PHES: One label, different goods?!

Detection of minimal hepatic encephalopathy: Normalization and optimization of the Psychometric Hepatic Encephalopathy Score. A neuropsychological and quantified EEG studyJournal of HepatologyVol. 49Issue 3PreviewPsychometric Hepatic Encephalopathy Score (PHES) and EEG are used to detect minimal hepatic encephalopathy (MHE). We aimed at standardizing PHES in Italy and comparing Italian, German and Spanish norms in EEG characterized cirrhotic patients. Full-Text PDF Psychometric testing has a long tradition in clinical hepatology. It has been used for decades for the assessment and follow-up of hepatic encephalopathy (HE). Initially hand-writing, or the construction of a five-pointed star were common methods to monitor HE. But, since the evaluation of these tests could not be standardized, alternative solutions were sought which combined convenience, practicability, sensitivity, specificity, and low costs in terms of money and time. A first approximation towards these goals was the introduction of the Number Connection Tests into the assessment of HE [[1]Zeegen R. Drinkwater J.E. Dawson A.M. Method for measuring cerebral dysfunction in patients with liver disease.Brit Med J. 1970; 2: 633-636Crossref PubMed Scopus (90) Google Scholar]. Then comprehensive batteries of psychometric tests were used, and it was shown that HE is characterized by a distinct pattern of cognitive impairment: alterations of attention, visuo-spatial perception and psychomotor function [2Gilberstadt S.J. Gilberstadt H. Zieve L. Buegel B. Collier R.O. Mc Clain J. Psychomotor performance defects in cirrhotic patients without overt encephalopathy.Arch Intern Med. 1908; 140: 519-521Crossref Scopus (155) Google Scholar, 3Rikkers L. Jenko P. Rudman D. Freides D. Subclinical hepatic encephalopathy: detection, prevalence, and relationship to nitrogen metabolism.Gastroenterology. 1978; 75: 462-469PubMed Google Scholar, 4Rehnström S. Simert G. Hansson J.A. Johnson G. Vang J. Chronic hepatic encephalopathy. A psychometric study.Scand J Gastroenterol. 1977; 12: 305-311Crossref PubMed Scopus (95) Google Scholar, 5Tarter R.E. Hegedus A.M. Van Thiel D.H. Schade R.R. Gavaler J.S. Starzl T.E. Nonalcoholic cirrhosis associated with neuropsychological dysfunction in the absence of overt evidence of hepatic encephalopathy.Gastroenterology. 1984; 86: 1421-1427Abstract Full Text PDF PubMed Scopus (170) Google Scholar]. While the verbal IQ was preserved in patients with overt hepatic encephalopathy the performance IQ was found to be decreased even in patients without clinical signs of HE [2Gilberstadt S.J. Gilberstadt H. Zieve L. Buegel B. Collier R.O. Mc Clain J. Psychomotor performance defects in cirrhotic patients without overt encephalopathy.Arch Intern Med. 1908; 140: 519-521Crossref Scopus (155) Google Scholar, 3Rikkers L. Jenko P. Rudman D. Freides D. Subclinical hepatic encephalopathy: detection, prevalence, and relationship to nitrogen metabolism.Gastroenterology. 1978; 75: 462-469PubMed Google Scholar]. The PHES – the psychometric hepatic encephalopathy score – has its roots in these flourishing times of neuropsychological assessment of HE. The term PHES was coined by Dr. Andres Blei in 2001. He suggested to name the sum score of the PSE-Syndrom-Test [[6]Schomerus H, Weissenborn K, Hamster W, Rückert N, Hecker H. PSE-Syndrom-Test. Swets Test Services. Frankfurt: Swets & Zeitlinger B.V.; 1999.Google Scholar], a test battery which had been especially developed for diagnosing HE, psychometric hepatic encephalopathy score (PHES) [[7]Weissenborn K. Ennen J.C. Schomerus H. Rückert N. Hecker H. Neuropsychological characterization of hepatic encephalopathy.J Hepatol. 2001; 34: 768-773Abstract Full Text Full Text PDF PubMed Scopus (622) Google Scholar]. The basis for the PSE-Syndrom-Test had been laid by the neuropsychologist Wolfgang Hamster† and the gastroenterologist Hans Schomerus† in the early eighties. They aimed to characterize the neuropsychological deficit typical for HE, and to develop a short battery that could be easily used in the clinic to make the diagnosis of minimal HE. Therefore, they presented a battery of more than 20 neuropsychological tests to patients with liver cirrhosis, patients with alcohol-toxic pancreatitis without liver cirrhosis, patients with alcohol-toxic neuropathy without cirrhosis and healthy controls [8Hamster W. Neuropsychologie der latenten portosystemischen Enzephalopathie. PhD thesis. Medizinische Fakultät der Eberhard-Karls-Universität Tübingen; 1982.Google Scholar, 9Schomerus H. Hamster W. Neuropsychological aspects of portal–systemic encephalopathy.Metab Brain Dis. 1998; 13: 361-377Crossref PubMed Scopus (117) Google Scholar]. They found deficits in concentration, attention and psychomotor function in the liver patients compared to controls. A discriminant analysis was performed to identify those variables which were capable of differentiating between patients with and without liver cirrhosis and between patients with HE and patients