Inhibitory Control Test for the Diagnosis of Minimal Hepatic Encephalopathy

Transcription

1 GASTROENTEROLOGY 2008;135: Inhibitory Control Test for the Diagnosis of Minimal Hepatic Encephalopathy JASMOHAN S. BAJAJ,* MUHAMMAD HAFEEZULLAH, JOSE FRANCO, RAJIV R. VARMA, RAYMOND G. HOFFMANN, JOSHUA F. KNOX, DARRELL HISCHKE, THOMAS A. HAMMEKE, STEVEN D. PINKERTON, and KIA SAEIAN *Division of Gastroenterology, Hepatology, and Nutrition, Virginia Commonwealth University and McGuire VA Medical Center, Richmond, Virginia; Division of Gastroenterology and Hepatology, Department of Population Health/Division of Biostatistics, Division of Neuropsychology, Department of Psychiatry and Behavioral Medicine, Medical College of Wisconsin, Milwaukee, Wisconsin See editorial on page Background & Aims: Minimal hepatic encephalopathy (MHE) is difficult to diagnose. The Inhibitory Control Test (ICT) measures response inhibition and has diagnosed MHE with 90% sensitivity and specificity in a selected population; high lure and low target rates indicated poor ICT performance. We studied the reliability and validity of ICT for MHE diagnosis. Methods: ICT was compared with a psychometric battery (standard psychometric tests [SPT]) for MHE diagnosis and overt hepatic encephalopathy (OHE) prediction. ICT was administered twice for test-retest reliability, before/after transvenous intrahepatic portosystemic shunting (TIPS), and before/after yogurt treatment. The time taken by 2 medical assistants (MA) to administer ICT was recorded and compared with that of a psychologist for cost analysis. Results: One hundred thirty-six cirrhotic patients and 116 age/education-matched controls were studied. ICT (>5 lures) had 88% sensitivity for MHE diagnosis with area under the curve for receiver operating characteristic. MHE-positive patients had significantly higher ICT lures (11 vs 4, respectively, P.0001) and lower targets (92% vs 97%, respectively, P.0001) compared with MHEnegative patients. The test/retest reliability for ICT lures (n 50, r 0.90, P.0001) was high. ICT and SPT were equivalent in predicting OHE (21%). ICT lures significantly worsened after TIPS (n 10; 5 vs 9, respectively; P.02) and improved after yogurt supplementation (n 18, 10 vs 5, respectively; P.002). The MAs were successfully trained to administer ICT; the time required for test administration and the associated costs were smaller for ICT than for SPT. Conclusions: ICT is a sensitive, reliable, and valid test for MHE diagnosis that can be administered inexpensively by MAs. Minimal hepatic encephalopathy (MHE) is a neurocognitive complication of cirrhosis, which has been reported in up to 80% of tested patients. 1 3 It is associated with increased progression to overt hepatic encephalopathy (OHE), diminished quality of life, and driving impairment as well as a high risk of traffic accidents. 4 8 MHE results in a specific spectrum of cognitive deficits, especially in the domains of attention, vigilance, response inhibition, and executive function Psychometric tests for MHE diagnosis specifically assess these cognitive domains. 12 The difficulty in applying these psychometric tests in the United States is the need for psychologic expertise for clinical administration and interpretation, the expensive and copyrighted testing procedures, and the lack of reimbursement by insurance companies. These issues were highlighted in a recent survey of the American Association for the Study of Liver Diseases (AASLD) in which, although most respondents believed MHE to be a serious problem that required testing, only a minority reported doing so. 13 Most respondents believed that a simple test that could be administered by clinic personnel would increase their testing rates. 13 The Inhibitory Control Test (ICT) is a computerized test of attention and response inhibition that has been used to characterize attention deficit disorder, schizophrenia, and traumatic brain injury. 10,14 17 ICT is similar to the continuous performance test and consists of presentation of several letters at 500-ms intervals. 18 Interspersed within these letters are the letters X and Y. The subject is instructed to respond to every X and Y during the initial part of the training run, which establishes the prepotent response. In the latter part of the training run, the subject is instructed only to respond when X and Y are alternating (called targets) and to inhibit from responding when X and Y are not alternating (called lures). After the training run, 6 test runs, which last approximately 2 minutes each, are administered with a total of 40 lures, 212 targets, and 1728 random letters in be- Abbreviations used in this paper: MHE, cirrhotic patients with minimal hepatic encephalopathy; MHE, cirrhotic patients without minimal hepatic encephalopathy; OHE, overt hepatic encephalopathy; SPT, standard psychometric tests; ICT, inhibitory control test; NCT-A, number connection test; DST, digit symbol; BDT, block design test; TIPS, intrahepatic portosystemic shunting by the AGA Institute /08/$34.00 doi: /j.gastro

2 1592 BAJAJ ET AL GASTROENTEROLOGY Vol. 135, No. 5 tween. At the end of the test, the lure and target response rates, lures, and target reaction times are automatically calculated. Lower lure response, higher target response, and shorter lure and target reaction times indicate good psychometric performance. ICT has been used for the diagnosis of MHE by our group in 50 nonalcoholic cirrhotic patients compared with 50 age- and education-matched controls with 90% sensitivity and specificity. 19 ICT was similar to the battery of standard psychometric tests (SPT) for prediction of OHE development and had good test-retest reliability. This was an initial report in highly selected and motivated patients and excluded all patients with an alcoholic etiology of cirrhosis. Furthermore, external validation was not performed in that initial study. 