BRITISH MEDICAL JOURNAL
6 JANUARY 1979
intrauterine exchange transfusion. Now that would have been a triumph for television! Perhaps I should stay silent and not pillory the poor doctor, who, after all, was only playing the same part that we all play to our patients-that of appearing to know everything. But a more serious point arises. When I find errors in programmes about which I know something I worry about the errors in programmes about which I know nothing. Television is such a suggestive medium. How many people, I wonder, would swear that they saw on the television monitor the cannula entering the baby's vein and later saw the doctor syringing blood backwards and forwards to and from the fetus ? Yet it did not happen. Are we all suffering from Galen's delusion? He "saw" invisible pores in the septum of the heart in order to satisfy his view of the world. How many invisible pores in our view ? Perhaps that was the point that Dr Miller was trying to make. T J HAMBLIN Department of Pathology, Royal Victoria Hospital, Bournemouth, Dorset
A plea to all vasectomists
SIR,-Requests for reversing a vasectomy are relatively uncommon, but (as expected just from the greater number of vasectomies being performed) these requests are progressively increasing. Reconstruction after a properly performed vasectomy is an easy procedure with at least a 70% chance of restoring a good sperm count, though the motility and pregnancy rate may be reduced by sperm antibodies in some cases. It is insufficiently realised that the most important factor which determines the success or otherwise of a vasovasostomy is how the original vasectomy was performed. Excision of long lengths of the vasa to prevent spontaneous reunion is an entirely unnecessary mutilation and may make reconstruction impossible. A less well known fault in technique is to perform the vasectomy too low down so that it involves the convoluted and thinner part of the vas; this makes the reconstruction much more difficult and obviously greatly reduces the chance of success. Whenever possible a vasectomy should be performed about the level of the head of the epididymis where the vas is a thick, straight tube. No more than 1 cm should be excised for histological purposes; prevention of spontaneous reunion requires only the placing of the two ends in different tissue planes. W KEITH YEATES Department of Urology, Newcastle University Hospitals, Newcastle upon Tyne
Physical therapy in chronic bronchitis
SIR,-Drs D A G Newton and H G Bevans (2 December, p 1525) are to be congratulated on their thorough and extensive clinical study. It is all the more unfortunate that they have chosen to misuse the term "intermittent positive pressure ventilation (IPPV)," which is widely accepted as the term referring to artificial mechanical ventilation of the lungs via an endotracheal or tracheostomy tube. The term "intermittent positive pressure breathing (IPPB)" is surely preferable. This is used to describe a short period of treatment
55
during which air (or air plus oxygen), often together with water and/or drugs, is delivered to a patient via a mouthpiece from a patienttriggered nebuliser. An even more appropriate term to describe this treatment, in view of the findings of Drs Newton and Bevans, is "giving them the Bird"' Their use of the term "IPPV" in this study is thus careless, and those who only read the summaries of papers could be seriously misled. JULIAN M LEIGH
(4) Other workers3 have shown that wellplanned physiotherapy can remove secretions and reduce airflow obstruction. This study was done in patients producing over 30 ml of sputum a day. Most patients in the study of Drs Newton and Bevans produced relatively small volumes of sputum even during exacerbations of infection. One would expect patients with small amounts of sputum to benefit less from postural drainage and IPPB than those with copious sputum. We consider that in patients with acute exacerbations of chronic bronchitis, especially Intensive Care Unit, Royal Surrey County Hospital, those who have large volumes of secretions,4 Guildford careftully planned physiotherapy and the ***We sent a copy of this letter to Dr Newton, administration of a bronchodilator before physiotherapy by IPPB or nebuliser is an whose reply is printed below.-ED, BMJ. essential part of treatment.
