cluster analysis of clinical semiology of psychogenic

Transkrypt

cluster analysis of clinical semiology of psychogenic
13
Epileptologia, 2007, 15: 13-28
CLUSTER ANALYSIS OF CLINICAL SEMIOLOGY OF
PSYCHOGENIC PSEUDOEPILEPTIC SEIZURES
ANALIZA KLASTROWA KLINICZNEJ SEMIOLOGII
PSYCHOGENNYCH NAPADÓW RZEKOMOPADACZKOWYCH
Joanna Jędrzejczak
Department of Neurology and Epileptology, Medical Centre for Postgraduate Education
00-416 Warsaw, Czerniakowska 231 Str.
SUMMARY
Wprowadzenie. Wraz z rozwojem długotrwałego
monitorowania wideo-EEG zaobserwowano, że
psychogenne napady rzekomopadaczkowe
(PNR) występują znacznie częściej niż wcześniej
uważano. Kliniczne rozróżnienie napadów padaczkowych (NP) i psychogennych napadów rzekomopadaczkowych często jest trudne. Ponadto, u tych samych pacjentów NP mogą często
współwystępować z PNR.
Cel. W celu wyjaśnienia mylącego obrazu klinicznego różnych rodzajów napadów porównano kliniczne cechy napadów u pacjentów z PNR i NP.
Celem pracy było opracowanie semiologicznej
klasyfikacji PNR na podstawie analizy wideometrycznej zarejestrowanych napadów.
Materiał i metoda. Analizowano kliniczny przebieg 312 PNR u 125 pacjentów z PNR. Wyróżniono 40 cech morfologicznych składających się na
obraz PNR. W celu identyfikacji objawów występujących razem w sposób systematyczny zastosowano analizę klastrową z hierarchiczną metodą grupowania mniejszych klastrów w większe.
Wyniki. Analiza klastrowa wykazała, że pewne
cechy współwystępują w sposób systematyczny.
Wyodrębniono cztery klastry. Klaster 1. (PS1) charakteryzuje się złożoną, gwałtowną symptomatologią ruchową. Klaster 2. (PS2) obejmuje pacjentów o prostej symptomatologii ruchowej. Klaster
3. (PS3) obejmuje pacjentów z dominującym drżeniem kończyn i całego ciała. Klaster 4. (PS4) charakteryzuje się objawami bezruchu. Porównanie
PNR w grupie pacjentów z PNR i grupie pacjentów z napadami mieszanymi ujawniło duże różnice semiologiczne.
Wnioski. Analiza klastrowa jest cennym i obiektywnym narzędziem do opracowania klasyfikacji
PNR opartej na całościowym obrazie klinicznym.
Introduction. With the growth of intensive EEG-video monitoring, it became apparent that psychogenic pseudoepileptic seizures (PPES) are
more common than was previously believed. Clinical differentiation between epileptic seizures
(ES) and PPES is often difficult. Moreover, ES and
PPES often coexist in the same patient.
Objective. To clarify the confusing clinical presentation of these different seizure types, the clinical
ictal characteristics of patients with ES and PPES
were compared. The purpose of the study was to
develop a semiological classification of psychogenic pseudoepileptic seizures based on analysis of video-EEG monitoring.
Material and methods. Clinical seizure semiology of 312 PPES in 125 patients with psychogenic
seizures was analysed. Forty morphological symptoms of the PPES clinical picture were distinguished. Cluster analysis was performed to identify systematically co-occurring symptoms. The
hierarchic agglomerative technique was applied,
i.e., symptoms were combined into clusters.
Results. Cluster analysis demonstrated that some
symptoms co-occur in a systematic way. Four
clusters were identified. Cluster 1 (PS1) is characterized by complex, violent motor symptomatology. Cluster 2 (PS2) represents patients with simple motor symptomatology. Cluster 3 (PS3) represents seizures with trembling of the limbs and
the entire body. Cluster 4 (PS4), motionless type,
is characterized by a single symptom, absence
of motor phenomena. Comparison of PPES in
a group of patients with PPES and a group of patients with mixed seizures revealed major semiological differences.
Conclusions. Cluster analysis is a valuable and
objective tool for classifying PPES on the basis of
Praca wpłynęła 27.11.2006 r.
Received November 27, 2006.
PAPERS • PRACE
STRESZCZENIE
14
Joanna Jędrzejczak
Może być ona pomocna w lepszym scharakteryzowaniu PNR i umożliwić opracowanie klasyfikacji etiologicznej.
the overall clinical semiology. It can help to improve
the characterisation of PPES and to develop an
etiological classification of the PPES phenomenon.
Słowa kluczowe: Psychogenne napady rzekomopadaczkowe – Klasyfikacja – Semiologia napadów
Key words: Psychogenic pseudoepileptic seizures – Classification – Semiology of seizures
INTRODUCTION
PAPERS • PRACE
A
lthough several researchers have tried
to emphasize the differences between
psychogenic pseudoepileptic seizures and
epileptic seizures by developing lists of differential symptoms (Molder, 1990; Lesser
i Krauss 1993; Porter, 1993), unfortunately there is no single clinical characteristic
which could enable to distinguish between
psychogenic seizures and epileptic seizures.
Moreover, everything that can be found in
epilepsy can also characterize nonepileptic seizures and vice versa.
One must therefore always base on
a group of features rather than on one isolated feature.
Selective use of one or two of a list of
symptoms may result in a swift diagnosis
but it may also result in unwilling false
positive diagnoses. There is no classical
diagnostic paradigm for psychogenic pseudoepileptic seizures and therefore diagnosis is still based on exclusion.
So there is the need to develop a comprehensive description and classification of
PPES on the basis of a group of features.
By analyzing PPES morphology we should
be able to develop a semiological classification system which takes into account the
typical characteristics of such seizures.
OBJECTIVE
T
he purpose of this study is to develop
a semiological classification of psychogenic pseudoepileptic seizures on the
basis of video-EEG monitoring of the seizures. Second goal is to compare semiology of psychogenic seizures in a group of
patients only with psychogenic seizures
and a group of patients with mixed seizures
(PPES and epileptic seizures) Cluster analysis is used to determine shared and nonshared features in these two groups.
MATERIAL AND METHOD
T
he study was run on 1353 epileptic
patients admitted consecutively to the
Department of Neurology and Epileptology, Medical Centre for Postgraduate Education, Warsaw in 1990-1999. The majority of the patients were referred as patients
with drug resistant epilepsy.
CLINICAL DESCRIPTION
All 1353 patients were thoroughly interviewed, examined neurologically and
submitted to additional tests (brain neuroimaging, interictal EEG and video-EEG).
Psychogenic pseudoepileptic seizures were
diagnosed in 125 patients (9.2%). There
were 100 women (80%) and 25 men (20%).
In this group, 90 patients (72%) had only
psychogenic pseudoepileptic seizures
(Group 1) and 35 patients (28%) had both
epileptic seizures and psychogenic pseudoepileptic seizures (Group II).
The mean age at admission in two
groups was 25 y. (15-52). The onset of psychogenic pseudoepileptic seizures was at
age 4-50 y. (M = 22) in Group I and at age