19 The aims of this present study were to (1) validate the ICT for the diagnosis of MHE in cirrhotic patients by comparing them with age- and education-matched controls; (2) assess its test-retest reliability; (3) determine its external validity by studying ICT before and after transvenous intrahepatic portosystemic shunting (TIPS), before and after yogurt supplementation, and its ability to predict onset of OHE; and (4) estimate the time and cost required for previously untrained medical assistants to administer the test in the hepatology clinic. Patients and Methods Outpatients with cirrhosis were recruited between September 2004 and December 2007 through clinical referral. Patients between ages 18 and 65 years, without current psychoactive drug use, not on OHE treatments, without history of OHE, and without alcohol use within 3 months were included. MHE Definition and SPT SPT battery for this study was the Number Connection Test-A (NCT-A) and 2 subtests of the Wechsler s Adult Intelligence Scale-III: Digit Symbol Test (DST) and Block Design (BDT) tests. 20 MHE was defined as a performance impaired 2 standard deviations beyond that of the age/education-matched controls on any test. This is modified from the Working Group on Hepatic Encephalopathy recommendation, which has previously been used to diagnose MHE and predict OHE and driving impairment in our population. 12,19,21 The study was divided into the following parts: (1) Cross-sectional study of SPT vs ICT for MHE diagnosis. (2) Test-retest reliability of SPT and ICT. (3) Follow-up of cirrhotic patients to detect OHE development. (4) External validity: (a) study change in ICT and SPT after TIPS placement and (b) study change in ICT and SPT after probiotic yogurt supplementation. (5) Time and cost analysis of ICT administration in hepatology clinics. For study parts 1, 2, and 3, a total of 256 patients without OHE or without current OHE therapy were considered for recruitment; 65 refused to participate, 33 had concurrent psychoactive drug use (antianxiety, antipsychotics, antiepileptics, and interferon), and 23 had consumed alcohol within 3 months. As a result, 135 cirrhotic patients were included in the study. Age- and educational status-matched healthy controls were recruited from the community through advertisements. Written informed consent was obtained from each participant. MHE Diagnosis With ICT vs SPT All patients and controls were administered the SPT and ICT during the same sitting under the supervision of a psychologist who was blinded to the nature of the participants. Parameters recorded were NCT-A completion time in seconds, raw score of DST and BDT, ICT lure and target response rates, and ICT lure and target reaction times. Control and cirrhotic groups were compared with respect to demographics and psychometric test results. Correlation between age with individual ICT and SPT results were performed for both groups. Cirrhotic patients were divided into those with MHE (MHE ) and those without MHE (MHE ) using SPT as the gold standard. Sensitivity and specificity of diagnosis of MHE using ICT data were compared using the receiver operating characteristic and area under the curve and logistic regression. A subgroup analysis of psychometric tests was performed within the cirrhotic group for hepatitis C-positive cirrhotic patients, those with alcoholic cirrhosis, and those with both. Test-Retest Reliability A randomly selected group of controls and cirrhotic patients underwent SPT and ICT twice at least 30 days apart to gauge the test-retest reliability. There was no change in the clinical status (Child Pugh score), addition of psychoactive medications, or development of new complications of cirrhosis (variceal bleeding, ascites, and others) in between the 2 testing times. Spearman rank correlation was used to correlate the NCT-A, BDT, DST, and ICT lure and target response of the first and second administration. Follow-Up to Evaluate OHE Development Patients who were enrolled in the cross-sectional study were followed up as a matter of clinical practice by the primary hepatologists for the detection of OHE. OHE was diagnosed when the patient was started on therapy and the precipitating factor was recorded. Development of OHE was compared between patients with MHE according to SPT vs ICT. Patients with discordant or concordant results on initial testing were compared. External Validity SPT and ICT changes after TIPS placement. Consecutive outpatient cirrhotic patients who underwent elective TIPS placement were studied with SPT and ICT

3 November 2008 DIAGNOSIS OF MINIMAL HEPATIC ENCEPHALOPATHY 1593 at least 30 days prior to and 30 days post-tips. Patients already on OHE therapy, those with psychoactive medications, and those who underwent emergent TIPS were excluded from this study. There was no primary prophylaxis protocol against OHE post-tips in place. Patients were reexamined post-tips for OHE development and chart review to evaluate for OHE development before that visit was performed. TIPS stent characteristics and reduction in hepatic portal venous pressure gradient were recorded. ICT and SPT parameters before and after TIPS were compared using paired t tests. SPT and ICT changes after probiotic yogurt supplementation in a randomized, controlled trial. Consecutive MHE patients were enrolled for a randomized, controlled trial of probiotic yogurt for MHE treatment. 22 All patients fulfilled the overall inclusion and exclusion criteria of the cross-sectional study and underwent SPT and ICT at baseline. Patients were randomized in a 2:1 ratio into the yogurt and the no-treatment group and were followed for 60 days, at the end of which SPT and ICT were both repeated at the same sitting by a scorer blinded to the randomization scheme. Paired t test within groups and 2 sample t tests were used to compare continuous variables, and Fisher exact test was used to compare binary variables between groups. Time and Cost Analysis of ICT Administration in the Clinic To assess the time and costs associated with ICT administration in a real-world clinic setting, 2 medical assistants (MAs; one 31 years old and other 64 years old; both women) were trained to administer ICT on a laptop computer in the hepatology clinic. After training on the computer for 30 minutes each, they were asked to administer the test to the principal investigator (J.S.B.) to evaluate their ability to administer this test. They then administered the test to 10 subjects each (5 cirrhotic patients and 5 controls) to ensure proficiency in test administration. After adequate training of the MAs, the time required for administration of ICT was recorded for the next 20 patients. The time required for administration of the standard psychometric battery by the psychologist was also determined for the same patients. Current procedural terminology codes for a MA assessment (E and M 1; 99211) and for neuropsychologic testing by a psychologist (96118) were used to estimate third-party charges and Medicare reimbursement. 