with alcoholic brain atrophy and healthy controls. The Digit Symbol Test, the Benton Test, the Line tracing Test (in particular the number of errors), reaction time to acoustic and visual stimuli, and the tests steadiness, aiming and long sticks from the "Motorische Leistungsserie", a battery aimed to detect alterations of fine motor skills, showed a high discrimination accuracy. For practicability reasons they decided to develop a paper-pencil-test battery that represented the affected domains such as visual-motor coordination, attention shift and motor speed and accuracy. Considering the results of the discriminant analysis the Digit Symbol Test (DST), a paper-pencil version of the Line Tracing Test (LTT) and instead of aiming the Serial Dotting Test (SDT) were included into the battery. Then the Number Connection Tests (NCT) A and B (also referred to as Trailmaking Tests) were added because these two tests were the most frequently used psychometric tests for the diagnosis of minimal HE at that time. The new battery was presented to more than 400 individuals to develop normative data [[10]Schönleber R. Standardisierung eines psychometrischen Tests zur Erfassung der hepatischen Enzephalopathie. Thesis. Medizinische Fakultät der Eberhard-Karls-Universität Tübingen; 1989.Google Scholar]. Most of the tested people were in-patients of the Department of Internal Medicine, part were forced to undergo a neuropsychological examination to keep their driving licence. Unfortunately, the age pattern of this "representative group" was disproportionate. Therefore, in a second step the battery was presented to a group of 120 healthy controls with a balanced pattern of age, education and occupation [[11]Ennen J. Diagnosestandardisierung der latenten portosystemischen Enzephalopathie mittels psychometrischer Testverfahren. Thesis. Medizinische Hochschule Hannover; 2000.Google Scholar]. Finally, in 1999, the battery was published by SWETS Test Services in German. Hamster and Schomerus made several attempts to publish their data regarding an optimal test combination for the diagnosis of HE. At the end of the eighties, however, psychometric testing – especially the application of paper-pencil-tests – had become obsolete. More "sophisticated" technical methods were expected to produce a higher sensitivity and a higher specificity than paper-pencil-tests [12Davies M.G. Rowan M.J. Feely J. EEG and event related potentials in hepatic encephalopathy.Metab Brain Dis. 1991; 6: 175-186Crossref PubMed Scopus (26) Google Scholar, 13Weissenborn K. Neurophysiological methods in the diagnosis of early hepatic encephalopathy.in: Bengtsson F. Jeppsson B. Almdahl T. Vilstrup H. Progress in hepatic encephalopathy and metabolic nitrogen exchange. CRC Press, Boca Raton1991: 27-39Google Scholar, 14Kircheis G. Wettstein M. Timmermann L. Schnitzler A. Häussinger D. Critical flicker frequency for quantification of low-grade hepatic encephalopathy.Hepatology. 2002; 35: 357-366Crossref PubMed Scopus (329) Google Scholar]. Drs. Hamster and Schomerus would be highly pleased to see their test battery finally being appreciated all over the world. After the standardization in Germany similar test batteries have also been standardized in Spain [[15]Romero Gómez M. Córdoba J. Jover R. del Olmo J. Fernández A. Flavià M. et al.Normality tables in the Spanish population for psychometric tests used in the diagnosis of minimal hepatic encephalopathy.Med Clin (Barc). 2006; 127: 246-249Crossref PubMed Scopus (59) Google Scholar] and, now, in Italy [[16]Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar] and the United Kingdom [[17]Marks M.E. Jackson C.D. Montagnese S. Jenkins C.W. Head M. Morris R.W. et al.Derivation of a normative UK database for the psychometric hepatic encephalopathy score (PHES): confounding effect of ethnicity and test scoring.J Hepatol. 2008; 48: S119Abstract Full Text PDF Google Scholar]. An altered test version has been standardized in India [[18]Thumburu KK, Kurmi R, Dhiman RK, Venkataramarao SH, Prabhakar S, Singh P, et al. Psychometric hepatic encephalopathy score, critical flicker frequency and P300 event-related potential for the diagnosis of minimal hepatic encephalopathy: evidence that psychometric hepatic encephalopathy score is enough. 13th ISHEN, Padua; 2008 [abstract].Google Scholar]. All test versions used so far are similar with regard to their principal structure. All combine a digit symbol test, a serial dotting test, a line tracing test and number or figure connection tests. Nevertheless, there are significant differences in details that make it difficult to compare the results achieved with the different test versions. Amodio et al. [[16]Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar], for example, used for their battery presented in this issue of journal the DST, SDT and LTT forms of the PSE-Syndrom-Test and added previously developed forms for the NCT A and B because of differences between the Italian and German alphabet sequences. In addition to the distribution of numbers and letters the dimension of circles and letters in these tests were different from the German forms. The Indian group substituted the NCT B with a test where the subjects had to connect figures belonging to a specific category because they have to deal with a high amount of non-alphabetized patients [[19]Dhiman R.K. Saraswat V.A. Verma M. Naik S.R. Figure connection test: a universal test for assessment of mental state.J Gastroenterol Hepatol. 