SIR,-I am grateful to Dr Leigh for pointing MARGARET E HODSON out the ambiguity in our summary. UnforDIANA GASKELL tunately our original text mentioned "interBARBARA A WEBBER mittent positive pressure-ventilation" once J C BATTEN Brompton Hospital, and we had thereafter abbreviated this to London SW3 "IPPB"; this was subedited to "IPPV" and we allowed this to pass rather than changing Newton, D A G, and Stephenson, A, Lancet, 1978, 2, 530. "ventilation" to "breathing." 2 Shenfield, G M, et al, American Review of Respiratory Dr Leigh's suggested title would no doubt Diseases, 1973, 108, 501. G M, Webber, B A, and Clarke, S W, unleash a spate of correspondence from 3 Cochrane, British Medical Journal, 1977, 2, 1181. frustrated readers no longer able to write to 4Lancet, 1978, 2, 1241. The Times. I suspect the Editor would prefer the occasional V for a B and hope that readers of the VMJ will read their articles right Dexamethasone in acute stroke
through. D A G NEWTON St James's University Hospital, Leeds
Sm,-The trial reported by Dr Graham Mulley and his colleagues (7 October, p 994) and the subsequent correspondence (28 October, p 1230; 25 November, p 1500; 9 December, p 1639) raise a number of important points. We are grateful to Dr R G Wilcox for supplying additional details of the patients in the trial. (1) Trial design-(i) The authors attached no
***We must admit to having compounded the original error by introducing "intermittent positive-pressure ventilation" into the summary and into a revised version of the title of the paper, which originally read "Physical therapy in exacerbations of chronic bronch- importance to separating the stroke patients into itis." We apologise for our part in this con- defined diagnostic subgroups before treatment was fusion of terms.-ED, BM7. allocated. Yet to consider them all as a single SIR,-It was with some concern that we read the paper by Drs D A G Newton and H G Bevans (2 December, p 1525). We would like to make the following comments: (1) This study has not separated the effects of breathing exercises and postural drainage from those of "IPPV" (presumably intermittent positive pressure breathing was intended) and minimal details are given about the administration of either treatment. (2) Physiotherapy given in their "standard fashion"' refers to postural drainage for three minutes in each of four different positions. Some patients with severe obstructive chronic bronchitis would not tolerate the prone and supine positions without signs of respiratory distress. With frequent changes in position and no relaxed controlled diaphragmatic breathing patients participating in such a treatment regimen would be exhausted; bronchial secretions would be mobilised but incompletely cleared in the period of time allowed. (3) In this study the bronchodilator was not co-ordinated with the time of physiotherapy and the bronchodilator was given by pressurised aerosol. It has been shown that a bronchodilator is more efficient given by intermittent positive pressure breathing (IPPB) than by pressurised aerosol.' Indeed, one of the indications for using IPPB in this type of patient is to give bronchodilator by a very effective means before physiotherapy.
patient group is illogical because the natural histories differ for infarction and haemorrhage and for cerebral and brainstem lesions. As the literature already indicates that dexamethasone is not effective in unselected acute strokes this further trial would have been better aimed at determining whether a particular patient subgroup will benefit or whether the effect of dexamethasone is related to the timing of administration. (ii) Importance was similarly not attached to stratifying patients according to level of consciousness; dependence was placed instead on random allocation. As a result twice as many patients in the dexamethasone group were fully alert and twice as many in the placebo group were responding only to pain; this prejudiced the placebo group from the outset of the study. (iii) Dr Mulley and his colleagues are to be congratulated on obtaining speedy admission (mean 5j h), but the delay in initiating treatment was different in the two groups-namely, 9 h in the placebo group and 12 h in the dexamethasone group. This difference would appear to prejudice the dexamethasone group and is difficult to explain in a double-blind trial. (2) Neurological scoring systems-Many stroke trials use different systems to "score" neurological deficit which, although claimed to be easy to use by the authors, never appear to be reproducible by other centres. As in this trial, the total score is usually obtained by adding separate scores for a number of variables; thus even when two total scores are the same the contributing subscores may be quite dissimilar. It is therefore both statistically unsound and misleading to say that the two groups were well matched on neurological scores at admission. The only way to obviate this difficulty is to define and analyse each parameter separately.'