23 Fixed costs for the copyrighted Wechsler s Adult Intelligence Scale-III instrument and materials and for computer time for ICT were excluded from the analyses because these costs are likely to be negligible when averaged over large numbers of test administrations. Results Cross-Sectional Study of the SPT and ICT for Diagnosing MHE Within the cirrhotic patients, there were 8 patients between 18 and 35 years, 67 between 36 and 55 years, and 48 patients between 56 and 65 years of age. Of the 116 controls, 23 were between 18 and 35 years, 66 between 36 and 55 years, and 27 between 56 and 65 years of age. Control and cirrhosis group results and MHE definition. There was no significant difference in demographics (Table 1). The cirrhotic group was significantly impaired with respect to SPT performance compared with controls. A patient with cirrhosis was considered to be MHE if any of the following conditions were met: NCT-A 35 seconds, BDT raw score 28, or DST raw score 66. Similarly, there was a significantly higher ICT Table 1. Comparison Between Controls and Patients With Cirrhosis Controls (n 116) Cirrhotic patients (n 135) P values Age (y) Sex (M/F) 63/53 72/63.88 Education (y) Etiology of cirrhosis (HCV/alcohol/HCV alcohol/others) 45/19/13/42 NCT-A (seconds) DST (raw score) BDT (raw score) Response to ICT lures (number out of 40) Response to ICT targets (percentage responded to) ICT lure reaction time (milliseconds) ICT target reaction time (milliseconds) NOTE. Controls were age- and educational status-matched with cirrhotic patients; sex distribution was similar between the groups. Others within the cirrhotic group included 14 patients with nonalcoholic steatohepatitis, 11 with primary sclerosing cholangitis, 9 with primary biliary cirrhosis, and the remainder with cryptogenic cirrhosis. There was a significantly impaired performance in each SPT test in cirrhotic patients compared with controls. A significantly worse ICT psychometric performance, ie, higher lure response rate, lower target response rate, and higher lure and target response time, was observed. HCV, hepatitis C only; SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test.

4 1594 BAJAJ ET AL GASTROENTEROLOGY Vol. 135, No. 5 Table 2. Comparison Between Patients With Cirrhosis With and Without Minimal Hepatic Encephalopathy MHE (n 87) MHE (n 48) P value Age (y) Sex (M/F) 46/41 26/22.89 Education (y) Child Pugh score distribution 58/26/3 41/7/0.05 (A/B/C) Chronic hepatitis C (%) 41 (47) 17 (35).19 Alcoholic cirrhosis (%) 22 (25) 10 (21).56 NCT-A (seconds) DST (raw score) BDT (raw score) Response to ICT lures (number out of 40) Response to ICT targets (percentage responded to) ICT lure reaction time (milliseconds) ICT target reaction time (milliseconds) NOTE. A cirrhotic patient was considered to be MHE if any of the following conditions were met: NCT-A 35 seconds, BDT raw score 28, or DST raw score 66. There was no significant difference in age, educational status, sex distribution, Child Pugh score, chronic hepatitis C, and alcoholic cirrhosis prevalence between cirrhotic patients with and without MHE. There was a significantly impaired performance in each SPT test in MHE cirrhotic patients compared with MHE cirrhotic patients. A significantly worse ICT psychometric performance, ie, higher lure response rate, lower target response rate, and higher lure and target response time, was observed. MHE, cirrhotic patients with minimal hepatic encephalopathy; MHE, cirrhotic patients without minimal hepatic encephalopathy; SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test. The MHE group had a significantly higher rate of lure response and longer lure and target reaction times. MHE patients also demonstrated a significantly lower target response compared with MHE (Table 2) patients. Sensitivity and specificity of ICT using SPT as the gold standard. A cut off of 5 lures per person was used to diagnose MHE based on receiver operating characteristic analysis and our prior study. 19 Of the 87 MHE patients by SPT, 76 had 5 lures, and 11 did not. Of the 48 MHE patients by SPT, 37 had 5 lures, and 11 had 5 lures. Sensitivity was 87%, and specificity was 77% for the diagnosis of MHE using ICT lure threshold of 5 lures (Figure 1A). The receiver operating characteristic curve had an area under the curve of for MHE diagnosis using lures (Figure 1B). lure response, a significantly lower ICT target response, and a longer target reaction time in the cirrhotic group compared with controls. Comparison within the cirrhotic group. Using the a priori definition, 87 patients were considered to be MHE, and the remaining 48 were MHE (Table 2). Therefore, 48 patients were not impaired in any test. Of the 87 MHE patients, 25 were impaired on all 3 tests, 24 patients on 2 tests, and 38 patients on 1 psychometric test. A significant positive correlation between number of impaired tests on the SPT and lure response (r 0.51, P.0001) and a significant negative correlation of targets with number of impaired tests on SPT (r 0.39, P.0001) was observed. There was also a significant negative correlation between ICT lures and targets (r 0.37, P.0001). There was no significant difference in demographics, prevalence of hepatitis C, or alcoholic etiology of cirrhosis between MHE and MHE groups. A borderline statistically significant predominance of Child Pugh B and C class cirrhotic patients in the MHE group was observed (P.05). Figure 1. (A) The individual value plot of lures in cirrhotic patients with and without MHE shows 88% sensitivity and 77% specificity for the diagnosis of MHE using a cut off of 5 lures and above. SPT was used as the gold standard for the diagnosis of MHE. The total number of lures in the ICT is 40. MHE, minimal hepatic encephalopathy; SPT, standard psychometric battery. (B) The receiver operating characteristic curve for the ICT lure response for the diagnosis of MHE using SPT as a gold standard showed that an area under the curve was There was no significant contribution of ICT targets, lure, and target reaction time to this area under the curve on logistic regression. MHE, minimal hepatic encephalopathy; SPT, standard psychometric battery.

