¹University Hospital Rotterdam-Dijkzigt, Dept. of Neurology
²Aphasia Foundation Rotterdam/ Rehabilitation Centre Rijndam.

Introduction

Problems in wordfinding constitute the most basic obstruction in communication in aphasic patients. The patients are either able to communicate to some extent by means of circumlocutions or they cannot communicate at all. Often, the listeners lose attention. The patient will either be frustrated or remains unaware of the meaninglessness of his or her message. Phonological and/or semantic disorders are reported to be mainly responsible for word finding deficits in aphasia and require a selective structural linguistic therapy.

The main goal of phonological therapy is to facilitate the selection and sequencing of speech sounds in speech production. A successful therapy should reduce the number of phonemic distortions in the patient’s speech. Therapeutic tools described in the literature are: naming with phonological cues, rhyme judgement, reading aloud and repetition.

The aim of semantic therapy is better recognition of the semantic features of content words. The therapy has been effective if it enhances the amount of adequate content words in spontaneous speech. This is accompanied by better comprehension. Word discrimination directed to the aspects of meaning is reported to be the most suitable therapy. Useful techniques are: auditory/written word/picture matching, word categorisation, judgement of semantic relatedness, antonym/synonym judgement for auditory word pairs.

Generally, structural therapy is applied in the initial stage of rehabilitation. The aim is two-fold:

1. to increase the patient’s insight into his main deficit
2. to improve the patient’s abilities in the most affected linguistic area.

After the period of structural linguistic therapy, the patient is stimulated to use his linguistic skills to exchange information in different communicative settings.

During the last two decades a number of papers have appeared on semantic or/and phonological intervention. Compared with phonological therapy, semantic therapy proves to be a better facilitator of naming tasks: both a long term effect and a generalisation effect to new objects have been observed. In contrast with phonological therapy, semantic therapy aims at the central level of language-processing models and therefore influences both language production and comprehension.

However, neither in the Netherlands nor in other countries, was there available a developed semantic therapy method, with various degrees of difficulty. The development of such a therapy seems to be justified with the premise that its efficacy has to be proved in every day language use. Progress in a naming task is not sufficient.

To explore the need for semantic therapy, the occurrence of semantic disorders was investigated in a group study. The Semantic Association Test was developed (Visch-Brink et al. 1996; Visch-Brink et al. in press); the patient is required to make semantic associations between words, pictures or words and pictures. Measured with the Semantic Association Test, 54% of the aphasic patients appeared to have a lexical semantic disorder. However, only concrete words are involved in the test. The percentage of patients with a lexical semantic disorder would probably be higher if abstract words were also included.

BOX, a lexical semantic therapy

The newly developed therapy programme, BOX (Visch-Brink et al. 1997; Visch-Brink 1997), provides a lexical semantic therapy, focusing on the interpretation of written words, sentences and texts without reference to pictures. The use of written words has the benefit that the dimension concrete-abstract may be built in, fulfilling a need for patients with moderate to mild forms of aphasia. Moreover, in contrast to auditorily presented words, written words give the patient the opportunity of processing the words over a considerable time. If the patient’s efforts result in a correct response, we may conclude that within the time spent on the decision process, the lexical semantic system is adequately addressed. A similar processing effort may in the long run facilitate access to the semantic system.

In developing BOX, we set ourselves the following aims:

1. We wanted the programme to be useful for aphasic patients with varying degrees of lexical semantic impairment.
2. The exercises had to cover a large number of topics, to meet individual points of interests.
3. There had to be considerable variation in the tasks to strengthen the generalisation effect.
4. We not only wanted tasks at the word level, but also at sentence and text level, plus a possibility of switching between those levels.

In its final form, the programme now consists of more than 1000 exercises at word, sentence and text level. All the exercises follow a pattern in which the patient has to make a choice out of a number of alternatives, or a false/correct decision. It is stressed in the literature that making semantic decisions is the pivot of lexical semantic therapy. ‘Practice in semantic word discrimination may improve the accuracy of word-finding more than therapy aimed specifically at output itself.’ (Butterworth et al. 1984).

There are eight different types of exercises:

I. Semantic Categories

II. Syntagmatic and Paradigmatic Relationship

III. Semantic Gradation

IV. Adjectives and Exclamations

V. Part-Whole Relationship

VI. Anomalous Sentences

VII. Semantic Definition

VIII. Semantic Context.

Most of the exercises contain three levels of difficulty.

For patients with a moderate semantic disorder, enough material is available for approximately six months of therapy, with therapy sessions of 90 minutes a week.