1995; 10: 14-23Crossref PubMed Scopus (74) Google Scholar]. Both alterations of the test (Fig. 1) have measurable influence on the test performance. The Indian solution even alters the test structure as their substitute for the NCT B assesses other cognitive functions than the NCT B (Table 1).Table 1Cognitive and motor functions addressed by the different components of the PHES – delivering test batteriesSub-testAssessable functionsNumber Connection Test APsychomotor speed; visual scanning efficiency, sequencing, attention, concentrationNumber Connection Test BAttention set shifting ability, psychomotor speed, visual scanning efficiency, sequencing, attention, concentrationFigure Connection Test [19]Dhiman R.K. Saraswat V.A. Verma M. Naik S.R. Figure connection test: a universal test for assessment of mental state.J Gastroenterol Hepatol. 1995; 10: 14-23Crossref PubMed Scopus (74) Google ScholarVisual perception, visual search, visual scanning efficiency, psychomotor speed, attention, concentration, working memoryDigit Symbol TestAssociative learning; graphomotor speed, cognitive processing speed, visual perception, working memorySerial Dotting TestMotor speedLine Tracing TestMotor speed and accuracy Open table in a new tab Besides differences in the structure, there are also marked differences in the evaluation procedure of the tests. All groups compare the subject's sub-test results to norms and score them with +1 to −3 points depending on their position in the +1 to −3 standard deviation range from the mean. Then a sum score is calculated which is referred to as Psychometric Hepatic Encephalopathy Score (PHES). Since all groups use the term PHES for the sum score of their test variation the difference between the batteries tends to be forgotten and PHES results from different study groups are compared without reservations. However, there are several very important differences in the calculation of the sum score. They become most evident considering the scoring of the line tracing test results. Hamster and Schomerus had shown in their discriminant analysis that performance time and number of errors added independently to the discrimination between patients with liver cirrhosis and their control groups with the number of errors being even better than the performance time. Therefore, they included both, performance time and errors into their sum score [[8]Hamster W. Neuropsychologie der latenten portosystemischen Enzephalopathie. PhD thesis. Medizinische Fakultät der Eberhard-Karls-Universität Tübingen; 1982.Google Scholar]. Amodio et al. [[16]Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar] argue that LTT performance time and errors are significantly related. Thus they measure LTT performance by "error-weighted time". Their argumentation is identical to that of the authors of the Spanish and British normal data, who also calculate a combined – though completely differently calculated – score for performance time and errors [15Romero Gómez M. Córdoba J. Jover R. del Olmo J. Fernández A. Flavià M. et al.Normality tables in the Spanish population for psychometric tests used in the diagnosis of minimal hepatic encephalopathy.Med Clin (Barc). 2006; 127: 246-249Crossref PubMed Scopus (59) Google Scholar, 17Marks M.E. Jackson C.D. Montagnese S. Jenkins C.W. Head M. Morris R.W. et al.Derivation of a normative UK database for the psychometric hepatic encephalopathy score (PHES): confounding effect of ethnicity and test scoring.J Hepatol. 2008; 48: S119Abstract Full Text PDF Google Scholar]. The argumentation is based on the data of healthy controls. It does not take into account that performance time and errors in the LTT represent different aspects of brain function, which may be altered independently by any pathophysiology (Fig. 2). Therefore, neuropsychologists suggest to assess performance speed and accuracy independently, if possible. The different handling of the LTT results in differences with regard to the range of the sum score. While the German version provides a range from +6 to −18 the Italian, Spanish and British versions provide a range from +5 to −15. Considering all changes of test structure and evaluation it becomes obvious that the cut-off score between normal and pathological results must be different between the different test versions, and must be determined for each of the different versions, separately. Since the test shall discriminate between normal brain function and hepatic encephalopathy the cut-off score should be determined by comparison of healthy controls and patients with clinically