5 November 2008 DIAGNOSIS OF MINIMAL HEPATIC ENCEPHALOPATHY 1595 Table 3. Comparison Between Patients With Cirrhosis With and Without Chronic Hepatitis C as the Etiology of Cirrhosis HCV positive (n 58) HCV negative (n 77) P value Age (y) NCT-A (seconds) DST (raw score) BDT (raw score) Response to ICT lures (number out of 40) Response to ICT targets (%) ICT lure reaction time (milliseconds) ICT target reaction time (milliseconds) NOTE. There was no significant difference in age, individual SPT performance, and all aspects of ICT performance, ie, lure response rate, target response rate, and lure and target reaction times, between cirrhotic patients with and without HCV infection (included 13 patients with HCV and alcohol). SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test. Correlation between SPT and ICT parameters within the cirrhotic group. There was a significant correlation between ICT lures and targets with SPT performance. ICT lures were positively correlated with NCT-A performance (r 0.41, P.0001) and negatively correlated with BDT (r 0.40, P.0001) and DST (r 0.36, P.0001). ICT targets were significantly negatively correlated with NCT-A (r 0.29, P.001) and positively correlated with BDT (r 0.29, P.001) and DST (r 0.41, P.001). Correlation of age with SPT and ICT results. There was a significant positive correlation between age and NCT-A (r 0.425, P.0001) in controls. BDT (r 0.437, P.0001) and DST (r 0.417, P.0001) demonstrated a corresponding significant negative correlation with age in controls. ICT lures were not significantly correlated with age in controls (r 0.121, P.23). Similar to the control group, there was a significant negative correlation between BDT (r 0.231, P.01) and age in cirrhotic patients. However, there was no significant correlation in the cirrhotic group between NCT-A (r 0.204, P.79) and DST (r 0.13, P.15) scores and age. ICT lures continued to demonstrate no significant correlation with age in the cirrhotic patients (r 0.06, P.51). Because some of the metrics showed age effects, cirrhotic patients were age matched to controls. Hepatitis C, alcoholic cirrhosis, and both subgroup analysis. Within the cirrhotic patients, 58 patients had chronic hepatitis C, and 77 did not. There was no statistically significant difference in SPT or ICT parameters between those with or without hepatitis C groups (Table 3). Similarly, no significant difference in ICT or SPT parameters was observed between those with an alcoholic etiology of cirrhosis compared with cirrhotic patients without alcohol as the etiology (Table 4). Thirteen patients had both alcohol- and chronic hepatitis C-associated cirrhosis; the performance on SPT (NCT-A: 36 8 seconds, BDT: 33 9, DST: 57 12) and ICT (targets 92 7 and lures 7 7) were statistically similar to patients with chronic hepatitis C and those with alcoholic cirrhosis. Test-Retest Reliability Results A randomly selected group of 20 controls and 30 cirrhotic patients from the patients enrolled above were administered the SPT and ICT days apart. No change in the clinical status of the patients and controls was observed in the interim. The 30 cirrhotic patients included 21 with hepatitis C, 5 with alcoholic cirrhosis, 3 with cryptogenic cirrhosis, and 1 with primary biliary cirrhosis; all were Child Pugh class A. Both ICT and SPT parameters had a strong test-retest correlation in cirrhotic patients as well as controls (Table 5). None of the test-retest correlations differed significantly between the controls and the cirrhotic patients. Table 4. Comparison Between Patients With Cirrhosis With and Without Alcohol as the Etiology of Cirrhosis Alcoholic etiology (n 32) Nonalcoholic etiology (n 103) P value Age (y) NCT-A (seconds) DST (raw score) BDT (raw score) Response to ICT lures (number out of 40) Response to ICT targets (%) ICT lure reaction time (milliseconds) ICT target reaction time (milliseconds) NOTE. All included patients had not consumed alcohol within 3 months before the study. There was no significant difference in age, individual SPT performance, and all aspects of ICT performance, ie, lure response rate, target response rate, and lure and target reaction times between cirrhotic patients with and without history of alcoholic liver disease (included 13 patients with both HCV and alcoholic etiology of cirrhosis). SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test.

