192 CORRESPONDENCE

delusional conditions with a hypochondriacal con Rogers and Clay made no effort to test statistically tent may respond, often dramatically, to treatment the hypothesis that imipramine is more effective than with the diphenylbutylpiperidine agent pimozide in placebo from the combined evidence of all 30 studies. a dosage of 2—8mgm daily, where a variety of Cochran (‘954) provides a method for combining treatment regimes has previously failecL It appears statistics from a number ofindependent trials so as to that this group is identified by the absence of a come out with an overall statement that the proba primary depressive illness,and that the delusional bility observed in the combined trials differs from content is probably related to a paranoid-type that which would be expected by chance alone. process of a schizophrenic nature. We agree with Using the data provided by Rogers and Clay, the Riding and Munro that dysmorphophobic (neurotic) probability that the differences between imipramine conditions do not respond to thisapproach. Dermato and placebo could be observed by chance alone is @ logical hypochondriasis, as described by Zaidens (f), less than This would seem to be sufficient is specifically mentioned by Bebbington and in evidence that tricyclics are really superior to placebo eluded in his poor-prognosis group. This is not an in treating depression, although, it is obvious to uncommon disorder, and where such conditions may anyone in the field that much more needs to be be classified as non-dysthymic delusional parasitosis. learned about the clinical use of tricyclics. Reilly, Jopling and Beard (@) and Riding and Kerry and Orme assert that Rogers and Clay Munro (2) have demonstrated that remarkable analysed only a small proportion of the published improvement may be forthcoming following treat trials on antidepressants and that there are many ment with pimozide. trials in which a placebo has achieved a better result It is difficult to decide whether either of Bebbing than an antidepressant, but that these trials were not ton's cases could be considered as properly delu included. We have reviewed the same studies on sional, though Case 2 sounds probable. We have imiprasnine and our tally checks almost exactly with noted that the personality in this case exhibited that of Rogers and Clay (Davis, ig@o; Davis et al, obsessional traits, and it is our impression that such 1969). In our analysis we did not uncover the ‘¿many traits are commonly present in patients who respond trials in which placebo has achieved a better result favourably to pimozide. than an antidepressant' which Kerry and Orme TERENCE M. REILLY claim were not included in the analysis ofRogers and A. W. BEARD Clay. In fact, we did not find that among studies Dept of P.@yc/zologwa1Medicine, which followed strict research design there were any The Middlesex Hospital, that reported placebo significantly more effective London, W.i than tricyclic antidepressants. There are, of course, many open trials of imipramine which report REFERENCES imipramine to be effective, and there are both open I.BEBBINGTON, P. E. (1976) British Journal of Psychiat,y, and placebo-controlled random assignment trials za8, 475—8. with other tricycle antidepressants which also show 2. RHIHIG, B. E. J. & Mwnto, A. (ig7@) Lancet, i, 4oo. that these tricyclics are effective agents in the treat @ 3. REILLY,T. M. (i@7@)Lancet,i, 4. ZAIDENS,S.H. (1950) PsychosomaticMedicine,12, 250—3 ment of depression. Both these sources of evidence 5. Rutiux, T. M., Jo@uuo, W. H. & Bww, A. W. would further support the efficacy of tricyclics. To be published. Jom@iM. DAVIS STEPHEN E. ERICKSEN IllinoisStatePsychiatricInstitute, Chicago,Illinois, USA CONTROLLED TRIALS OF IMIPRAMINE DEAR Sm, REFERENCES DrsRogersand Clay (Journal, Dec. :975, 127, p 599) COCHRAN, W. J. (,g@) Some methods of strengthening reviewed 30 studies comparing imipramine to placebo the Common x2 test. Biometrics,10, 417—5:. and noted that the great majority of them show DAvis, J. M. (1970) Treatment of affective disorders in imipramine to be significantly superior to placebo. clinical practice. In Changing Patterns in Psychiatric Care (ed. T. Rothman), pp 197—228. New York: Kerry and Orme (Journal, March 1976, 128, p 310) Crown PublishingCo.

have questioned this interpretation and say that it —¿ KLERMAN, G. L. & ScrnLDIutAuT,J.J. (1969) Drugs would be unfortunate if the results of this particular used in the treatment of depression. Psychopharmacology: statistical review were accepted uncritically as A Review of Progiess 1957—1967 (ed. D. Efrom), evidence that imipramine is therapeutically effective. pp 1057-64. US Government Printers Office.

Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 21:30:17, subject to the Cambridge Core terms of use.