-that the psychiatric symptoms are secondary to being labelled-is unlikely, given that low blood pressure in the United Kingdom is likely to be a cause for felicitation. Methodological limitations must be considered, but the collection of data by observers who were uninterested in this association at the time, and the strength of the associations reported, means that the findings cannot be easily dismissed. Neither sets of authors speculate on the cause of this association. What could they be? Could hypoperfusion impair cerebral monoamine activity? Could hypoperfusion lead to tissues, particularly muscle, being deprived of an adequate oxygen supply, causing the symptoms of fatigue?'0 Could the presence of depression or anxiety, from whatever cause, be associated with a persistent change in vasomotor tone? Evidence to support these speculations is lacking. Another option is to regard both as manifestations of interindividual variation, characterised by certain psychological responses (fatigue or depression) on the one hand and physiological state (persistent hypotension) on the other. (Another example of linked psychological responses and physiological states is the increased and delayed hypertensive response to certain stresses observed in people with type A behaviour.") Such a constitution was, in fact, recognised earlier this century,'2 and the neurasthenic patient has always been said to have lower than average blood pressure. Should anything be done now? Keeping a very open mind
is important. It could be that the Anglophone medical community has been working with a blinkered assumption that as high blood pressure is "bad" for the patient so low blood pressure must be "good." It may be that psychiatric morbidity modestly impairing the quality of life, associated with low blood pressure, has been overlooked. Secondly, it seems worth while to explore evidence that antihypotensive regimens help such morbidity. Clinical trials could be considered in collaboration with German or French colleagues. It is time for British doctors to put their Euroscepticism to the test. ANTHONY MANN Professor of Epidemiological Psychiatry and Vice Dean, Institute of Psychiatry, London SE5 8AF 1 Paver L. Medicine and culture: van'eties of treatmetnt in the United States, England, West Germany and France. New York: Henry Holt, 1988. 2 Pemberton J. Does constitutional hypotension exist? BMJ 1989;298:660-2. 3 Robbins JM, Korda H, Shapiro MF. Treatment for a nondisease: the case of low blood pressure. Soc Sci Med 1982;16:27-33. 4 Harrison TR. Principles of internal medicine. New York: McGraw Hill, 1987. 5 Tonkin A, Wing L. Hypotension: assessment and management. Medj Aust 1990;153:474-85. 6 Robinson S. Hypotension: the ideal normal blood pressure. N EnglJ7 Med 1940;233:407-16. 7 Mann A. Hypertension: psychological aspects and diagnostic impact in a clinical trial. Psychol Med
MonogrSuppl 1984;5:1-35. 8 Pilgrim J, Stansfeld S, Marmot M. Low blood pressure, low mood? BMJ 1992;304:75-8. 9 Wessely S, Nickson J, Cox B. Symptoms of low blood pressure: a population study. BMJ7
1990;301:362-5. 10 Simpson LO. Symptoms of low blood pressure. BM,7 1990;301:815. 11 Manuck SB, Craft S, Gold KJ. Coronary-prone behaviour pattern and cardiosascular disease. Psychophysiology 1978; 1 5:403-1 1. 12 Dally J. Low blood pressure: its causes and significance. London: Heinemann, 192
Guillain-Barre syndrome revisited Pathogenesis still unknown When Guillain, Barre, and Strohl made their original observations on two patients with progressive weakness, areflexia, and a high concentration of protein in their cerebrospinal fluid' there was heated debate over the relation of their syndrome with its good prognosis to the more serious and fatal case of ascending paralysis that Landry had described in detail some 60 years before.2 Subsequent authors have taken a broad view of the syndrome, which has come to be known as the Guillain-Barre syndrome, and have used the term to encompass several different clinical presentations and outcomes. Diagnostic criteria for the Guillain-Barre syndrome are wide and select a group of acute idiopathic neuropathies on the basis of progressive limb weakness and areflexia without