Pilot study I

In a preliminary evaluation of the programme’s effectiveness, BOX was applied to two patients with chronic aphasia, more than a year post onset. Both patients were tested before and after therapy by the Aachener Aphasia Test, and by the Semantic Association Test. Finally, to compare the function word/content word ratio in spontaneous speech before and after therapy, we used the Spontaneous Speech analysis, developed by Joanette and Goulet in 1990. Patient A was a left handed 79-year-old man with a severe Broca’s aphasia, who had been a road paver. During six months the patient received lexical semantic therapy by means of BOX, in two weekly sessions of an hour. The test results before and after therapy show an improvement in visual comprehension, subpart of the AAT, and of lexical semantic abilities, reflected in the performance on the Semantic Association Test (see table 1). The profile obtained from the Spontaneous Speech Analysis showed a difference over time. There was no difference in the ratio of content/function words before and after therapy. However, within the category content words, after therapy, the number of nouns decreased, whereas the number of verbs increased.

Patient B was a 71-year-old right handed housewife with a transcortical sensory aphasia. Lexical semantic therapy started one-year post onset, and continued for two months with a frequency of two weekly sessions of one hour. The AAT-results failed to show improvement, but there was an increase in the results on the Semantic Association Test, especially of performance on the verbal versions (see table 2). The improvement seemed to be independent of the input channel. The distribution of the various word groups in the spontaneous speech samples, analysed following Joanette and Goulet (1990), followed the same pattern as in patient A. After therapy the total number of meaningful verbs increased slightly. However, in contrast to patient A, patient B also produced more nouns in the second speech sample.

Those case studies showed that lexical semantic therapy was useful in two patients with severe aphasia and a lexical semantic deficit, who were both more than one year post onset CVA.

However, there was no control therapy.

Pilot study II

In the second pilot study, BOX therapy is compared with phonological therapy as a control therapy, using a cross-over design. Both spontaneous speech in daily life situations and the performance on different language modalities and linguistic levels are used as evaluative measurements. Eight aphasic stroke patients, five men and three women, with unilateral lesions in the left hemisphere completed the therapy. The demographic and medical data are shown in table 3. The patients were at least one year post onset. The phonological control therapy was directed at the selection and ordering of phonemes, and is based on a method described by Cubelli and colleagues in 1988. Four steps with increasing difficulty were built-in: (i) selection of existing words, (ii) compiling words out of syllables, (iii) analysing words into phonemes and (iv) making pronounceable sequences out of separate letters. At every stage the patient had to pronounce the word. No fixed order of exercises was prescribed for either of the therapeutic programmes. All speech therapists decided which exercises were most suitable at any given time for the patient being treated.

Four patients first underwent BOX, and then the phonological programme, and four patients underwent the therapy programmes in reverse order. Each programme was applied for 20 hours with a minimum of one-and-a-half and a maximum of three hours a week. Evaluative tests were administered before, after and in-between both therapy methods. The following tests were used: (i) the Aachener Aphasia Test to measure the patients’ performance on language modalities, (ii) the Semantic Association Test (words) and Synonym Judgement of written word pairs (PALPA, Kay et al. 1992; Dutch version, Bastiaanse et al. 1995) to measure semantic processing, (iii) Auditory lexical decision (PALPA), Judgement of Minimal pairs of existing and nonexisting words, auditorily presented (PALPA), Repetition of nonwords, to measure phonological processing. The Amsterdam Nijmegen Everyday Language Test (ANELT, Blomert et al. 1994) was used to measure the patient’s functioning in everyday language situations. The patient’s spontaneous speech is requested in ten different situations. The responses are assessed by two judges on two 5-points scales. Scale A, understandability, is a measure of the adequacy of the message; scale B measures intelligibility, that is the clarity of speech, independent of its content. The test results before and after therapeutic methods are presented in table 4 (AAT), in table 5 (ANELT and semantic processing), and in table 6 (phonological processing).

The tests with a positive change of scores in relation to the application of only one of the therapeutic methods were: (i) AAT-Repetition and AAT- Written language, in relation to the phonological programme, and (ii) AAT-Comprehension, ANELT-Scale A and Synonym Judgement in relation to BOX. The main finding of the study was that four out of eight patients made progress on the ANELT-Scale A after BOX-therapy (case 1, case 4, case 5 and case 8). It might be inferred that lexical semantic therapy is a valuable tool to restore verbal communication. Additional therapy with the focus on integration of linguistic skills in communicative situations seems to be superfluous.

Progress in naming was not better on BOX than on the Phonological Programme or vice-versa. This effect was expected, in view of the fact that both lexical semantic and phonological disorders cause naming deficits. Only a specific error analysis could reflect the effectiveness of BOX, as compared to that of the phonological programme. Such an analysis still has to be performed.