6 1596 BAJAJ ET AL GASTROENTEROLOGY Vol. 135, No. 5 Table 5. Test-Retest Reliability of ICT and SPT Performance Controls (n 20) Correlation coefficient OHE Development P value Cirrhotic patients (n 30) Correlation coefficient P value NCT-A (seconds) DST (raw score) BDT (raw score) Response to ICT lures Response to ICT targets NOTE. There was a significant correlation between the first and second administration of SPT components and ICT performance in controls and cirrhotic patients. SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test. All patients (apart from 15 who refused) were followed up for a mean of months since first enrollment for OHE development. Nineteen patients were diagnosed to have OHE; OHE was diagnosed 9 6 months after enrollment. Of the 19 patients, 6 developed OHE after acute variceal bleeding, 7 after infections requiring hospital admission, 3 after TIPS, and the remaining 3 diagnosed after undergoing an operation. All 19 patients had been diagnosed with MHE using both ICT and SPT. None of the 22 patients with discordant test results (11 who were positive for MHE using SPT but not ICT and 11 who were positive for MHE using ICT but not SPT) developed OHE on follow-up. There was no statistically significant difference in the SPT or ICT parameters of MHE patients who developed OHE compared with those who did not. External Validity Study Results Testing before and after TIPS. Fifteen consecutive patients undergoing elective TIPS placement within February 2007 and December 2007 were approached, of which 3 were on psychoactive medications and 2 refused to participate. The remaining 10 patients were initially tested 26 5 days prior to the TIPS placement. The patients underwent TIPS for refractory ascites (n 6), hepatic hydrothorax (n 2), and for control of chronic hemorrhage from rectal and stomal varices (n 2). Covered TIPS (Viatorr stent, Flagstaff, AZ) placement was successful in all patients, and a mean hepatic portal venous pressure gradient reduction of mm Hg post-tips was achieved. All were discharged home in a satisfactory condition. Three patients (2 underwent TIPS for refractory ascites and 1 for chronic bleeding) were noted to have developed OHE by clinical examination by their primary hepatologist. All 3 were started on lactulose and underwent the second visit testing while they were on lactulose treatment. The second visit with SPT and ICT testing occurred 35 8 days post-tips. There was a significant increase in lure response on the ICT post-tips with or without inclusion of the 3 post-tips OHE patients (Supplementary Figure 1; see Supplementary material online at www. gastrojournal.org and Table 6). ICT target response rate significantly worsened post-tips only when all patients were included. Similarly NCT-A performance was significantly affected only when all patients were considered. BDT and DST scores were impaired after TIPS placement, but this did not reach statistical significance (Table 6). Probiotic yogurt trial ICT results. Twenty-five patients were enrolled: 17 were randomized into the yogurt group and 8 into the no-treatment group. Three yogurt group patients did not complete the study (2 stopped it because of palatability issues, and 1 died of an unrelated Pseudomonas aeruginosa septicemia after spending time in a public hot tub). Two patients in the no-treatment group developed OHE in the 60-day study period. Therefore, 14 yogurt-assigned and 6 no-treatment-assigned patients completed the study. 22 There was significant improvement in psychometric test performance across all SPT tests in the yogurt group (Table 7). 22 Paralleling this was a statistically significant improvement in ICT lure and ICT target performance Table 6. Psychometric Performance Before and After Tips Before TIPS (n 10) Post-TIPS including OHE (n 10) Post-TIPS without OHE (n 7) Paired t test of pre- and post-tips patients (including OHE) (n 10) Paired t test of pre- and post-tips (excluding OHE) (n 7) NCT-A (seconds) DST (raw score) BDT (raw score) ICT lures ICT targets NOTE. Three patients developed OHE after TIPS and were tested while they were on lactulose therapy. There was a trend toward significant impairment of tests of the SPT after TIPS. ICT lure response rate increased and ICT target response rate decreased significantly after TIPS, indicating an impairment of ICT performance compared with before TIPS. TIPS, transvenous intrahepatic portosystemic shunting; SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test.

7 November 2008 DIAGNOSIS OF MINIMAL HEPATIC ENCEPHALOPATHY 1597 Table 7. Psychometric Performance Before and After Probiotic Yogurt in a Randomized Controlled Trial Yogurt baseline (n 17) No-Rx baseline (n 8) Yogurt end (n 14) No-Rx end (n 6) Yogurt within group, P values No-Rx within group, P values NCT-A (seconds) DST (raw score) BDT (raw score) ICT lures ICT targets NOTE. Probiotic yogurt therapy resulted in significant improvement in psychometric function in all SPT components in the probiotic yogurtrandomized group only. Similarly there was a significant improvement in ICT performance (increased ICT lure response and decreased ICT target response rates). There was no significant change in SPT or ICT in the no-treatment randomized group. No-Rx, no treatment; SPT, standard psychometric tests; ICT, Inhibitory Control Test; NCT-A, Number Connection Test; DST, Digit Symbol Test; BDT, Block Design Test. within the yogurt group (Supplementary Figure 2; see Supplementary material online at org). In contrast, there was no improvement in SPT or ICT parameters within the no-treatment group. 