356 The agents for a double-blind trial are often prepared in a single batch. If this is the case, it is important, especially in a prolonged trial or one conducted in tropical climates, to examine the preparations again at the end of the study for their matching quality (as well as their drug content), procedures which we consider should be routine in any such trial. In the present investigation one pair of agents which were known to be a good colour match when freshly prepared developed, over a period of about 18 months, colour changes which made them readily distinguishable. On the other hand, if more than one batch is prepared for a trial, then each should be checked to ensure that the match is good within each batch. It is quite clear from the present investigation that it may be very difficult to produce indistinguishable preparations for use in "double-blind" trials and that, as a consequence, such studies are often not doubleblind at all. There is no alternative but to accept this fundamental shortcoming of the technique. We therefore recommend that in planning a controlled clinical trial there should be close collaboration from the outset between the investigator and the pharmaceutical firm. They should first discuss the study with particular reference to unbiased assessments of the results to see whether this can be achieved without recourse to the "double-blind" method. If the design of the study necessitates the method, which is likely to be the case when the assessment of the effect of treatment involves the recording of current symptoms, then they must consider the circumstances of the study and whether or not it is to be cross-over. In particular they must consider how easily the agents to be used can be matched. In this respect, Joyce4 states, "In not one of twenty clinical trials with which I have been connected have the first attempts at identical formulation
by the manufacturers been satisfactory." Furthermore, when the agents are prepared the matching properties should not be evaluated casually, but should be studied in a formal way along the lines of our study, as Joyce4 has also suggested. If a close match either is not possible or else has not been achieved, then the organisation of the study must take this into account and be adapted as far as is possible to ensure that the assessments remain unbiased. A suitable designs 9 may effectively conceal from the clinician and the patient what treatment the patient is receiving. The facts that stand out are that, unless the organisation of the study is appropriate, the interpretation of the results may remain in doubt, and that the "double-blind" study is an idol which may, and often does, have feet of clay. grateful to Dr T. B. Binns, Dr D. M. Burley, Dr S. Radhakrishna, and Dr I. Sutherland for helpful suggestions during the preparation of the report. We
are
REFERENCES 1. 2. 3. 4.
5. 6. 7. 8. 9.
Lionel, N. D. W., Herxheimer, A. Br. med. J. 1970, iii, 637. Reiffenstein, R. J., Schiltroth, A. J., Todd, D. M. Can. med.
Ass. J. 1968, 99, 1134. Mahon, W. A., Daniel, E. E. ibid 1964, 90, 565. Joyce, C. R. B. in Psychopharmacology Dimensions and Perspectives (edited by M. Balint); p.215. London, 1968. Fox, W., Robinson, D. K., Tall, R., Mitchison, D. A., Kent, P. W., MacFadyen, D. M. Tubercle, 1969, 50, 125. Freestone, D. S. Lancet, 1969, ii, 98. Tech. Rep. Scr. Wld Hlth Org. 1974, no. 536. Prineas, R. J. Med. J. Aust. 1971, ii, 425. British Medical Journal, 1970, iii, 597.
Letters
to
the Editor
SI UNITS IN MEDICINE
SIR,-The introduction of SI units has generated a great deal of correspondence in your columns. I am sure that Dr Ltster (Jan. 31, p. 244) is not alone in wishing to know the extent to which health authorities have complied with circular
HSC(IS)140. Health authorities were asked to let the Department have a short report on the situation within their regions as at Dec.1, 1975. There was no set format so that a number of the replies tended to be subjective. However, it would appear that by now most Regions have brought SI into widespread use. 1 region and in addition 3 areas and 9 districts are known to be behindhand for a variety of reasons, but expect to complete the exercise by September, 1976; a further 2 areas and 5 districts have made no arrangements as yet to introduce SI. Allied to the general resistance to the abolition of the mittimetre of mercury, an appreciable number of laboratories have not adopted the pascal for blood-gas measurements. A case is to be put to the E.E.C. in a few months’ time advocating the retention of the millimetre of mercury as a unit of measurement in clinical medicine. I fully appreciate the concern that has been demonstrated over the introduction of SI and hope that this information will go some way to assure those who are still uncommitted that elsewhere the National Health Service has virtually completed the changeover. Department of Health & Social Security, Alexander Fleming House, Elephant & Castle,
H. YELLOWLEES,
London SE1 6BY.