evidence of porphyria or exposure to lead or toxins. Mild sensory signs, lack of fever, raised concentrations of protein in the cerebrospinal fluid, and neurophysiological evidence of demyelination are all helpful but not essential for the diagnosis. There remains no laboratory test to identify the disease reliably. A prospective study of the syndrome's outcome found that the median time to complete functional recovery was nine months, but one in five patients were left unable to work at 12 months and a further one in eight had died.3 Poor outcome is associated with more rapidly progressive disease, the need for assisted ventilation, the presence of small distally evoked muscle action potentials, and older age. Antecedent infections with numerous agents have been reported, but convincing controlled data are available only for campylobacter, cytomegalovirus, and Epstein-Barr virus.' The nature of the antecedent infection has little effect on the clinical features or outcome with the exception of campylobacter, which correlates with poor outcome, perhaps reflecting a different pathogenesis. About 3% of patients with the BMJ
VOLUME 304
11 JANUARY 1992
acute disease develop a chronic or relapsing course. These patients differ in having a higher prevalence of certain HLAs (CW7 and B8) and responding to immunosuppression.5 Necropsy studies suggest that different pathological processes recur within the syndrome, contributing to the variation in clinical features and outcome. A perivascular inflammatory infiltrate was prominent in 12 cases in an influential study of the syndrome in 19696 and came to be regarded as the pathological hallmark of the disease. Later reports have suggested that lymphocytic infiltration is prominent in only some cases and many show demyelination without obvious lymphocytes.7 A few very severe cases that fulfil diagnostic criteria for Guillain-Barre syndrome have pathological evidence of an axonal neuropathy.8 Such patients have peripheral nerves that are difficult to stimulate electrically and recover poorly. How infection triggers the disease is unknown. The generation of autoimmune responses because of similarities in structure between infective agents and host proteins is the most favoured hypothesis, but hard evidence is lacking. It is tempting to speculate that the pathogenesis of cases of Guillain-Barre syndrome with appreciable lymphocytic infiltration is mediated by T cells while those with few lymphocytes present are mediated by antibodies. Efforts to show cellular immune hypersensitivity to neural antigens in patients with Guillain-Barre syndrome have been disappointing despite evidence supporting their role in experimental allergic neuritis in animal models. Lymphocytes from the blood of a few patients respond to the P2 protein of myelin.9 Complement fixing antibodies to neural antigens have been found occasionally by many authors but have been described in more than three quarters 65
of patients with the syndrome when a sensitive but technically demanding technique was used.'0 This assay detected IgM antibodies that seemed to react with an antigen in myelin, called the Forssman antigen, and fell in titre in the first few weeks of the neuropathy. A separate IgG antibody which reacts with gangliosides GM, and GDb occurs in about 15% of patients." These antibodies seem more frequent in patients with severe disease and are associated with previous infection with campylobacter. Such antibodies could be secondary to whatever process is causing the demyelination, and international collaborative efforts are underway to clarify their role. Research into the role of specific treatment for the syndrome has been more rewarding. Trials of plasma exchange have shown convincing benefits in the more severe cases of the syndrome, particularly in those patients who become bedbound and receive plasma exchange within one to two weeks of the onset of their disease.'2 3 Patients with milder disease who receive plasma exchange later in the course of their illness may benefit, but the relevant studies have not been performed. Plasma exchange is expensive and, although morbidity is small in experienced hands, other treatment options are