overt HE. Amodio et al. [[16]Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar] give reasons for their cut-off score of ⩽−4 by comparing their psychometric results with the EEG results as they had not examined patients with clinically overt HE. This approach is comprehensible. However, like all other substitute measures EEG is not able to fully represent HE. Even patients with clinically overt HE may present with normal EEG [[13]Weissenborn K. Neurophysiological methods in the diagnosis of early hepatic encephalopathy.in: Bengtsson F. Jeppsson B. Almdahl T. Vilstrup H. Progress in hepatic encephalopathy and metabolic nitrogen exchange. CRC Press, Boca Raton1991: 27-39Google Scholar]. The paper by Amodio et al. [[16]Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar] underscores the principal usefulness of the PSE-Syndrom-Test test battery for diagnosing minimal HE. However, it also highlights facts that have to be considered in psychometric testing principally: the influence of age, education, occupation and socio-cultural background and practice effects (see Table 2).Table 2Advantages and limitations of the psychometric test batteries that have been standardized for the assessment of minimal hepatic encephalopathy 6Schomerus H, Weissenborn K, Hamster W, Rückert N, Hecker H. PSE-Syndrom-Test. Swets Test Services. Frankfurt: Swets & Zeitlinger B.V.; 1999.Google Scholar, 15Romero Gómez M. Córdoba J. Jover R. del Olmo J. Fernández A. Flavià M. et al.Normality tables in the Spanish population for psychometric tests used in the diagnosis of minimal hepatic encephalopathy.Med Clin (Barc). 2006; 127: 246-249Crossref PubMed Scopus (59) Google Scholar, 16Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar, 17Marks M.E. Jackson C.D. Montagnese S. Jenkins C.W. Head M. Morris R.W. et al.Derivation of a normative UK database for the psychometric hepatic encephalopathy score (PHES): confounding effect of ethnicity and test scoring.J Hepatol. 2008; 48: S119Abstract Full Text PDF Google Scholar, 18Thumburu KK, Kurmi R, Dhiman RK, Venkataramarao SH, Prabhakar S, Singh P, et al. Psychometric hepatic encephalopathy score, critical flicker frequency and P300 event-related potential for the diagnosis of minimal hepatic encephalopathy: evidence that psychometric hepatic encephalopathy score is enough. 13th ISHEN, Padua; 2008 [abstract].Google ScholarAdvantagesLimitations•High sensitivity•High specificity•High reliability•High validity•Simplicity•Bedside tests•Low cost•Need for representative norm data for each sample studied considering effects of age, education, occupation, gender and socio-cultural background•Non-applicability of the NCT B in illiterate subjects•Need for the control of practice effects Open table in a new tab Importantly, normal data that have been elaborated in a representative sample from one country cannot be used for the evaluation of individual data in other countries without control. Due to alterations in educational levels and daily living activities from one generation to the other, neuropsychologists recommend the periodic re-evaluation of normal data even within a distinct population. How come, despite of all these drawbacks, the psychometric hepatic encephalopathy score (PHES) based on one or the other variation of the original PSE-Syndrom-Test is increasingly used for diagnosing minimal HE, worldwide? The battery is practical and easy to apply, sensitive and specific, and cheap. Practice effects in the range as in the present study by Amodio et al. [[16]Amodio P. Campagna F. Olianas S. Iannizzi P. Mapelli D. Penzo M. 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: 346-353Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar] have not been observed by other groups [[11]Ennen J. Diagnosestandardisierung der latenten portosystemischen Enzephalopathie mittels psychometrischer Testverfahren. Thesis. Medizinische Hochschule Hannover; 2000.Google Scholar]. None of the neuro- or psychophysiological methods used competitively has proven to be of greater use for diagnosing mHE, so far. This holds especially true also for the most recently recommended measure – the critical flicker frequency [[20]Romero-Gómez M. Córdoba J. Jover R. del Olmo J.A. Ramirez M. Rey R. et al.Value of the critical flicker frequency in patients with minimal hepatic encephalopathy.Hepatology. 2007; 45: 879-885Crossref PubMed Scopus (249) Google Scholar]. It is foreseeable that the norm differences between the different European countries will even out within the next decade. It appears worthwhile to directly compare meanwhile the raw data of the different representative samples of different countries, and to see whether the results would be more comparable after the deletion of one or the other sub-test, or after identical handling of the raw data. At best, this combined with a summarisation of all norm data collected to date would result in more widely applicable norms, which, for example, could be used for the assessment of HE in international multi-centre trials.