Chief Medical Officer
SiR,—More than
of the percentage of teaching have undergone willing or enforced change to SI, do we not need to know the effects? Does the change please or irritate clinicians, does it protect or endanger patients, facilitate or impair patient care? Does it clarify or confuse, help or hinder clinical diagnosis? Is there a consensus for an uncontrolled yet logical extension of molar reporting of, say, blood salicylate level to aspirin dosage? Do clinicians approve of SI in practice? Only a survey of opinion would reveal this and settle the mass-concentration molar and mm Hg/kilopascal controversies. Most clinicians would adopt a consensus view, but this has not been sought and is not expressed by the number of laboratories that have changed to SI. A. N. G. CLARK Brighton General Hospital, Elm Grove, Brighton BN2 3EW. JOANNA SHELDON and district
a statement
hospitals who
BASIC SCIENCES, SCIENCE, AND MEDICAL EDUCATION
SIR,-Professor Campbell’s article (Jan. 17, p. 134) is succinctly argued, lucid, and bold. He seems to be well aware that to talk about science always irritates medical audiences, and he must be prepared to accept the barrage that will inevitably follow. Surely he will be accused of "semantics" and Popperism Professor Campbell admits that science is not essential fOf health care. There the followers of Illich would agree, and their appeal for more common sense and humanitarianismis fair though one-sided. However, the introduction of more science in medical eduation would further undermine the pseudomilitary structure and discipline within the medical hierarchy held together t? power, authority, and tradition. Methods of scientific though: are powerful weapons against obscurantism and authef
357 tarianism. The student trained in scientific thought would be and would ask for hard data. Pharmaceutical industries would suffer financially because their claims would be rejected as profit biased. There are a few inherent snags and contradictions in teaching science in medical schools. First, somebody must teach it to teachers. Secondly, those who grasp complex problems of epistemology will become humble because they find that the definite answers they were looking for did not exist. Their newly acquired awareness of ignorance will be of no use in their role of doctor from whom patients will expect to have knowledge. Sir George Pickering’ summarised this beautifully: "From the time of Galen to our own ... medicine always presented a facade of systematic knowledge, or alleged knowledge, Infor, like religion, medicine could not tolerate ignorance tellectual nihilism is the very stuff of which scientists are made, but it is scarcely convenient for a practising physician." Thirdly, doctors are split by the dichotomy of wishing to be taken seriously as scientists and, at the same time, to retain their elitist power and pomp.23 While the scientist freely shares his knowledge and ignorance in discussion, the doctor keeps his patient in darkness by necessity or habit. While science can play with general hypotheses at leisure, the doctor must give an instant ad-hoc judgement or treatment for an individual patient. Another problem, clearly exposed by Professor Campbell, is essentialism in medical thought. Contemporary medicine continues to be based on an aristotelian classification of diseases. In the grey area of non-health, new nosological territories are being charted and named after their discoverers, as if they have existed since the creation. Beyond this area lies the Blessed Land of Normal, often referred to in the mythology of medicine but never mapped. The Classical Case is viewed as materialisation of the Idea or Essence of Disease, as if all diseases were disorders of kind and not of degree. One of the first critics of Professor Campbell, Dr West (Jan. 31, p. 242), says: "A clear description of a rash is just as much a part of science as a definition of a kilopascal ..." In fact, neither has anything to do with nature and processes of science with which Professor Campbell is concerned. The first is just a description (i.e., an observational statement) and the other is an arbitrary definition. It was Aristotle who believed that science consists of definitions and by defining everything we would know everything; but that was more than 2000 years
encouraged to question "experience"
...
