Rapid Responses to:

PAPERS: Helen L Crimlisk, Kailash Bhatia, Helen Cope, Anthony David, C David Marsden, and Maria A Ron Slater revisited: 6 year follow up study of patients with medically unexplained motor symptoms BMJ 1998; 316: 582-586 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Paroxysmal neurological disorders in medically unexplained motor dysfunctions

K R Nair   (26 December 1999)

Revisiting Slater revisited:

David PJ Osborn, Vicky J Holt, SHO in liaison psychiatry.   (2 October 2002)

Paroxysmal neurological disorders in medically unexplained motor dysfunctions 26 December 1999
K R Nair, consultant neurologist,Cosmopolitan Hospital, Trivandrum Send response to journal: Re: Paroxysmal neurological disorders in medically unexplained motor dysfunctions	Sir, This interesting series of cases suggest that the clinicians should be aware of the possible late occurence of the neurological disorders in patients with unexplained motor symptoms. Many of these patients are labelled as primary psychiatric problems by General practitioners, psychiatrists and even neurologists. In such a series studied by us ( not published ) we found that paroxysmal neurologic syndromes do occur in about 4-5% of cases. They are Paroxysmal kinesigenic dystonia, Lance's syndrome and occasionally dystonia with diurnal variation. Most of these cases are in very young patients. Young onset Parkinson's disease some times manifest intially as anxiety induced tremors only. Dr.K.R.Nair, Trivandrum
Revisiting Slater revisited: 2 October 2002
David PJ Osborn, MRC Research Fellow & Honorary SpR in Psychiatry. Dept of Psychiatry and Behavioural Sciences,RFUCMS, UCL, NW3 2PF, Vicky J Holt, SHO in liaison psychiatry. Send response to journal: Re: Revisiting Slater revisited:	We recently critically appraised this well known paper, while looking for the best evidence regarding the prognosis of a clinical case of conversion disorder. We were surprised to note a couple of errors in the paper, which apparently remain uncorrected, four years on. One odds ratio (OR) in the paper shows an effect opposite to the conclusions, whilst another makes no sense. The odds ratios for “getting better” from medically unexplained symptoms, in the main results table (4), are incorrect for the prognostic factors of both age; “adjusted OR: 0.73, (95% CI: 5.72-4.14)” and shorter duration of symptoms “unadjusted OR 0.24, (0.08-0.74), adjusted OR 0.11 (0.02-0.67)”. An adjusted OR of 0.11 would, of course, indicate that shorter symptom duration gave a worse outcome. However, if the raw numbers in the table are correct, having symptoms for less than a year gives a positive unadjusted odds ratio of 7.17 (1.81- 30.87). The latter result concurs with the conclusion of the text, that a shorter duration of symptoms is a good prognostic factor in patients with medically unexplained symptoms. We wonder how much attention the table itself has received during the publication process and on the occasions when the paper has been cited subsequently . The broad confidence intervals for a number of the prognostic factors also stand out in the table, but have attracted no comment. For instance, a change in marital status resulted in a fully adjusted OR of 33.66; (2.52-444.61), and co morbid psychiatric disorder gave an OR of 7.34; (1.29-42.28). These wide ranges, resulting from the small numbers studied, afford us little certainty about the true effect size of these prognostic factors. The errors do not affect the paper’s main conclusions, which remain robust. They were, however, important to us when trying to use the paper to estimate the prognosis of our patient. The lesson is clearly that we must take little for granted when we appraise literature, however influential the paper or journal.