In conclusion, the trial reported by Dr
BRITISH MEDICAL JOURNAL
56
Mulley and his colleagues has given no further guidance in the use of dexamethasone in any stroke patient. Whether dexamethasone is of value in selected cases of acute cerebral infarction has yet to be determined. F CLIFFORD ROSE RuDY CAPILDEO TIMOTHY STEINER Department of Neurology, Charing Cross Hospital, London W6
Capildeo, R, and Rose, F C, in Progress in Stroke Research (1), ed R M Greenhalgh and F Clifford Rose. London, Pitman Medical. In press.
phenomenon is that, if true, it completely contradicts basic, longstanding, conventional medical teaching on tuberculous meningitis. Before accepting this as just another example of how badly wrong they got things in "the old days" I thought that the supporting evidence for the report warranted at least cursory investigation. This proved disappointing. In the four cases reported by Emond and McKendrick there was no clinical evidence of tuberculous infection in three and in the fourth case, in which a previously negative Mantoux test became positive, no evidence of tuberculosis was found at necropsy 10 days after the institution of antituberculosis therapy. In fact the only evidence of tuberculosis in any of the patients was that Mycobacterium tuberculosis was isolated from the cerebrospinal fluid (CSF). This might appear on the face of it pretty conclusive evidence, but the authors themselves raise the possibility of accidental contamination. They discount this on the grounds that for it to happen four times is too much of a coincidence. I would suggest that if a laboratory technique is faulty to the degree that accidental contamination is possible it is more likely rather than less likely that the contamination will be repeated. I think it was incumbent on these workers to demonstrate that no possibility of contamination existed to enable them to draw such a surprising conclusion. In no case was the Ziehl-Neelsen film from the CSF reported to contain acid/alcohol-fast bacilli, and if we discount the subsequent isolation of M tuberculosis from these specimens the obvious diagnosis in each case would have been virus meningitis. The depressed cellmediated immunity which accompanies virus infections and causes false-negative Mantoux tests2 would account for the transient negative Mantoux test in patient 4. The further claim that elevated IgM levels in case 2 were evidence of bacterial rather than virus meningitis was contradicted by Buchanan,3 who stated that it is not possible to distinguish between the various types of meningitis on the basis of immunoglobulin patterns. The final conclusion must be that spontaneous cure in tuberculous meningitis has not been demonstrated by the evidence presented in this paper. It is hardly in the best interests of medicine for an influential column such as a leading article in the BMJ to accept the findings of a publication and to endorse it by quotation without first casting a critical eye on the contents. The mere fact that a paper achieves publication does not mean that it automatically rates acceptance, never mind quotation.
SIR,-The report by Dr Graham Mulley and others (7 October, p 994) of the ineffectiveness of dexamethasone treatment in stroke confirms the findings of previous controlled doubleblind studies referred to in their paper. However, I would question the validity of clinical evaluation at one year as a test of the effectiveness of this therapy, although such late evaluation has been used in other therapeutic studies in stroke.' Mortality in the months following the ictus is an unreliable indicator of the effectiveness of treatment administered during the acute stage of stroke. The essential beneficial effect of dexamethasone in cerebral lesions is to reduce cerebral oedema, and this is the rationale for its use in acute cerebrovascular lesions. However, the cause of mortality from stroke in the first week differs from that in the ensuing three months. In our own series of 467 patients admitted to an acute stroke unit2 death in the first week was due to coning (that is, cerebral oedema) in 72% of those who expired, while the remainder died during the subsequent months of pulmonary or cardiac complications. The effect of steroid therapy, therefore, on these varying pathological states would be quite different. Since death in the first week is an effect of raised intracranial pressure mainly due to cerebral oedema, only death and possibly immediate morbidity in the acute phase are reliable markers of such a therapeutic effect. The late disability (at one year) represents a mixture of factors, including spasticity, the efficacy of rehabilitation, and the patient's motivation, all of which are factors largely impossible to match or measure. The similar disability in both drug and placebo groups may simply reflect the insensitivity of the method to evaluate these immeasurable variables. Finally, if steroids did in fact reduce cerebral oedema more severely brain-damaged patients might survive in the drug-treated Department of Pathology, group so that there would be more severely Victoria disabled patients in the treated group after Blackpool,Hospital, Lancs one year than in the placebo group.
