A Simplified Psychometric Evaluation for the Diagnosis of Minimal Hepatic Encephalopathy

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2011;9: A Simplified Psychometric Evaluation for the Diagnosis of Minimal Hepatic Encephalopathy OLIVIERO RIGGIO,* LORENZO RIDOLA,* CHIARA PASQUALE,* ILARIA PENTASSUGLIO,* SILVIA NARDELLI,* FEDERICA MOSCUCCI,* MANUELA MERLI,* SARA MONTAGNESE, PIERO AMODIO, and CARLO MERKEL *Department of Clinical Medicine, Center for the Diagnosis and Treatment of Portal Hypertension, Sapienza University of Rome, Rome; and Department of Clinical and Experimental Medicine, University of Padua, Padua, Italy BACKGROUND & AIMS: The psychometric hepatic encephalopathy score (PHES), which includes 5 psychometric tests, is a standard for the diagnosis of minimal hepatic encephalopathy (HE). We investigated whether a simplified PHES (SPHES) is as useful as the whole PHES. METHODS: The PHES was determined for 79 cirrhotic patients (the training group), who were followed up for the development of overt HE. Backward logistic regression was performed by eliminating stepwise variables removal did not impair regression. A separate series of 65 patients was used as a validation group. RE- SULTS: The PHES was abnormal in 45 patients. The SPHES, determined from the digit symbol, serial dotting, and line tracing tests, did not differ significantly from the full PHES; 24 of the 79 patients developed overt HE. The likelihood of developing overt HE was higher among patients with an abnormal PHES (log-rank P.003) or SPHES (P.004). By using Cox regression and model for end-stage liver disease scores to analyze data from patients with previous HE and transjugular intrahepatic portosystemic shunts, PHES (relative risk, 4.16; P.003) and SPHES (relative risk, 3.70; P.004) were the only variables associated with the development of overt HE. The accuracy of the SPHES was confirmed in the validation group. CONCLUSIONS: A simplified PHES is as good as the PHES in diagnosing minimal HE and in predicting the occurrence of overt HE. Keywords: Psychometric Evaluation; Minimal Hepatic Encephalopathy; Mild Cognitive Impairment; Liver Cirrhosis. Minimal hepatic encephalopathy (MHE) is detectable in 20% to 60% of cirrhotic patients depending on the tools used for its diagnosis and on the severity of the underlying liver disease. 1 3 MHE is associated with a reduced quality of life 4 and an increased frequency of car accidents and traffic violations 5 7 and is a risk factor for the development of overt HE. 8,9 The optimal MHE diagnosis is a matter of debate. The psychometric hepatic encephalopathy score (PHES), which includes 5 paper-and-pencil tests, 10 is the recommended standard. 11 However, PHES is not widely used. According to a recent survey, although MHE is considered a significant problem by 84% of the members of the American Association for the Study of Liver Diseases, MHE is never screened by 38% of them 12 because psychometric testing is time consuming, difficult, and not standardized. To overcome these problems, computerized psychometric tests 13 recently have been proposed as a simplified tool for MHE detection. However, the accuracy for MHE diagnosis as well as the capacity of predicting the development of overt HE by the Inhibitory Control Test has to be confirmed. 14 To increase the feasibility of MHE screening in routine clinical practice, we performed a study to establish whether the PHES battery can be simplified to a lower number of subtests, without decreasing its diagnostic accuracy or its ability to identify patients at risk of overt HE. Patients and Methods From March to October 2009, all consecutive cirrhotic patients without overt HE (evaluated by the West Haven criteria, 15 the clinical HE staging scale, 16 and a pool of standardized closed questions) admitted to the Gastroenterology Unit in Rome were eligible for the study. Exclusion criteria were overt HE within 15 days before the study, alcohol/psychoactive drug intake (positive alcoholemia and/or benzodiazepines or opioid urine metabolites), and unrelated neurologic disease. Although there is some evidence of persistent cognitive impairment after an episode of overt HE, 17,18 patients with a history of overt HE for more than 15 days were not excluded. These patients often are affected by MHE and are prompt to develop overt HE in the follow-up period, thus representing a demonstrative sample to study the diagnostic