"early" group were admitted within 9 h, and, though the sumthey were treated in this period, the details
mary states that given in the text
effects of
variance with this claim. At best, the cysteamine only be assessed in the six "early" and the results seem to confirm our earlier finding of are at
can
patients, almost complete protection against paracetamol hepatotoxicity.’ The Newcastle workers, however, seem reluctant to accept this, despite the fact that in these patients the maximum levels of aspartate aminotransferase (A.S.T.) and ferritin were significantly lower with less histological evidence of liver damage. In most of the comparisons made by Douglas et al. the data from the six "early" and the twelve "late" patients were pooled even though most of the late group were admitted and treated far too late for cysteamine to be effective. Their conclusions are not valid. The "paracetamol index" described by Douglas et al. is inappropriate and further confuses the issue. There is enormous individual variation in the plasma-paracetamol half-life after overdosage,6 and it is possible for the plasma concentration to be well above and well below the line in fig. 1 of their paper at different times in the same patient. The value obtained for the paracetamol index may thus depend more on the time that the patient is admitted after ingestion than on the severity of intoxication. Furthermore, it can be calculated that even if the plasma concentrations decline in parallel with the line in their fig. 1 the "paracetamol index" at any given time will itself decrease exponentially with a half-time of 3 h. We have treated nineteen severely poisoned patients with cysteamine given within 10 h (mean 6-8 h) of ingestion of paracetamol. All but one had a plasma-paracetamol well above the line in fig. 1 in the article by Douglas et al. None had significant hepatic or renal damage, and there were no deaths. In
PROTECTIVE EFFECT OF CYSTEAMINE GIVEN WITHIN
10 h OF
INGESTION
IN PATIENTS WITH SEVERE PARACETAMOL POISONING
ago. Mater Misericordiæ Dublin, Ireland.
Hospital,
PETR SKRABANEK
CYSTEAMINE FOR PARACETAMOL POISONING
S 1000 units/I), and three died in hepatic failure (see table). Furthermore, many of our treated patients were at risk because of alcoholism or previous consumption of drugs causing induction of hepatic microsomal enzymes 7 -an important aspect not considered by Douglas et al. Seven other patients who could not be given cysteamine for 11.5-26-5 h after overdosage all developed severe hepatic necrosis (mean A.S.T. > 4470 units/1). Our results with cysteamine given within 10 h to patients with severe paracetamol poisoning reinforce our previously expressed reservationsconcerning the ethics of controlled trials of cysteamine in this situation. We hope that the confusing report by Douglas et al. will not cause cysteamine to be withheld from patients at risk of severe liver damage or death after paracetamol overdosage. Regional Poisoning Treatment Centre, and University Department of Therapeutics, Royal Infirmary, Edinburgh EH3 9YW.
L. F. PRESCOTT
J. PARK A. T. PROUDFOOT
6. Prescott, L. F., Wright, N., Roscoe, P., Brown, S. S. ibid. 1971, i, 519. 7. Wright, N., Prescott, L. F. Scott. med. J. 1973, 18, 56. 8. Prescott, L. F., Matthew, H. Lancet, 1974, i, 998.
by the manufacturers been satisfactory." Furthermore, when the agents are prepared the matching properties should not be evaluated casually, but should be studied in a formal way along the lines of our study, as Joyce4 has also suggested. If a close match either is not possible or else has not been achieved, then the organisation of the study must take this into account and be adapted as far as is possible to ensure that the assessments remain unbiased. A suitable designs 9 may effectively conceal from the clinician and the patient what treatment the patient is receiving. The facts that stand out are that, unless the organisation of the study is appropriate, the interpretation of the results may remain in doubt, and that the "double-blind" study is an idol which may, and often does, have feet of clay. grateful to Dr T. B. Binns, Dr D. M. Burley, Dr S. Radhakrishna, and Dr I. Sutherland for helpful suggestions during the preparation of the report. We
are
REFERENCES 1. 2. 3. 4.
5. 6. 7. 8. 9.
Lionel, N. D. W., Herxheimer, A. Br. med. J. 1970, iii, 637. Reiffenstein, R. J., Schiltroth, A. J., Todd, D. M. Can. med.
Ass. J. 1968, 99, 1134. Mahon, W. A., Daniel, E. E. ibid 1964, 90, 565. Joyce, C. R. B. in Psychopharmacology Dimensions and Perspectives (edited by M. Balint); p.215. London, 1968. Fox, W., Robinson, D. K., Tall, R., Mitchison, D. A., Kent, P. W., MacFadyen, D. M. Tubercle, 1969, 50, 125. Freestone, D. S. Lancet, 1969, ii, 98. Tech. Rep. Scr. Wld Hlth Org. 1974, no. 536. Prineas, R. J. Med. J. Aust. 1971, ii, 425. British Medical Journal, 1970, iii, 597.