JOHN W NORRIS MacLachlan Stroke Unit, University of Toronto Mathews W B, et al, Brain, 1976, 99, 193. 2Bril, V, .qorris, J W, and Hachinski, V C, Canadian J7ournal of Neurological Sciences, 1977, 4, 218.
Recovery from tuberculous meningitis SIR,-Your interesting leading article on virus meningitis (25 November, p 1451) referred to a report by Emond and McKendrick1 of four cases of spontaneous recovery from tuberculous meningitis. The significance of this surprising
J S CARGILL
Emond, R T D, and McKendrick, G D W, Lancet, 1973, 2, 234. 2 British Medical Journal, 1970, 4, 573. 3 Buchanan, N, Lancet, 1973, 2, 677.
6 JANUARY 1979
these four patients, all had clinical manifestations suggesting meningeal infection (though one was very mild), three had CSF lymphocytosis, and Mycobacterium tuberculosis was cultured from the CSF of all four patients. What further evidence of tuberculous meningitis would Dr Cargill like to have? The authors deal fully with the problem of accidental contamination being responsible for the positive cultures and consider that the coincidence of this happening in four similar cases would be most unlikely. On the evidence available the authors' conclusions are valid. If Dr Cargill prefers to believe that the four positive CSF cultures in these patients are artefacts, thus discounting the authors' claim, she must also explain Cramer and Bickel's 1922 account1 of 43 proved cases of tuberculous meningitis which resolved spontaneously. Macgregor and Green in 1937,2 reviewing the pathology of tuberculosis, concluded that meningitis may be arrested and undergo clinical cure even when the CSF has been infected. Serious meningitis is the term sometimes applied to this form of tuberculous meningitis, which has a good prognosis (including spontaneous recovery) and was well reviewed by Lincoln and Sewell in 1963.3 The fact that two of these reports were published in the pre-antibiotic era indicates that Dr Cargill's comment on "how badly wrong they got things in 'the old days' " is unjustified. -ED, BMJ7. Cramer, A, and Bickel, G, Annales de Medecine, 1922, 12, 226. 2Macgregor, A R, and Green, C A, J'ournal of Pathology and Bacteriology, 1937, 2, 613. 3Lincoln, E H, and Sewell, E M, Tuberculosis in Children. New York, McGraw-Hill, 1963.
Anaemia in Indian childhood cirrhosis
SIR,-Dr M S Tanner and his colleagues in their article on Indian childhood cirrhosis presenting in Britain with orcein-positive deposits in the liver and kidney (30 September, p 928) state that "anaemia is a recognised feature of Indian childhood cirrhosis, but it has not been reported to be haemolytic." I would like to bring it to their attention that haemolytic anaemia has in fact been reported in this type of cirrhosis,' though it was emphasised that the degree of anaemia was not always related to the severity of haemolysis. I wouild like to learn whether the levels of caeruloplasmin and reduced glutathione of erythrocytes were determined in this patient. SINASI OZSOYLU Department of Pediatric Medicine, Hacettepe University Children's Medical Center, Ankara, Turkey
Perkash, A, et al, Archives of Disease in Childhood, 1971, 46, 46.