needed. Steroids were a popular treatment until an influential but small controlled trial in 1978 failed to provide evidence of benefit."' A trial conducted by the Medical Research Council is in progress to assess a possible role for methylprednisolone, but preliminary results suggest that it is ineffective. '5 Gammaglobulin is the latest treatment option, and interim results of a Dutch trial were encouraging (FCA Van der Meche et al, Peripheral Neuropathy Association meeting, Oxford 1990). Further studies will be needed to define its role. Such treatment is crude, and a better understanding of the pathogenesis of the disease may eventually lead to more specific modes of treatment. Until then there is no doubt that excellent intensive care is vital to reduce morbidity and mortality. Autonomic complications are difficult to manage
and may rarely lead to sudden death. It is worrying that mortality in a British survey could be as high as 13%3 whereas centres specialising in neurological intensive care quote a rate of I 3%. 6 Such figures are clearly influenced by many factors, including selection bias, but they are understandably used by those who favour increased specialisation in the management of this fascinating and challenging condition. JOHN WINER
Senior Lecturer, University Department and Regional Centre for Neurology, University of Birmingham, Queen Elizabeth Hospital, Birmingham B15 2TH I Guillain G, Barre JA, Strohl A. Sur un syndrome de radiculonevrite avec hyperalbuminose du liquide cephalorachidien sans reaction cellulaire. Remarques sur les caracteres cliniques et graphiques des reflexes tendineux. Bulletin de la Societet Medicale des Hopitaux Paris 1916;40: 1462-70. 2 Landrv 0. Note sur la paralvsie ascendante aigue. Gazette Hebdomadaire Medicine Chirurgie 1859;6:472-88. 3 Winer JB, Hughes RAC, Osmond C. A prospective study of acute idiopathic neuropathy. 1. Clinical features and their prognostic value. J Neurol Neurosurg Psychiatry 1988;51:605-12. 4 Winer JB, Hughes RAC, Anderson MIJ, Jones DM, Kangro H, Watkins RPF. A prospective study of acute idiopathic neuropathy. 2. Antecedent events. J Neurol Neurosurg Psychiatry 1988;51:6 13-8. S Hughes RAC. Chronic idiopathic demyelinating polvradiculoneuropathy. In: Guillain-Barre syndrome. London: Springer-Verlag, 1990:205-46. 6 Asbury AK, Arnason BG, Adams RD. The inflammatory lesion in idiopathic polyneuritis. Its role in pathogenesis. Medicine 1969;48:173-215. 7 Honosvar M, Tharakan JKJ, Hughes RAC, Leibowitz S, Winer JB. A clinicopathological studv of Guillain-Barre syndrome: nine cases and literature review. Brain 1991;114:1245-70. 8 Feasby TE, Gilbert JJ, Brown WF, Bolton CF, Hahn AF, Koopman WJ, et al. An acute axonal form of Guillain-Barre polyneuropathy. Brain 1986;109:1115-26. 9 Taylor WA, Brostoff SW, Hughes RAC. P. specific lymphocyte transformation in Guillain-Barre syndrome and chronic idiopathic demyelinating polvradiculoneuropathy. J Neurol Sci
1991;104:52-5. 10 Koski CL, Humphrey R, Shin ML. Anti-peripheral myelin antibodies in patients with demyelinating neuropathy: quantitative and kinetic determination of serum antibodies by complement component and fixation. Proc Natl Acad Sci USA 1985;82:905-9. 11 Walsh FS, Cronin M, Koblar S, Doherty P, Winer J, Leon A, et al. Association between glyconjugate antibodies and campylobacter infection in patients with Guillain-Barre syndrome.
J Neurosimmunol 1991;34:43-51. 12 Guillain-Barre Syndrome Study Group. Plasmapheresis for acute Guillain-Barre syndrome.
Neurology 1985;35:1096-104. 13 French Cooperative Group in Plasma Exchange in Guillain-Barre Syndrome. Efficiency of plasma exchange in Guillain-Barre syndrome: role of replacement fluids. Ann Neurol 1987;22:753-61. 14 Hughes RAC, Davis JMN, Perkin GD, Pierce JM. Controlled trial of prednisolone in acute polyneuropathy. Lancet 1978;ii:750-3. 15 Hughes RAC. Ineffectiveness of high dose intrasenous methvlprednisolone in Guillain-Barre syndrome. Lancet 1991;338:1142. 16 Ropper AH, Shahani BT. Diagnosis and management of acute areflexic paralysis with emphasis on Guillain-Barre syndrome. In: Asbury AK, Gilliatt RW, eds. Peripheral nerve disorders. A practical approach. London: Butterworth, 1984.
