Prognostic Studies

For many practical purposes it is advantageous to frame operational rather than theoretical definitions. As Guze has emphasized, a diagnosis has an essentially predictive function. If we make a diagnosis of "hysteria " on a group of patients we must expect them to show as a group some uniformity in the later course of the condition, and in the aetiological factors uncovered in the course of time. Follow-up studies can accordingly be expected to provide a useful check. The follow-up studies of "hysteria" show that the later state of patients so classified depends on the mode of selection-that is, on the nature of the operational definition.

    Thus the two authors who selected patients by the presence of conversion symptoms in an uncomplicated setting both found favourable recovery rates. Out of the 90 patients with recent onset whom Carter (1949) succeeded in contacting after a lapse of four to six years, 70% were well, and all but seven of them were able to work. However, two of these patients had become schizophrenic. Ljungberg (1957) made a longer follow-up, and the results were not quite so favourable. After a year 62% of his patients were symptom-free, and of those who were not two-thirds were still able to work. However, after the first year, in which most of the recoveries took place, subsequent progress was very slow; after five years a quarter of the patients were still suffering from symptoms, and after 15 years one-fifth. In the course of time 3.1% of his propositi became schizophrenic, 2.4 % had a manic-depressive psychosis, and 3.3% developed epilepsy. Similar findings were made in a small study by Gatfield and Guze (1962). During a follow-up period of 3 to 10 years 4 out of 24 patients with conversion symptoms developed clear signs of neurological disease (tabes, motor paresis, basal ganglia disease, cerebral tumour), with an additional case in which the seizures proved to be focal. These authors concluded that conversion symptoms arise with a variety of psychiatric and neurological conditions as cause; and that the prognosis is not good, many patients becoming chronic.

    Similar views have been based on investigations along purely clinical lines, without follow-up. Chodoff and Lyons (1958) studied a carefully sifted group of patients with conversion reactions, from which all neurological disorders had been excluded. The personalities of these patients were of all kinds: passive-aggressive, emotionally unstable, inadequate, schizoid, paranoid, etc. ; but there was only one example of the hysterical personality, as this term is commonly understood. The authors make the penetrating comment that the traditional description of the "hysterical personality" is a description of women in the words of men, and that it is a caricature of femininity. Ziegler, Imboden, and Meyer (1960) also found that only a minority of patients with conversion reactions showed "hysterical" personalities. On the other hand, they noticed a striking association with depression; 30% of their patients had depressive symptoms, and when they examined a sample of 100 depressives they found conversion symptoms in 28. Stephens and Kamp (1962) also found no predominance of hysterical personalities in patients diagnosed as suffering from hysterical and dissociative syndromes: they thought the commonest personality deviation was of the passive-dependent type.

    For research purposes there are great advantages in leaving the selection of cases to the arbitrary decisions of others. "Hysterics " selected in this way prove to be extremely heterogeneous. Ziegler and Paul (1954) took all the women who had been diagnosed as suffering from "psychoneurosis hysteria" at the Boston Psychopathic Hospital from 1927 to 1932. There were 66 of these women, and 22 had been readmitted to hospital later with a psychotic diagnosis, including 12 diagnoses of dementia praecox, 9 of manic-depressive psychosis, and 2 of organic psychosis. The authors conclude: "These cases were thought to have a kind of uniformity twenty-five years ago; now the extreme diversity in mental status raises the question whether the given criteria are not so imprecise as to be nonfunctional; whether they do not create the illusion of defining an entity where there is none."

    A similar conclusion was reached by me (Slater, 1961) on a sample of 24 pairs of twins from the Bethlem and Maudsley Hospitals. The clinical, genetical, and prognostic aspects of this study all converged to indicate that the illnesses classified as hysterical under section 311 of the International Classification of Diseases were heterogeneous. Errors of diagnosis were prominent. The "hysterics" in this material included patients suffering from focal brain lesions, epilepsy, schizophrenia, endogenous depression, and anxiety states. Pyschogenic and physical causative factors were varied and non-specific. No specific genetical factor was found to play a part. Whatever may have been the case at the time of diagnosis, no unifying thread of any kind could be found with which this group of patients could be tied together a few years later.