***We sent a copy of this letter to Dr Tanner and his colleagues, whose reply is printed below.-ED, BM_J.
***Dr Cargill is unfortunately mistaken in her statement that the possibility of spontaneous recovery from tuberculous meningitis "completely contradicts basic, longstanding, conventional medical teaching on tuberculous meningitis." Spontaneous recovery in tuberculous meningitis is well documented though raresufficiently unusual for Emond and McKendrick to report four such instances in the paper quoted in the leading article. Of
SIR,-Although Perkash et al provided evidence for haemolysis in Indian children with cirrhosis, their data are difficult to interpret because the children studied were aged between 8 months and 12 years, whereas the entity "Indian childhood cirrhosis" occurs largely in the 1-3-year age group.' Furthermore, histological confirmation ofthe diagnosis was available in only 20 of 55 cases. Thus
6 JANUARY 1979
intrauterine exchange transfusion. Now that would have been a triumph for television! Perhaps I should stay silent and not pillory the poor doctor, who, after all, was only playing the same part that we all play to our patients-that of appearing to know everything. But a more serious point arises. When I find errors in programmes about which I know something I worry about the errors in programmes about which I know nothing. Television is such a suggestive medium. How many people, I wonder, would swear that they saw on the television monitor the cannula entering the baby's vein and later saw the doctor syringing blood backwards and forwards to and from the fetus ? Yet it did not happen. Are we all suffering from Galen's delusion? He "saw" invisible pores in the septum of the heart in order to satisfy his view of the world. How many invisible pores in our view ? Perhaps that was the point that Dr Miller was trying to make. T J HAMBLIN Department of Pathology, Royal Victoria Hospital, Bournemouth, Dorset
A plea to all vasectomists
SIR,-Requests for reversing a vasectomy are relatively uncommon, but (as expected just from the greater number of vasectomies being performed) these requests are progressively increasing. Reconstruction after a properly performed vasectomy is an easy procedure with at least a 70% chance of restoring a good sperm count, though the motility and pregnancy rate may be reduced by sperm antibodies in some cases. It is insufficiently realised that the most important factor which determines the success or otherwise of a vasovasostomy is how the original vasectomy was performed. Excision of long lengths of the vasa to prevent spontaneous reunion is an entirely unnecessary mutilation and may make reconstruction impossible. A less well known fault in technique is to perform the vasectomy too low down so that it involves the convoluted and thinner part of the vas; this makes the reconstruction much more difficult and obviously greatly reduces the chance of success. Whenever possible a vasectomy should be performed about the level of the head of the epididymis where the vas is a thick, straight tube. No more than 1 cm should be excised for histological purposes; prevention of spontaneous reunion requires only the placing of the two ends in different tissue planes. W KEITH YEATES Department of Urology, Newcastle University Hospitals, Newcastle upon Tyne
Physical therapy in chronic bronchitis
SIR,-Drs D A G Newton and H G Bevans (2 December, p 1525) are to be congratulated on their thorough and extensive clinical study. It is all the more unfortunate that they have chosen to misuse the term "intermittent positive pressure ventilation (IPPV)," which is widely accepted as the term referring to artificial mechanical ventilation of the lungs via an endotracheal or tracheostomy tube. The term "intermittent positive pressure breathing (IPPB)" is surely preferable. This is used to describe a short period of treatment
55
during which air (or air plus oxygen), often together with water and/or drugs, is delivered to a patient via a mouthpiece from a patienttriggered nebuliser. An even more appropriate term to describe this treatment, in view of the findings of Drs Newton and Bevans, is "giving them the Bird"' Their use of the term "IPPV" in this study is thus careless, and those who only read the summaries of papers could be seriously misled. JULIAN M LEIGH
(4) Other workers3 have shown that wellplanned physiotherapy can remove secretions and reduce airflow obstruction. This study was done in patients producing over 30 ml of sputum a day. Most patients in the study of Drs Newton and Bevans produced relatively small volumes of sputum even during exacerbations of infection. One would expect patients with small amounts of sputum to benefit less from postural drainage and IPPB than those with copious sputum. We consider that in patients with acute exacerbations of chronic bronchitis, especially Intensive Care Unit, Royal Surrey County Hospital, those who have large volumes of secretions,4 Guildford careftully planned physiotherapy and the ***We sent a copy of this letter to Dr Newton, administration of a bronchodilator before physiotherapy by IPPB or nebuliser is an whose reply is printed below.-ED, BMJ. essential part of treatment.