accuracy of tests for MHE. One hundred cirrhotic patients were examined: 79 were included (Supplementary Table 1) and 21 were excluded because of overt HE. At inclusion, none of the patients was taking lactulose or nonabsorbable antibiotics. Some patients were taking diuretics (58%) or -blockers (36%). Informed written consent to participate in the study was obtained. The Sapienza University of Rome Ethical Committee approved the study (Rif. 1720/ ). The psychometric evaluation was performed in a quiet room, with no distracting noises. Patients underwent the PHES battery, including the digit symbol test (DST), trail-making test A (TMT-A) and trail-making test B (TMT-B), the serial dotting test (SDT), and the line tracing test (LTT). 10 Each test was scored against age- and education-adjusted norms for the Italian population. The PHES is the sum of integer scores of each test computed from the adjusted Z values, as follows: score Abbreviations used in this paper: DST, digit-symbol test; LTT, linetracing test; MELD, model for end-stage liver disease; MHE, minimal hepatic encephalopathy; PHES, psychometric hepatic encephalopathy score; SDT, serial-dotting test; SPHES, simplified psychometric hepatic encephalopathy score; TMT-A, trail-making test A; TMT-B, trail-making test B by the AGA Institute /$36.00 doi: /j.cgh

2 614 RIGGIO ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 7 3 for Z 3, score 2 for 3 Z 2, score 1 for 2 Z 1, score 0 for 1 Z 1, score 1 for Z 1. A PHES of 4 or less was considered abnormal. 19 Patients then were followed up with planned visits. Both the patients and their families were instructed to strictly adhere to the follow-up evaluation and to contact the medical staff immediately should any alteration in the patients mental status occur. Patients were censored at the time of overt HE, death, liver transplantation, or when a transjugular intrahepatic portosystemic shunt (TIPS) procedure was performed. No treatment was prescribed for HE prophylaxis. Validation Study Sixty-five patients with cirrhosis without overt HE, observed and followed-up in the Department of Clinical and Experimental Medicine at the University of Padua, assessed with the PHES, and spectrally analyzed electroencephalogram, 19 followed up for up to 46 months (mean, 22 mo), were used to validate the SPHES. Statistical Analysis A backward logistic regression analysis was performed by incorporating, as a first step, the 5 Z scores of the variables included in the PHES, and by subsequently eliminating stepwise those variables that could be removed without impairing the regression (Wald statistic was lowest with a P value.10, and if the overall chi-square of the model was not decreased significantly from the previous step). To compare the ability of PHES and of the simplified PHES (SPHES) in predicting overt HE, Kaplan Meier curves were used and compared by the Mantel Cox test. In addition, Cox multiple regressions were performed by evaluating the PHES or SPHES potential predictive value by including the Child Pugh class or model for end-stage liver disease (MELD), the presence of TIPS, and history of overt HE as possible confounding factors; a forward stepwise method was used. The results are presented as mean standard deviation. Results According to PHES, 45 patients had MHE (Table 1). In the MHE group, the number of patients with a previous history of overt HE was significantly higher (15 vs 3 patients; P.01). The logistic analysis showed that a model containing only DST, SDT, and LLT was similar to that containing the whole set of tests (Supplementary Table 2). The SPHES therefore was defined according to the regression coefficients and the Z scores of the 3 remaining tests (see Appendix). The rate of patients classified with MHE according to the PHES or SPHES was not significantly different (45 of 79 vs 42 of 79; P NS). During a follow-up period of months, 20 patients died, 4 were submitted to liver transplantation, and 4 underwent a TIPS procedure. Twenty-four patients developed overt HE. The probability of developing overt HE was higher in patients with MHE according to PHES or SPHES compared with those without MHE (Figure 1A). All patients but 5 were concordantly classified based on PHES and SPHES. On Cox regression analysis, including MELD score, TIPS, previous overt HE, and PHES/SPHES, only the latter were associated significantly with overt HE development (Supplementary Table 3). Table 