis patients, deemed at risk of decompensation (62% male; median 56 years, Child-Pugh Score 8 [IQR 7–9], MELD-Na 15 [IQR 11–19], UKELD 53 [IQR 49–56], CLIF-AD score 47 [IQR 40–52] and 17 controls with no clinical/biochemical evidence of liver disease (53% male; median 58 years), were included. Compared to controls, cirrhosis patients performed significantly worse in CL-ART (24s [IQR 18–27] vs 16s [IQR 13–18], p<0.001); EncephalApp (206s [IQR 176–271] vs 154s [132–165], p<0.001) and PHES (-3 (IQR -6/-2] vs 0[0–1], p<0.001). CL-ART showed a good correlation with both EncephalApp (r=0.796, p<0.001, [figure 1A]) and PHES tests (r=-0.668, p<0.001, [Figure 1B]). Using PHES and clinical assessment, 16 patients (35.5%) had an HE diagnosis. Mean CL-ART times were significantly higher in patients with HE (27s vs 22s, p<0.001). AUROC to diagnose HE was 0.84 (95% CI=0.72–0.96) for CL-ART. A cut-off of >23 seconds in CL-ART time showed a sensitivity of 81% and specificity of 62%, to diagnose HE.

Abstract P64 Figure 1

A) Scatter plot showing correlation between CL-ART and EncephalApp results for each patient; B) Scatter plot showing correlation between CL-ART and PHES results for each patient

Conclusions CL-ART is a novel smartphone app, enabling remote assessment of HE. It shows good comparability with other standard tests and high sensitivity for diagnostic performance in identifying HE. Given its ease-of-use, rapid testing (usually <30s) and smartphone application, it provides a convenient tool for remote, sustainable management of cirrhosis.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.