SIR,-I am grateful to Dr Leigh for pointing MARGARET E HODSON out the ambiguity in our summary. UnforDIANA GASKELL tunately our original text mentioned "interBARBARA A WEBBER mittent positive pressure-ventilation" once J C BATTEN Brompton Hospital, and we had thereafter abbreviated this to London SW3 "IPPB"; this was subedited to "IPPV" and we allowed this to pass rather than changing Newton, D A G, and Stephenson, A, Lancet, 1978, 2, 530. "ventilation" to "breathing." 2 Shenfield, G M, et al, American Review of Respiratory Dr Leigh's suggested title would no doubt Diseases, 1973, 108, 501. G M, Webber, B A, and Clarke, S W, unleash a spate of correspondence from 3 Cochrane, British Medical Journal, 1977, 2, 1181. frustrated readers no longer able to write to 4Lancet, 1978, 2, 1241. The Times. I suspect the Editor would prefer the occasional V for a B and hope that readers of the VMJ will read their articles right Dexamethasone in acute stroke
through. D A G NEWTON St James's University Hospital, Leeds
Sm,-The trial reported by Dr Graham Mulley and his colleagues (7 October, p 994) and the subsequent correspondence (28 October, p 1230; 25 November, p 1500; 9 December, p 1639) raise a number of important points. We are grateful to Dr R G Wilcox for supplying additional details of the patients in the trial. (1) Trial design-(i) The authors attached no
***We must admit to having compounded the original error by introducing "intermittent positive-pressure ventilation" into the summary and into a revised version of the title of the paper, which originally read "Physical therapy in exacerbations of chronic bronch- importance to separating the stroke patients into itis." We apologise for our part in this con- defined diagnostic subgroups before treatment was fusion of terms.-ED, BM7. allocated. Yet to consider them all as a single SIR,-It was with some concern that we read the paper by Drs D A G Newton and H G Bevans (2 December, p 1525). We would like to make the following comments: (1) This study has not separated the effects of breathing exercises and postural drainage from those of "IPPV" (presumably intermittent positive pressure breathing was intended) and minimal details are given about the administration of either treatment. (2) Physiotherapy given in their "standard fashion"' refers to postural drainage for three minutes in each of four different positions. Some patients with severe obstructive chronic bronchitis would not tolerate the prone and supine positions without signs of respiratory distress. With frequent changes in position and no relaxed controlled diaphragmatic breathing patients participating in such a treatment regimen would be exhausted; bronchial secretions would be mobilised but incompletely cleared in the period of time allowed. (3) In this study the bronchodilator was not co-ordinated with the time of physiotherapy and the bronchodilator was given by pressurised aerosol. It has been shown that a bronchodilator is more efficient given by intermittent positive pressure breathing (IPPB) than by pressurised aerosol.' Indeed, one of the indications for using IPPB in this type of patient is to give bronchodilator by a very effective means before physiotherapy.
patient group is illogical because the natural histories differ for infarction and haemorrhage and for cerebral and brainstem lesions. As the literature already indicates that dexamethasone is not effective in unselected acute strokes this further trial would have been better aimed at determining whether a particular patient subgroup will benefit or whether the effect of dexamethasone is related to the timing of administration. (ii) Importance was similarly not attached to stratifying patients according to level of consciousness; dependence was placed instead on random allocation. As a result twice as many patients in the dexamethasone group were fully alert and twice as many in the placebo group were responding only to pain; this prejudiced the placebo group from the outset of the study. (iii) Dr Mulley and his colleagues are to be congratulated on obtaining speedy admission (mean 5j h), but the delay in initiating treatment was different in the two groups-namely, 9 h in the placebo group and 12 h in the dexamethasone group. This difference would appear to prejudice the dexamethasone group and is difficult to explain in a double-blind trial. (2) Neurological scoring systems-Many stroke trials use different systems to "score" neurological deficit which, although claimed to be easy to use by the authors, never appear to be reproducible by other centres. As in this trial, the total score is usually obtained by adding separate scores for a number of variables; thus even when two total scores are the same the contributing subscores may be quite dissimilar. It is therefore both statistically unsound and misleading to say that the two groups were well matched on neurological scores at admission. The only way to obviate this difficulty is to define and analyse each parameter separately.'