22 Time and Cost Analysis of ICT Administration in the Clinic The average time required for the medical assistants to administer the ICT in the clinic was 14 3 minutes compared with 32 7 minutes for SPT administration by a trained psychologist. Medicare reimbursement for more than 30 and less than 60 minutes of a psychologist s time to administer the SPT is $112, and the third-party billing charge is $348. In contrast, for ICT, an MA s time would cost $18 for Medicare and $41 for third-party insurance. The Institutional Review Board at the Medical College of Wisconsin approved this protocol. Discussion The current study demonstrates that the ICT is simple to administer and has a high sensitivity, area under the curve for diagnosis, and test-retest reliability for the diagnosis of MHE compared with SPT. ICT has external validity for MHE because it predicts OHE development, improves after successful MHE therapy, and worsens after TIPS placement. This study also demonstrates that ICT can be administered in clinics by MAs after a single training session, which makes ICT a less expensive method for MHE diagnosis than SPT. ICT measures response inhibition and attention, 2 basic cognitive domains that are affected in MHE. 24 The outcomes of ICT (lures, targets, and reaction times) provide measures of separate but complementary aspects of impairment in MHE. Lure response is an act of commission, signifying a defect in response inhibition. 18 Response inhibition is an essential aspect of executive function in the brain, which controls an individual s ability to inhibit a response that is perhaps prepotent but incorrect in a circumstance. 11,25 In the current study, all subjects were instructed to avoid responding to lures in the training session. Despite this, patients with MHE responded to a significantly higher number of lures compared with patients without MHE and healthy controls. Response inhibition underlies each action that is performed by the subject, and its impairment can be inherently responsible for potentiating wrong decisions in psychometric testing and in daily life, such as during driving. On the ICT, errors of omission produce a significantly lower target detection rate in MHE patients compared with controls and patients without MHE. These errors are considered primary errors of attention in that they represent a lapse in focus. Errors of omission along with longer lure and target reaction times are commonly associated with diminished processing speed and impairments in visuomotor function. 16 It is likely that MHE affects attention circuits that lead to functional impairments in these patients. 26 In addition, ICT lures response was also negatively correlated with ICT target performances, indicating that both these aspects of attention deficits coexist. Undoubtedly, the effects of these attention deficits on MHE patients performance in daily life are compounded when combined with impaired response inhibition. The standard battery used for the diagnosis for MHE in this study consists of 3 tests that evaluate the attention and visuomotor coordination. Previous studies have demonstrated a significantly worse performance of MHE patients on similar batteries and have recommended these tests for MHE diagnosis. 12 NCT-A or trail-making test involves connecting numbers from 1 to 25 accurately while being timed; it is a test of visual scanning speed, visual attention, and psychomotor tracking. DST, in which a subject has to transcribe nonsense symbols paired with numbers after referring to a key, is a test of mental speed and visuomotor coordination and memory. BDT is a test of visuospatial reasoning that requires spatial orientation, manipulation, and problem solving to rapidly reproduce designs shown to them using blocks. ICT impairment, for both lures and targets, were significantly correlated with impairment in NCT-A, DST, and BDT and also with the number of tests that were impaired. Because all SPT tests rely on accurate visual and attention processing, therefore, it is to be expected

8 1598 BAJAJ ET AL GASTROENTEROLOGY Vol. 135, No. 5 that the ICT, which is also a test of attention and processing speed, would correlate with performances on these tests. As shown in prior studies, performances on NCT-A, BDT, and DST significantly worsen with increasing age, a finding also demonstrated in our study. 9,27 To account for this, cirrhotic patients were matched on age to controls. In sharp contrast, ICT lure response rates do not significantly change with increasing age. This is an important distinction between SPT and ICT because the norms of SPT in most populations are age dependent, whereas ICT performance in controls and cirrhotic patients are not related to age. This potentially broadens and simplifies the applicability of ICT in the clinic setting. There are several confounding variables that can affect the diagnosis of MHE. Chronic hepatitis C infection has been proposed to result in neurocognitive deficits even before the development of cirrhosis. 28 In this study, there was no significant difference in SPT or ICT parameters patients with or without chronic hepatitis C infection. This has been similar to our prior published experience with ICT and shows that, in our population, the effect of cirrhosis probably outweighs the effect of chronic hepatitis C infection. Similarly, there was no significant difference in the psychometric performance on SPT and ICT in patients with alcoholic vs nonalcoholic etiology of cirrhosis. This is important to establish because the initial ICT study specifically excluded those with alcoholic cirrhosis, even those with proven alcohol abstinence. These findings further increase the applicability of the ICT across a broad range of prevalent chronic liver diseases. Both inter-/intraobserver reliability and test-retest reliability are important components of the overall validation for a diagnostic test. ICT is a patient-administered instrument that has an automated, computerized analysis system. Therefore inter-/intraobserver reliability is not important. There was a high correlation of ICT lures and targets between repeated administrations, indicating excellent test-retest reliability. This is essential because psychometric performance is an end point of MHE trials, and differentiating between learning and therapeutic effect of an MHE intervention is important. 