Preventing infection in laboratories A new code promotes even safer practices Before 1974 only limited parts of health care premises were subject to health and safety inspection. When the Health and Safety at Work Act 1974 extended occupational and safety legislation to all areas of health care the need for guidance covering clinical laboratories became clear. The original code of practice on preventing laboratory acquired infections, the Howie code,' was issued in 1978, but since then it has been modified, added to, and partially superseded. Now there is a completely new code of practice.25 The original Howie code was based on a report of a working party set up by the Department of Health and Social Security to consider ways of preventing laboratory acquired infection. The report was never published-only the code, which immediately prompted some controversy. Although it had been widely circulated beforehand for consultation, its publication was followed by considerable discussion about the real and apparent cost implications of the code.67 It was always clear that it would require reviewing and updating, and this has now been done by technical subcommittees of the Health and Safety Commission's health services advisory group. 66
The new code does not have the force of law, but, as with its predecessor, health and safety inspectors (and courts and tribunals) are likely to use it as the benchmark for what is reasonable and acceptable in the laboratory. Nevertheless, one ofthe original criticisms of the Howie code was that it was sometimes overinterpreted by overzealous safety inspectors, and overinterpretation remains a danger. Strict application of any code may so restrict activity that it defeats the object of the exercise. Too tight restrictions on laboratories (or other health care departments) may prevent them from fulfilling their main task of contributing to health care. In essence, the prevention of infection in clinical laboratories requires the application of common sense. It does not require large amounts of money, but it does mean that all staff coming into contact with biological specimens should be trained about the dangers of inappropriate activity both to themselves and to others. Furthermore, planners will have to include adequate provision for safety in new laboratories, and these measures must not be adversely affected by considerations of economy. Collins summarised the problem as follows: "Diseases cost money, but good techniques cost no more than BMJ VOLUME 304
11 JANUARY 1992
is important. It could be that the Anglophone medical community has been working with a blinkered assumption that as high blood pressure is "bad" for the patient so low blood pressure must be "good." It may be that psychiatric morbidity modestly impairing the quality of life, associated with low blood pressure, has been overlooked. Secondly, it seems worth while to explore evidence that antihypotensive regimens help such morbidity. Clinical trials could be considered in collaboration with German or French colleagues. It is time for British doctors to put their Euroscepticism to the test. ANTHONY MANN Professor of Epidemiological Psychiatry and Vice Dean, Institute of Psychiatry, London SE5 8AF 1 Paver L. Medicine and culture: van'eties of treatmetnt in the United States, England, West Germany and France. New York: Henry Holt, 1988. 2 Pemberton J. Does constitutional hypotension exist? BMJ 1989;298:660-2. 3 Robbins JM, Korda H, Shapiro MF. Treatment for a nondisease: the case of low blood pressure. Soc Sci Med 1982;16:27-33. 4 Harrison TR. Principles of internal medicine. New York: McGraw Hill, 1987. 5 Tonkin A, Wing L. Hypotension: assessment and management. Medj Aust 1990;153:474-85. 6 Robinson S. Hypotension: the ideal normal blood pressure. N EnglJ7 Med 1940;233:407-16. 7 Mann A. Hypertension: psychological aspects and diagnostic impact in a clinical trial. Psychol Med
MonogrSuppl 1984;5:1-35. 8 Pilgrim J, Stansfeld S, Marmot M. Low blood pressure, low mood? BMJ 1992;304:75-8. 9 Wessely S, Nickson J, Cox B. Symptoms of low blood pressure: a population study. BMJ7
1990;301:362-5. 10 Simpson LO. Symptoms of low blood pressure. BM,7 1990;301:815. 11 Manuck SB, Craft S, Gold KJ. Coronary-prone behaviour pattern and cardiosascular disease. Psychophysiology 1978; 1 5:403-1 1. 12 Dally J. Low blood pressure: its causes and significance. London: Heinemann, 192
Guillain-Barre syndrome revisited Pathogenesis still unknown When Guillain, Barre, and Strohl made their original observations on two patients with progressive weakness, areflexia, and a high concentration of protein in their cerebrospinal fluid' there was heated debate over the relation of their syndrome with its good prognosis to the more serious and fatal case of ascending paralysis that Landry had described in detail some 60 years before.2 Subsequent authors have taken a broad view of the syndrome, which has come to be known as the Guillain-Barre syndrome, and have used the term to encompass several different clinical presentations and outcomes. Diagnostic criteria for the Guillain-Barre syndrome are wide and select a group of acute idiopathic neuropathies on the basis of progressive limb weakness and areflexia without evidence of porphyria or exposure to lead or toxins. Mild sensory signs, lack of fever, raised concentrations of protein in the cerebrospinal fluid, and