In conclusion, the trial reported by Dr
BRITISH MEDICAL JOURNAL
56
Mulley and his colleagues has given no further guidance in the use of dexamethasone in any stroke patient. Whether dexamethasone is of value in selected cases of acute cerebral infarction has yet to be determined. F CLIFFORD ROSE RuDY CAPILDEO TIMOTHY STEINER Department of Neurology, Charing Cross Hospital, London W6
Capildeo, R, and Rose, F C, in Progress in Stroke Research (1), ed R M Greenhalgh and F Clifford Rose. London, Pitman Medical. In press.
phenomenon is that, if true, it completely contradicts basic, longstanding, conventional medical teaching on tuberculous meningitis. Before accepting this as just another example of how badly wrong they got things in "the old days" I thought that the supporting evidence for the report warranted at least cursory investigation. This proved disappointing. In the four cases reported by Emond and McKendrick there was no clinical evidence of tuberculous infection in three and in the fourth case, in which a previously negative Mantoux test became positive, no evidence of tuberculosis was found at necropsy 10 days after the institution of antituberculosis therapy. In fact the only evidence of tuberculosis in any of the patients was that Mycobacterium tuberculosis was isolated from the cerebrospinal fluid (CSF). This might appear on the face of it pretty conclusive evidence, but the authors themselves raise the possibility of accidental contamination. They discount this on the grounds that for it to happen four times is too much of a coincidence. I would suggest that if a laboratory technique is faulty to the degree that accidental contamination is possible it is more likely rather than less likely that the contamination will be repeated. I think it was incumbent on these workers to demonstrate that no possibility of contamination existed to enable them to draw such a surprising conclusion. In no case was the Ziehl-Neelsen film from the CSF reported to contain acid/alcohol-fast bacilli, and if we discount the subsequent isolation of M tuberculosis from these specimens the obvious diagnosis in each case would have been virus meningitis. The depressed cellmediated immunity which accompanies virus infections and causes false-negative Mantoux tests2 would account for the transient negative Mantoux test in patient 4. The further claim that elevated IgM levels in case 2 were evidence of bacterial rather than virus meningitis was contradicted by Buchanan,3 who stated that it is not possible to distinguish between the various types of meningitis on the basis of immunoglobulin patterns. The final conclusion must be that spontaneous cure in tuberculous meningitis has not been demonstrated by the evidence presented in this paper. It is hardly in the best interests of medicine for an influential column such as a leading article in the BMJ to accept the findings of a publication and to endorse it by quotation without first casting a critical eye on the contents. The mere fact that a paper achieves publication does not mean that it automatically rates acceptance, never mind quotation.
SIR,-The report by Dr Graham Mulley and others (7 October, p 994) of the ineffectiveness of dexamethasone treatment in stroke confirms the findings of previous controlled doubleblind studies referred to in their paper. However, I would question the validity of clinical evaluation at one year as a test of the effectiveness of this therapy, although such late evaluation has been used in other therapeutic studies in stroke.' Mortality in the months following the ictus is an unreliable indicator of the effectiveness of treatment administered during the acute stage of stroke. The essential beneficial effect of dexamethasone in cerebral lesions is to reduce cerebral oedema, and this is the rationale for its use in acute cerebrovascular lesions. However, the cause of mortality from stroke in the first week differs from that in the ensuing three months. In our own series of 467 patients admitted to an acute stroke unit2 death in the first week was due to coning (that is, cerebral oedema) in 72% of those who expired, while the remainder died during the subsequent months of pulmonary or cardiac complications. The effect of steroid therapy, therefore, on these varying pathological states would be quite different. Since death in the first week is an effect of raised intracranial pressure mainly due to cerebral oedema, only death and possibly immediate morbidity in the acute phase are reliable markers of such a therapeutic effect. The late disability (at one year) represents a mixture of factors, including spasticity, the efficacy of rehabilitation, and the patient's motivation, all of which are factors largely impossible to match or measure. The similar disability in both drug and placebo groups may simply reflect the insensitivity of the method to evaluate these immeasurable variables. Finally, if steroids did in fact reduce cerebral oedema more severely brain-damaged patients might survive in the drug-treated Department of Pathology, group so that there would be more severely Victoria disabled patients in the treated group after Blackpool,Hospital, Lancs one year than in the placebo group.