1. Comparison Between Patients With and Without MHE According to a PHES Score of 4 or Less Included in the Training Group MHE (n 34) MHE (n 45) P value Age, y NS Sex, M/F 23/11 26/19 NS Education, y NS Child Pugh class, 19/9/6 14/22/9 NS A/B/C Child Pugh score NS MELD score NS Patients with history 31/3 30/15.01 of HE, no/yes, n PHES score TMT-A, s TMT-B, s DS, n SDT, s LTT, s LTT, errors NOTE. 2 or Student t test were used for unpaired data; mean SD shown. Validation Study The validation group (65 patients; age, 54 9 y; male/ female, 45/20; etiology of cirrhosis (alcohol/virus/other), 20/ 35/10; MELD score, 12 7), consisted of younger patients (age, 54 9vs60 13 y; P.014) with more compensated liver disease (MELD score, 12 7vs14 6; P.06), and fewer TIPS procedures (1 vs 6). The education level was similar ( vs y; P.7). In the validation group, 18 of 65 patients (28%) had MHE based on PHES and 13 had MHE based on SPHES. The mean dominant electroencephalogram frequency was lower in patients with MHE according to both PHES ( vs Hz; P.003) and SPHES ( vs Hz; P.01). As observed in the training group, both PHES and SPHES were significant predictors of occurrence of overt HE (Mantel Cox test, 11.75; P.001; and Mantel Cox test, 9.20; P.002; respectively) also in the validation group (Figure 1B). Discussion This study shows that PHES can be simplified and that TMT-A and TMT-B could be removed without a significant loss in diagnostic ability. The present study also confirms that patients with MHE are at risk of developing overt HE 8,9 and that the SPHES maintains the ability to identify these individuals. To confirm the validity of this observation, the SPHES assessment of MHE using the coefficients derived from the original group was tested in a separate group of patients. In this group, PHES and SPHES also agreed in diagnosing MHE, were equally effective in predicting the risk of overt HE, and they also showed a relationship with electroencephalogram slowing. Therefore, it would appear that the clinical value of SPHES is confirmed outside the group of patients used to develop it. The neuropsychological basis of the battery redundancy is not necessarily easy to explain but may relate to the fact that all the tests included are affected, to some extent, by psychomotor slowing. Indeed, TMT-A and TMT-B were included in the PHES

3 July 2011 SIMPLIFIED DIAGNOSIS FOR MHE 615 Figure 1. Cumulative probability of developing overt HE in the (A) training series and in the (B) validation group in patients with (solid line) or without (dotted line) MHE according to PHES or SPHES. in consideration of their widespread use in patients with cirrhosis because they had not been selected as independently diagnostic in the German discriminant analysis. 10 The TMT-A is certainly the least demanding of the 5 tests. DST and TMT-B are probably of a very similar difficulty and they both test complex attention. 20 Thus, SPHES, which does not include TMT-A and TMT-B, may be considered more difficult than PHES. Patients with MHE and attention deficits may be more likely to forget the LTT and the SDT instructions during the test execution and could be labeled inappropriately as affected by encephalopathy. The empiric demonstration of a similar performance of the 2 models in terms of diagnostic accuracy was therefore necessary. On the other hand, the use of SPHES decreases the time required for MHE screening. In addition, SPHES is independent on different alphabets (TMT-B implies use of alphabetic sequences), which may result in a more widespread application. In conclusion, this study suggests that a simplified psychometric evaluation using only 3 of the 5 tests included in the PHES is as efficient as PHES in detecting patients with MHE and in predicting the subsequent occurrence of overt HE. Because SPHES performance is not inferior and is faster than PHES, a gold standard for the diagnosis of MHE, it may contribute to increasing routine screening for this important complication of cirrhosis. Appendix SPHES e [( 12.6 (4.4 DST) (1.43 SDT) (2.20 LTT)] /1 e [( 12.6 (4.4 DST) (1.43 SDT) (2.20 LTT)], where DST, SDT, and LTT are the Z score of DST, SDT, and LTT tests, respectively. According to the logistic regression model, values of SPHES greater than 0.50 were diagnostic for the presence of MHE. Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenter-