5,29 The good test-retest reliability of ICT lures would make it a good test for application in trials. This study also addressed the external validity of the ICT by evaluating the ability of ICT to predict OHE and define post-tips and post-mhe therapy alterations in psychometric performance. Patients with MHE have a significantly higher rate of developing OHE, the detection of which is part of routine care. 4 Sixteen percent of included patients developed OHE in this study. Extending our previous experience with ICT, all OHE patients had been diagnosed with MHE using both ICT and SPT. 19 However, there was no significant difference in ICT lures or SPT individual performance between the patients who develop OHE and those who did not. TIPS is an excellent model for testing external validity of ICT because it can adversely affect psychometric function It has been hypothesized that patients with TIPS have a combination of portal hypoperfusion and increased availability of nitrogenous compounds because of the shunting. This results in worsening psychometric function and possibly OHE in 5% 35% of patients. 32,33 In this study, there was a significant worsening of SPT performance 1 month after TIPS placement compared with the pre-tips performance. Importantly, ICT lures worsened after TIPS placement in all but 1 patient. When the 3 patients without OHE after TIPS were excluded, ICT lures were the only psychometric test that was significantly impaired after TIPS placement. Although the number of patients was limited, this finding demonstrates that ICT performance worsens after TIPS. This is an important facet of external validity for ICT in patients with cirrhosis. Treatment for MHE is an attainable goal because gutbased therapy with lactulose, probiotics, and prebiotics has demonstrated improvement in psychometric functioning and quality of life. 34,35 Our group performed a randomized controlled trial of a probiotic yogurt vs no treatment in nonalcoholic cirrhotic patients over 60 days. 22 There was a significant improvement in individual tests of SPT in the yogurt group but not in the notreatment group. Patients who were randomized to yogurt in parallel to SPT improvement also had improvement in ICT lure response rate. In the no-treatment group, similar to the SPT, ICT lure response did not change. 22 This finding further establishes the external validity of ICT as a test that changes with change in the clinical status. The AASLD survey highlighted that a simple test that can be administered by clinic personnel would increase the chances of MHE testing in clinic. 13 In the present study, 2 previously untrained MAs were trained to administer the ICT in the hepatology clinic, and the time required to administer the test was compared with the SPT. Both administration time and associated costs were much smaller for ICT than for SPT. In contrast to SPT, which requires an additional appointment with a psychology specialist, ICT can be administered in the clinic by MAs while patients await their appointment with the hepatologist. Because of the high sensitivity and ease of administration, ICT would be useful as a screening test in clinics, which will aid in the decision of whether to treat or to send the patient for further testing. A modified version of the ICT will be made freely available to be downloaded after the trial has been published, which can increase the availability of MHE testing. In addition, because ICT involves recognizing specific letters, it can potentially be administered to non-english-speaking subjects with minimal modifications.

9 November 2008 DIAGNOSIS OF MINIMAL HEPATIC ENCEPHALOPATHY 1599 This study has several limitations. The educational status of the cirrhotic patients is high compared with the general population; however, despite this, cirrhotic patients as a whole and MHE patients in particular performed significantly worse on ICT and SPT compared with education-matched controls. Similar to our previous experience, most cirrhotic patients were Child Pugh class A or B. This is probably due to the high likelihood of Child Pugh class C patients being on OHE therapy, which was an exclusion criterion for this study. There is no current gold standard for MHE diagnosis, and this study used a modification of a recommended SPT battery; however, this study as well as our prior experience has found this method of MHE diagnosis to successfully predict OHE and driving impairment. 19,21 In addition, there was a good correlation between the number of impaired SPT and ICT results. The sample size for the TIPS and yogurt trial aspect were limited, but they did demonstrate a significant change in ICT that was similar to the change in SPT that was expected after TIPS and therapy. In summary, ICT is a sensitive test for the diagnosis of MHE, which can also predict OHE similar to SPT and has good external validity and test-retest reliability. ICT can be administered by clinic personnel without the need for psychologic expertise, which makes it an inexpensive option for MHE diagnosis. Supplementary Data Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.gastro References 1. Mullen K, Ferenci P, Bass NM, et al. An algorithm for the management of hepatic encephalopathy. Semin Liv Dis 2007; 27: Li YY, Nie YQ, Sha WH, et al. Prevalence of subclinical hepatic encephalopathy in cirrhotic patients in China. World J Gastroenterol 2004;10: Das A, Dhiman RK, Saraswat VA, et al. Prevalence and natural history of subclinical hepatic encephalopathy in cirrhosis. J Gastroenterol Hepatol 2001;16: Romero-Gomez M, Boza F, Garcia-Valdecasas MS, et al. Subclinical hepatic encephalopathy predicts the development of overt hepatic encephalopathy. Am J Gastroenterol 2001;96: Prasad S, Dhiman RK, Duseja A, et al. Lactulose improves cognitive functions and health-related quality of life in patients with cirrhosis who have minimal hepatic encephalopathy. Hepatology 2007;45: Wein C, Koch H, Popp B, et al. Minimal hepatic encephalopathy impairs fitness to drive. Hepatology 2004;39: Bajaj JS, Hafeezullah M, Hoffmann RG, et