JOHN W NORRIS MacLachlan Stroke Unit, University of Toronto Mathews W B, et al, Brain, 1976, 99, 193. 2Bril, V, .qorris, J W, and Hachinski, V C, Canadian J7ournal of Neurological Sciences, 1977, 4, 218.
Recovery from tuberculous meningitis SIR,-Your interesting leading article on virus meningitis (25 November, p 1451) referred to a report by Emond and McKendrick1 of four cases of spontaneous recovery from tuberculous meningitis. The significance of this surprising
J S CARGILL
Emond, R T D, and McKendrick, G D W, Lancet, 1973, 2, 234. 2 British Medical Journal, 1970, 4, 573. 3 Buchanan, N, Lancet, 1973, 2, 677.
6 JANUARY 1979
these four patients, all had clinical manifestations suggesting meningeal infection (though one was very mild), three had CSF lymphocytosis, and Mycobacterium tuberculosis was cultured from the CSF of all four patients. What further evidence of tuberculous meningitis would Dr Cargill like to have? The authors deal fully with the problem of accidental contamination being responsible for the positive cultures and consider that the coincidence of this happening in four similar cases would be most unlikely. On the evidence available the authors' conclusions are valid. If Dr Cargill prefers to believe that the four positive CSF cultures in these patients are artefacts, thus discounting the authors' claim, she must also explain Cramer and Bickel's 1922 account1 of 43 proved cases of tuberculous meningitis which resolved spontaneously. Macgregor and Green in 1937,2 reviewing the pathology of tuberculosis, concluded that meningitis may be arrested and undergo clinical cure even when the CSF has been infected. Serious meningitis is the term sometimes applied to this form of tuberculous meningitis, which has a good prognosis (including spontaneous recovery) and was well reviewed by Lincoln and Sewell in 1963.3 The fact that two of these reports were published in the pre-antibiotic era indicates that Dr Cargill's comment on "how badly wrong they got things in 'the old days' " is unjustified. -ED, BMJ7. Cramer, A, and Bickel, G, Annales de Medecine, 1922, 12, 226. 2Macgregor, A R, and Green, C A, J'ournal of Pathology and Bacteriology, 1937, 2, 613. 3Lincoln, E H, and Sewell, E M, Tuberculosis in Children. New York, McGraw-Hill, 1963.
Anaemia in Indian childhood cirrhosis
SIR,-Dr M S Tanner and his colleagues in their article on Indian childhood cirrhosis presenting in Britain with orcein-positive deposits in the liver and kidney (30 September, p 928) state that "anaemia is a recognised feature of Indian childhood cirrhosis, but it has not been reported to be haemolytic." I would like to bring it to their attention that haemolytic anaemia has in fact been reported in this type of cirrhosis,' though it was emphasised that the degree of anaemia was not always related to the severity of haemolysis. I wouild like to learn whether the levels of caeruloplasmin and reduced glutathione of erythrocytes were determined in this patient. SINASI OZSOYLU Department of Pediatric Medicine, Hacettepe University Children's Medical Center, Ankara, Turkey
Perkash, A, et al, Archives of Disease in Childhood, 1971, 46, 46.
***We sent a copy of this letter to Dr Tanner and his colleagues, whose reply is printed below.-ED, BM_J.
***Dr Cargill is unfortunately mistaken in her statement that the possibility of spontaneous recovery from tuberculous meningitis "completely contradicts basic, longstanding, conventional medical teaching on tuberculous meningitis." Spontaneous recovery in tuberculous meningitis is well documented though raresufficiently unusual for Emond and McKendrick to report four such instances in the paper quoted in the leading article. Of
SIR,-Although Perkash et al provided evidence for haemolysis in Indian children with cirrhosis, their data are difficult to interpret because the children studied were aged between 8 months and 12 years, whereas the entity "Indian childhood cirrhosis" occurs largely in the 1-3-year age group.' Furthermore, histological confirmation ofthe diagnosis was available in only 20 of 55 cases. Thus