Letters
to
the Editor
SI UNITS IN MEDICINE
SIR,-The introduction of SI units has generated a great deal of correspondence in your columns. I am sure that Dr Ltster (Jan. 31, p. 244) is not alone in wishing to know the extent to which health authorities have complied with circular
HSC(IS)140. Health authorities were asked to let the Department have a short report on the situation within their regions as at Dec.1, 1975. There was no set format so that a number of the replies tended to be subjective. However, it would appear that by now most Regions have brought SI into widespread use. 1 region and in addition 3 areas and 9 districts are known to be behindhand for a variety of reasons, but expect to complete the exercise by September, 1976; a further 2 areas and 5 districts have made no arrangements as yet to introduce SI. Allied to the general resistance to the abolition of the mittimetre of mercury, an appreciable number of laboratories have not adopted the pascal for blood-gas measurements. A case is to be put to the E.E.C. in a few months’ time advocating the retention of the millimetre of mercury as a unit of measurement in clinical medicine. I fully appreciate the concern that has been demonstrated over the introduction of SI and hope that this information will go some way to assure those who are still uncommitted that elsewhere the National Health Service has virtually completed the changeover. Department of Health & Social Security, Alexander Fleming House, Elephant & Castle,
H. YELLOWLEES,
London SE1 6BY.
Chief Medical Officer
SiR,—More than
of the percentage of teaching have undergone willing or enforced change to SI, do we not need to know the effects? Does the change please or irritate clinicians, does it protect or endanger patients, facilitate or impair patient care? Does it clarify or confuse, help or hinder clinical diagnosis? Is there a consensus for an uncontrolled yet logical extension of molar reporting of, say, blood salicylate level to aspirin dosage? Do clinicians approve of SI in practice? Only a survey of opinion would reveal this and settle the mass-concentration molar and mm Hg/kilopascal controversies. Most clinicians would adopt a consensus view, but this has not been sought and is not expressed by the number of laboratories that have changed to SI. A. N. G. CLARK Brighton General Hospital, Elm Grove, Brighton BN2 3EW. JOANNA SHELDON and district
a statement
hospitals who
BASIC SCIENCES, SCIENCE, AND MEDICAL EDUCATION
SIR,-Professor Campbell’s article (Jan. 17, p. 134) is succinctly argued, lucid, and bold. He seems to be well aware that to talk about science always irritates medical audiences, and he must be prepared to accept the barrage that will inevitably follow. Surely he will be accused of "semantics" and Popperism Professor Campbell admits that science is not essential fOf health care. There the followers of Illich would agree, and their appeal for more common sense and humanitarianismis fair though one-sided. However, the introduction of more science in medical eduation would further undermine the pseudomilitary structure and discipline within the medical hierarchy held together t? power, authority, and tradition. Methods of scientific though: are powerful weapons against obscurantism and authef
357 tarianism. The student trained in scientific thought would be and would ask for hard data. Pharmaceutical industries would suffer financially because their claims would be rejected as profit biased. There are a few inherent snags and contradictions in teaching science in medical schools. First, somebody must teach it to teachers. Secondly, those who grasp complex problems of epistemology will become humble because they find that the definite answers they were looking for did not exist. Their newly acquired awareness of ignorance will be of no use in their role of doctor from whom patients will expect to have knowledge. Sir George Pickering’ summarised this beautifully: "From the time of Galen to our own ... medicine always presented a facade of systematic knowledge, or alleged knowledge, Infor, like religion, medicine could not tolerate ignorance tellectual nihilism is the very stuff of which scientists are made, but it is scarcely convenient for a practising physician." Thirdly, doctors are split by the dichotomy of wishing to be taken seriously as scientists and, at the same time, to retain their elitist power and pomp.23 While the scientist freely shares his knowledge and ignorance in discussion, the doctor keeps his patient in darkness by necessity or habit. While science can play with general hypotheses at leisure, the doctor must give an instant ad-hoc judgement or treatment for an individual patient. Another problem, clearly exposed by Professor Campbell, is essentialism in medical thought. Contemporary medicine continues to be based on an aristotelian classification of diseases. In the grey area of non-health, new nosological territories are being charted and named after their discoverers, as if they have existed since the creation. Beyond this area lies the Blessed Land of Normal, often referred to in the mythology of medicine but never mapped. The Classical Case is viewed as materialisation of the Idea or Essence of Disease, as if all diseases were disorders of kind and not of degree. One of the first critics of Professor Campbell, Dr West (Jan. 31, p. 242), says: "A clear description of a rash is just as much a part of science as a definition of a kilopascal ..." In fact, neither has anything to do with nature and processes of science with which Professor Campbell is concerned. The first is just a description (i.e., an observational statement) and the other is an arbitrary definition. It was Aristotle who believed that science consists of definitions and by defining everything we would know everything; but that was more than 2000 years
encouraged to question "experience"
...