Patients with varying types of aphasia (Broca’a, Wernicke’s and Anomic) and different degrees of severity did benefit from BOX-therapy. This finding is in line with the fact that in a large group of aphasic patients (N=78), the performance on the verbal Semantic Association Test was not related to aphasia type or severity, as measured by the Token Test (Visch-Brink et al. 1996).

Our primary intention was to develop a therapy, which had the potential to be useful for a lot of aphasic patients. This ‘formal-semantic’ therapy, presenting the lexical form and the semantically associated words (see also Le Dorze et al. 1994), is aimed at activating features that distinguish between semantically related items. Variation in the nature and difficulty of the exercises provides the patient with the opportunity to practise semantic differentiations between words in different contexts. The main goal of the programme is to facilitate semantic processing in patients with a ‘general’ lexical semantic disorder (Franklin, 1993), which should have a positive influence on semantic processing in new situations. The progress observed on the ANELT is in line with this assumption.

Current research

We wished to demonstrate the efficacy of BOX therapy with a large group of aphasic patients. A randomised controlled clinical trial would be the most appropriate way of providing evidence for lexico-semantic therapy in aphasia. A randomised test to demonstrate the efficacy of BOX therapy is currently being carried out in 30 clinical centres across the Netherlands.

A type of phonological therapy recently developed (FIKS, Van Rijn et al., in press) is being used as a control therapy. This programme is closer to BOX in terms of variability and level of difficulty that the phonological therapy used in Pilot study II. The ANELT is our primary measure for evaluating the results. Patients with both lexico-semantic and phonological difficulties in the period between 3 and 5 months post-CVA are included. For a year post-onset, patients receive 60 hours of either lexico-semantic or phonological therapy. In order to discover both negative and positive factors affecting the efficacy of therapy, we will have detailed neurological information, information on neuropsychological disorders, language function and the patients’ psychosocial condition. A detailed description of the patients during therapy may answer the question about which aspects of BOX therapy are essential for which patients. The most important aim of this therapy is to establish the value of lexico-semantic therapy in improving the communication abilities of aphasic patients.

REFERENCES

Bastiaanse R, Bosje M, & Visch-Brink EG, Psycholinguïstische Testbatterij voor de Taalverwerking van Afasiepatiënten. Een bewerking van J Kay, R Lesser en M Coltheart. Psycholinguistic Assessments of Language Processing in Aphasia, 1995. Hove : Lawrence Erlbaum Associates.

Blomert L, Kean M-L, Koster Ch, et Schokker J. Amsterdam-Nijmegen Everyday Language Test: Construction, reliability and validity. Aphasiology 1994 ; 8 : 381-407.

Butterworth B, Howard D, et McLoughlin P. The semantic deficit in aphasia : the relationship between semantic errors in auditory comprehension and picture naming. Neuropsychologia 1984; 22: 409-426.

Cubelli R, Foresti A, et Consolini T. Reeducation strategies in conduction aphasia. Journal of Communication Disorders 1988 ; 21-249.

Franklin SE. Researching the treatment of anomia: the case for single cases. In FJ Stachowiak et al. (Eds.), Developments in the Assessment and Rehabilitation of Brain-Damaged Patients, 1993. Tübingen Gunter Narr Verlag.

Joanette Y. et Goulet P. Narrative discourse in right-brain-damaged right-handers. In Y Joanette and HH Brownell (Eds.), Discourse ability and brain damage, 1990. New York, Hong Kong : Springer Verlag.

Le Dorze G, Boulay N, Gaudreau J. et Brassard C. The contrasting effects of a semantic versus a formal-semantic technique for the facilitation of naming in a case of anomia. Aphasiology 1994 ; 8, 127-143.

Van Rijn M, Booy L, et Visch-Brink EG. FIKS, een fonologisch therapieprogramma voor afatische patiënten, in press ; Lisse : Swets & Zeitlinger.

Visch-Brink EG, Bajema IM, et Van de Sandt-Koenderman ME. Lexical semantic therapy: BOX. Aphasiology 1997; 11: 1057-1078.

Visch-Brink EG. Reply, Let’s do semantics. Wanted: an experienced therapist. Aphasiology 1997 ; 11 : 1106-1115.

Visch-Brink EG, Denes G, et Stronks D. Visual and verbal processing in aphasia. Brain and Language 1996 ; 55 : 130-132.

Visch-Brink EG, Denes G, et Stronks D. De Semantische Associatie Test. Swets & Zeitlinger, in press.

____________________________________________________________

_____________________________________________________

_________________________________________________________

_____________________________________________

___________________________________________________