neurophysiological evidence of demyelination are all helpful but not essential for the diagnosis. There remains no laboratory test to identify the disease reliably. A prospective study of the syndrome's outcome found that the median time to complete functional recovery was nine months, but one in five patients were left unable to work at 12 months and a further one in eight had died.3 Poor outcome is associated with more rapidly progressive disease, the need for assisted ventilation, the presence of small distally evoked muscle action potentials, and older age. Antecedent infections with numerous agents have been reported, but convincing controlled data are available only for campylobacter, cytomegalovirus, and Epstein-Barr virus.' The nature of the antecedent infection has little effect on the clinical features or outcome with the exception of campylobacter, which correlates with poor outcome, perhaps reflecting a different pathogenesis. About 3% of patients with the BMJ
VOLUME 304
11 JANUARY 1992
acute disease develop a chronic or relapsing course. These patients differ in having a higher prevalence of certain HLAs (CW7 and B8) and responding to immunosuppression.5 Necropsy studies suggest that different pathological processes recur within the syndrome, contributing to the variation in clinical features and outcome. A perivascular inflammatory infiltrate was prominent in 12 cases in an influential study of the syndrome in 19696 and came to be regarded as the pathological hallmark of the disease. Later reports have suggested that lymphocytic infiltration is prominent in only some cases and many show demyelination without obvious lymphocytes.7 A few very severe cases that fulfil diagnostic criteria for Guillain-Barre syndrome have pathological evidence of an axonal neuropathy.8 Such patients have peripheral nerves that are difficult to stimulate electrically and recover poorly. How infection triggers the disease is unknown. The generation of autoimmune responses because of similarities in structure between infective agents and host proteins is the most favoured hypothesis, but hard evidence is lacking. It is tempting to speculate that the pathogenesis of cases of Guillain-Barre syndrome with appreciable lymphocytic infiltration is mediated by T cells while those with few lymphocytes present are mediated by antibodies. Efforts to show cellular immune hypersensitivity to neural antigens in patients with Guillain-Barre syndrome have been disappointing despite evidence supporting their role in experimental allergic neuritis in animal models. Lymphocytes from the blood of a few patients respond to the P2 protein of myelin.9 Complement fixing antibodies to neural antigens have been found occasionally by many authors but have been described in more than three quarters 65
of patients with the syndrome when a sensitive but technically demanding technique was used.'0 This assay detected IgM antibodies that seemed to react with an antigen in myelin, called the Forssman antigen, and fell in titre in the first few weeks of the neuropathy. A separate IgG antibody which reacts with gangliosides GM, and GDb occurs in about 15% of patients." These antibodies seem more frequent in patients with severe disease and are associated with previous infection with campylobacter. Such antibodies could be secondary to whatever process is causing the demyelination, and international collaborative efforts are underway to clarify their role. Research into the role of specific treatment for the syndrome has been more rewarding. Trials of plasma exchange have shown convincing benefits in the more severe cases of the syndrome, particularly in those patients who become bedbound and receive plasma exchange within one to two weeks of the onset of their disease.'2 3 Patients with milder disease who receive plasma exchange later in the course of their illness may benefit, but the relevant studies have not been performed. Plasma exchange is expensive and, although morbidity is small in experienced hands, other treatment options are needed. Steroids were a popular treatment until an influential but small controlled trial in 1978 failed to provide evidence of benefit."' A trial conducted by the Medical Research Council is in progress to assess a possible role for methylprednisolone, but preliminary results suggest that it is ineffective. '5 Gammaglobulin is the latest treatment option, and interim results of a Dutch trial were encouraging (FCA Van der Meche et al, Peripheral Neuropathy Association meeting, Oxford 1990). Further studies will be needed to define its role. Such treatment is crude, and a better understanding of the pathogenesis of the disease may eventually lead to more specific modes of treatment. Until then there is no doubt that excellent intensive care is vital to reduce morbidity and mortality. Autonomic complications are difficult to manage
and may rarely lead to sudden death. It is worrying that mortality in a British survey could be as high as 13%3 whereas centres specialising in neurological intensive care quote a rate of I 3%. 6 Such figures are clearly influenced by many factors, including selection bias, but they are understandably used by those who favour increased specialisation in the management of this fascinating and challenging condition. JOHN WINER