4 616 RIGGIO ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 7 ology and Hepatology at and at doi: / j.cgh References 1. Amodio P, Montagnese S, Gatta A, et al. Characteristics of minimal hepatic encephalopathy. Metab Brain Dis 2004;19: Quero JC, Hartmann IJ, Meulstee J, et al. The diagnosis of subclinical hepatic encephalopathy in patients with cirrhosis using neuropsychological tests and automated electroencephalogram analysis. Hepatology 1996;24: Amodio P, Del Piccolo F, Marchetti P, et al. Clinical features and survival of cirrhotic patients with subclinical cognitive alterations detected by the number connection test and computerized psychometric tests. Hepatology 1999;29: Prasad S, Dhiman RK, Duseja A, et al. Lactulose improves cognitive function and health-related quality of life in cirrhotic patients with minimal hepatic encephalopathy. Hepatology 2007; 45: 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, Schubert CM, et al. Minimal hepatic encephalopathy is associated with motor vehicle crashes: the reality beyond the driving test. Hepatology 2009;50: Kircheis G, Knoche A, Hilger N, et al. Hepatic encephalopathy and fitness to drive. Gastroenterology 2009;137: Romero-Gómez M, Córdoba J, Jover R, et al. Value of the critical flicker frequency in patients with minimal hepatic encephalopathy. Hepatology 2007;45: Hartmann IJ, Groeneweg M, Quero JC, et al. The prognostic significance of subclinical hepatic encephalopathy. Am J Gastroenterol 2000;95: Weissenborn K, Ennen JC, Schomerus H, et al. Neuropsychological characterization of hepatic encephalopathy. J Hepatol 2001; 34: Ferenci P, Lockwood A, Mullen K, et al. Hepatic encephalopathy definition, nomenclature, diagnosis and quantification. 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: Bajaj JS, Hafeezullah M, Franco J, et al. Inhibitory control test for the diagnosis of minimal hepatic encephalopathy. Gastroenterology 2008;135: Amodio P, Ridola L, Schiff S, et al. Improving detection of minimal hepatic encephalopathy using the inhibitory control task. Gastroenterology 2010;139: Conn HO, Leevy CM, Vlahcevic ZR, et al. Comparison of lactulose and neomycin in the treatment of chronic portal-systemic encephalopathy. A double blind controlled trial. Gastroenterology 1977; 72: Ortiz M, Córdoba J, Doval E, et al. Development of a clinical hepatic encephalopathy staging scale. Aliment Pharmacol Ther 2007;15: Bajaj JS, Schubert CM, Heuman DM, et al. Persistence of cognitive impairment after resolution of overt hepatic encephalopathy. Gastroenterology 2010;138: Riggio O, Ridola L, Pasquale C, et al. Evidence of persistent cognitive impairment after resolution of overt hepatic encephalopathy. Clin Gastroenterol Hepatol 2011;9: Amodio P, Campagna F, Olianas S, et al. Detection of minimal hepatic encephalopathy: normalization and optimization of the Psychometric Hepatic Encephalopathy Score. A neuropsychological and quantified EEG study. J Hepatol 2008;49: Lezak MD. Neuropsychological assessment. 3rd ed. New York: Oxford University Press, Reprint requests Address requests for reprints to: Oliviero Riggio, MD, Department of Clinical Medicine, Sapienza University of Rome, Viale dell Università 37, Rome, Italy. oliviero.riggio@uniroma1.it; fax: (39) (06) /(39) (06) Acknowledgments The authors thank Professor Karin Weissenborn for the psychometric hepatic encephalopathy score forms. Conflicts of interest The authors disclose no conflicts. Funding This study was partially supported by a Young Investigator Award given by the Italian Society of Gastroenterology (L.R.).

5 616.e1 RIGGIO ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 7 Supplementary Table 1. Demographic and Clinical Characteristics of the Training Group Patients, n 79 Age, y Sex, M/F 49/30 Education, y Etiology of cirrhosis, alcohol/virus/other 17/52/10 Child Pugh class, A/B/C 33/31/15 MELD score 14 6 Esophageal varices, absent/present 42/36 a NOTE. Mean SD shown. a One patient did not undergo upper endoscopy. Supplementary Table 2. Statistical Parameters of the Backward Stepwise Logistic Regression Analysis Model Overall 2 Decrease in 2 P 5 variables variables variables Final model Variable SE ( ) DS SDT LTT Constant Supplementary Table 3. Cox Regression Models Predicting Overt HE Variables SE ( ) P RR Overall 2 P Increased 2 P Risk of overt HE Variables tested: PHES, MELD score, TIPS, previous overt HE PHES / Variables tested: SPHES, MELD score, TIPS, previous overt HE SPHES / SE, standard error; RR, relative risk.