al. Minimal hepatic encephalopathy: a vehicle for accidents and traffic violations. Am J Gastroenterol 2007;102: Groeneweg M, Quero JC, De Bruijn I, et al. Subclinical hepatic encephalopathy impairs daily functioning. Hepatology 1998;28: Weissenborn K, Ennen JC, Schomerus H, et al. Neuropsychological characterization of hepatic encephalopathy. J Hepatol 2001; 34: Ford JM, Gray M, Whitfield SL, et al. Acquiring and inhibiting prepotent responses in schizophrenia: event-related brain potentials and functional magnetic resonance imaging. Arch Gen Psychiatry 2004;61: Schiff S, Vallesi A, Mapelli D, et al. Impairment of response inhibition precedes motor alteration in the early stage of liver cirrhosis: a behavioral and electrophysiological study. Metab Brain Dis 2005;20: Ferenci P, Lockwood A, Mullen K, et al. Hepatic encephalopathy definition, nomenclature, diagnosis, and quantification: final report of the working party at the 11th World Congresses of Gastroenterology, Vienna, Hepatology 2002;35: Bajaj JS, Etemadian A, Hafeezullah M, et al. Testing for minimal hepatic encephalopathy in the United States: an AASLD survey. Hepatology 2007;45: Epstein JN, Johnson DE, Varia IM, et al. Neuropsychological assessment of response inhibition in adults with ADHD. J Clin Exp Neuropsychol 2001;23: Konrad K, Gauggel S, Manz A, et al. Inhibitory control in children with traumatic brain injury (TBI) and children with attention deficit/ hyperactivity disorder (ADHD). Brain Inj 2000;14: Garavan H, Ross TJ, Stein EA. Right hemispheric dominance of inhibitory control: an event-related functional MRI study. Proc Natl Acad Sci U S A 1999;96: Pliszka SR, Liotti M, Woldorff MG. Inhibitory control in children with attention-deficit/hyperactivity disorder: event-related potentials identify the processing component and timing of an impaired right-frontal response-inhibition mechanism. Biol Psychiatry 2000;48: Ballard JC. Assessing attention: comparison of response-inhibition and traditional continuous performance tests. J Clin Exp Neuropsychol 2001;23: Bajaj JS, Saeian K, Verber MD, et al. Inhibitory control test is a simple method to diagnose minimal hepatic encephalopathy and predict development of overt hepatic encephalopathy. Am J Gastroenterol 2007;102: Wechsler D. Wechsler Adult Intelligence Scale-III. San Antonio, TX; Psychological Corp, Bajaj JS, Hafeezullah M, Hoffmann RG, et al. Navigation skill impairment: another dimension of the driving difficulties in minimal hepatic encephalopathy. Hepatology 2008;47: Bajaj JS, Saeian K, Christensen K, et al. Probiotic yogurt for the treatment of minimal hepatic encephalopathy. Am J Gastroenterol 2008;103: AMA. Current Procedural Terminology. Chicago, IL; American Medical Association, Weissenborn K, Giewekemeyer K, Heidenreich S, et al. Attention, memory, and cognitive function in hepatic encephalopathy. Metab Brain Dis 2005;20: Walker AJ, Shores EA, Trollor JN, et al. Neuropsychological functioning of adults with attention deficit hyperactivity disorder. J Clin Exp Neuropsychol 2000;22: Ortiz M, Cordoba J, Jacas C, et al. Neuropsychological abnormalities in cirrhosis include learning impairment. J Hepatol 2006; 44: Ortiz M, Jacas C, Cordoba J. Minimal hepatic encephalopathy: diagnosis, clinical significance and recommendations. J Hepatol 2005;42(Suppl):S45 S Perry W, Hilsabeck RC, Hassanein TI. Cognitive dysfunction in chronic hepatitis C: a review. Dig Dis Sci 2008;53:

10 1600 BAJAJ ET AL GASTROENTEROLOGY Vol. 135, No Watanabe A, Sakai T, Sato S, et al. Clinical efficacy of lactulose in cirrhotic patients with and without subclinical hepatic encephalopathy. Hepatology 1997;26: Sanyal AJ, Freedman AM, Shiffman ML, et al. Portosystemic encephalopathy after transjugular intrahepatic portosystemic shunt: results of a prospective controlled study. Hepatology 1994;20: Jalan R, Elton RA, Redhead DN, et al. Analysis of prognostic variables in the prediction of mortality, shunt failure, variceal rebleeding and encephalopathy following the transjugular intrahepatic portosystemic stent-shunt for variceal haemorrhage. J Hepatol 1995;23: Riggio O, Masini A, Efrati C, et al. Pharmacological prophylaxis of hepatic encephalopathy after transjugular intrahepatic portosystemic shunt: a randomized controlled study. J Hepatol 2005;42: Mullen KD, Ghanta RK, Putka BS. Prevention of first overt episode of hepatic encephalopathy after TIPS: no easy task. Hepatology 2006;43: Liu Q, Duan ZP, Ha da K, et al. Synbiotic modulation of gut flora: effect on minimal hepatic encephalopathy in patients with cirrhosis. Hepatology 2004;39: Malaguarnera M, Greco F, Barone G, et al. Bifidobacterium longum with fructo-oligosaccharide (FOS) treatment in minimal hepatic encephalopathy: a randomized, double-blind, placebocontrolled study. Dig Dis Sci 2007;52: Received April 8, Accepted July 17, Address requests for reprints to: Jasmohan S Bajaj, MD, MS, Division of Gastroenterology, Hepatology and Nutrition, Virginia Commonwealth University and McGuire VA Medical Center, 1201 Broad Rock Blvd, Richmond, VA. jasmohan.bajaj@va.gov; fax: (804) Supported in part by GCRC grant number M01-RR00058 and by the New Investigator Research Affairs Committee grant at the Medical College of Wisconsin (to J.S.B.). The authors disclose no conflicts.

11 November 2008 DIAGNOSIS OF MINIMAL HEPATIC ENCEPHALOPATHY 1600.e1 Supplementary Figure 1. Lure response rate before and after TIPS placement showed a significant worsening in lure response rate after TIPS placement. Three patients developed OHE and were re-tested while they were on lactulose. Total number of lures is 40. Supplementary Figure 2. There was a significant improvement in ICT lures after probiotic yogurt supplementation in a randomized controlled trial. The patients randomized to the no-treatment group did not show any significant difference in lure response rate before and after the trial. Solid lines indicate patients randomized to probiotic yogurt and dotted lines were patients randomized to the no-treatment group.