"early" group were admitted within 9 h, and, though the sumthey were treated in this period, the details
mary states that given in the text
effects of
variance with this claim. At best, the cysteamine only be assessed in the six "early" and the results seem to confirm our earlier finding of are at
can
patients, almost complete protection against paracetamol hepatotoxicity.’ The Newcastle workers, however, seem reluctant to accept this, despite the fact that in these patients the maximum levels of aspartate aminotransferase (A.S.T.) and ferritin were significantly lower with less histological evidence of liver damage. In most of the comparisons made by Douglas et al. the data from the six "early" and the twelve "late" patients were pooled even though most of the late group were admitted and treated far too late for cysteamine to be effective. Their conclusions are not valid. The "paracetamol index" described by Douglas et al. is inappropriate and further confuses the issue. There is enormous individual variation in the plasma-paracetamol half-life after overdosage,6 and it is possible for the plasma concentration to be well above and well below the line in fig. 1 of their paper at different times in the same patient. The value obtained for the paracetamol index may thus depend more on the time that the patient is admitted after ingestion than on the severity of intoxication. Furthermore, it can be calculated that even if the plasma concentrations decline in parallel with the line in their fig. 1 the "paracetamol index" at any given time will itself decrease exponentially with a half-time of 3 h. We have treated nineteen severely poisoned patients with cysteamine given within 10 h (mean 6-8 h) of ingestion of paracetamol. All but one had a plasma-paracetamol well above the line in fig. 1 in the article by Douglas et al. None had significant hepatic or renal damage, and there were no deaths. In
PROTECTIVE EFFECT OF CYSTEAMINE GIVEN WITHIN
10 h OF
INGESTION
IN PATIENTS WITH SEVERE PARACETAMOL POISONING
ago. Mater Misericordiæ Dublin, Ireland.
Hospital,
PETR SKRABANEK
CYSTEAMINE FOR PARACETAMOL POISONING
S 1000 units/I), and three died in hepatic failure (see table). Furthermore, many of our treated patients were at risk because of alcoholism or previous consumption of drugs causing induction of hepatic microsomal enzymes 7 -an important aspect not considered by Douglas et al. Seven other patients who could not be given cysteamine for 11.5-26-5 h after overdosage all developed severe hepatic necrosis (mean A.S.T. > 4470 units/1). Our results with cysteamine given within 10 h to patients with severe paracetamol poisoning reinforce our previously expressed reservationsconcerning the ethics of controlled trials of cysteamine in this situation. We hope that the confusing report by Douglas et al. will not cause cysteamine to be withheld from patients at risk of severe liver damage or death after paracetamol overdosage. Regional Poisoning Treatment Centre, and University Department of Therapeutics, Royal Infirmary, Edinburgh EH3 9YW.
L. F. PRESCOTT
J. PARK A. T. PROUDFOOT
6. Prescott, L. F., Wright, N., Roscoe, P., Brown, S. S. ibid. 1971, i, 519. 7. Wright, N., Prescott, L. F. Scott. med. J. 1973, 18, 56. 8. Prescott, L. F., Matthew, H. Lancet, 1974, i, 998.