Senior Lecturer, University Department and Regional Centre for Neurology, University of Birmingham, Queen Elizabeth Hospital, Birmingham B15 2TH I Guillain G, Barre JA, Strohl A. Sur un syndrome de radiculonevrite avec hyperalbuminose du liquide cephalorachidien sans reaction cellulaire. Remarques sur les caracteres cliniques et graphiques des reflexes tendineux. Bulletin de la Societet Medicale des Hopitaux Paris 1916;40: 1462-70. 2 Landrv 0. Note sur la paralvsie ascendante aigue. Gazette Hebdomadaire Medicine Chirurgie 1859;6:472-88. 3 Winer JB, Hughes RAC, Osmond C. A prospective study of acute idiopathic neuropathy. 1. Clinical features and their prognostic value. J Neurol Neurosurg Psychiatry 1988;51:605-12. 4 Winer JB, Hughes RAC, Anderson MIJ, Jones DM, Kangro H, Watkins RPF. A prospective study of acute idiopathic neuropathy. 2. Antecedent events. J Neurol Neurosurg Psychiatry 1988;51:6 13-8. S Hughes RAC. Chronic idiopathic demyelinating polvradiculoneuropathy. In: Guillain-Barre syndrome. London: Springer-Verlag, 1990:205-46. 6 Asbury AK, Arnason BG, Adams RD. The inflammatory lesion in idiopathic polyneuritis. Its role in pathogenesis. Medicine 1969;48:173-215. 7 Honosvar M, Tharakan JKJ, Hughes RAC, Leibowitz S, Winer JB. A clinicopathological studv of Guillain-Barre syndrome: nine cases and literature review. Brain 1991;114:1245-70. 8 Feasby TE, Gilbert JJ, Brown WF, Bolton CF, Hahn AF, Koopman WJ, et al. An acute axonal form of Guillain-Barre polyneuropathy. Brain 1986;109:1115-26. 9 Taylor WA, Brostoff SW, Hughes RAC. P. specific lymphocyte transformation in Guillain-Barre syndrome and chronic idiopathic demyelinating polvradiculoneuropathy. J Neurol Sci
1991;104:52-5. 10 Koski CL, Humphrey R, Shin ML. Anti-peripheral myelin antibodies in patients with demyelinating neuropathy: quantitative and kinetic determination of serum antibodies by complement component and fixation. Proc Natl Acad Sci USA 1985;82:905-9. 11 Walsh FS, Cronin M, Koblar S, Doherty P, Winer J, Leon A, et al. Association between glyconjugate antibodies and campylobacter infection in patients with Guillain-Barre syndrome.
J Neurosimmunol 1991;34:43-51. 12 Guillain-Barre Syndrome Study Group. Plasmapheresis for acute Guillain-Barre syndrome.
Neurology 1985;35:1096-104. 13 French Cooperative Group in Plasma Exchange in Guillain-Barre Syndrome. Efficiency of plasma exchange in Guillain-Barre syndrome: role of replacement fluids. Ann Neurol 1987;22:753-61. 14 Hughes RAC, Davis JMN, Perkin GD, Pierce JM. Controlled trial of prednisolone in acute polyneuropathy. Lancet 1978;ii:750-3. 15 Hughes RAC. Ineffectiveness of high dose intrasenous methvlprednisolone in Guillain-Barre syndrome. Lancet 1991;338:1142. 16 Ropper AH, Shahani BT. Diagnosis and management of acute areflexic paralysis with emphasis on Guillain-Barre syndrome. In: Asbury AK, Gilliatt RW, eds. Peripheral nerve disorders. A practical approach. London: Butterworth, 1984.
Preventing infection in laboratories A new code promotes even safer practices Before 1974 only limited parts of health care premises were subject to health and safety inspection. When the Health and Safety at Work Act 1974 extended occupational and safety legislation to all areas of health care the need for guidance covering clinical laboratories became clear. The original code of practice on preventing laboratory acquired infections, the Howie code,' was issued in 1978, but since then it has been modified, added to, and partially superseded. Now there is a completely new code of practice.25 The original Howie code was based on a report of a working party set up by the Department of Health and Social Security to consider ways of preventing laboratory acquired infection. The report was never published-only the code, which immediately prompted some controversy. Although it had been widely circulated beforehand for consultation, its publication was followed by considerable discussion about the real and apparent cost implications of the code.67 It was always clear that it would require reviewing and updating, and this has now been done by technical subcommittees of the Health and Safety Commission's health services advisory group. 66
The new code does not have the force of law, but, as with its predecessor, health and safety inspectors (and courts and tribunals) are likely to use it as the benchmark for what is reasonable and acceptable in the laboratory. Nevertheless, one ofthe original criticisms of the Howie code was that it was sometimes overinterpreted by overzealous safety inspectors, and overinterpretation remains a danger. Strict application of any code may so restrict activity that it defeats the object of the exercise. Too tight restrictions on laboratories (or other health care departments) may prevent them from fulfilling their main task of contributing to health care. In essence, the prevention of infection in clinical laboratories requires the application of common sense. It does not require large amounts of money, but it does mean that all staff coming into contact with biological specimens should be trained about the dangers of inappropriate activity both to themselves and to others. Furthermore, planners will have to include adequate provision for safety in new laboratories, and these measures must not be adversely affected by considerations of economy. Collins summarised the problem as follows: "Diseases cost money, but good techniques cost no more than BMJ VOLUME 304
11 JANUARY 1992
