agency. The sex ratio of referrals was two men to one woman; the sex ratio of the study population was 1 3 men to one woman. Referrals were made to general practitioners (37); dentists ( 15) for atlantoaxial x ray examination (10-ILeicester hosted the Special Olympics last year, and many Down's syndrome patients wanted to participate); ophthalmologv (nine); physiotherapy, speech therapy, and occupational therapy (six); geneticists, chiropodists, dietitians, and orthopaedic surgeons (six). In those referred to general practitioners 33 medical problems were diagnosed: raised blood pressure (four), glycosuria (three), proteinuria (two), hypothyroidism (three), hyperthyroidism (one), skin lesions (eight), anxiety (one), depression (one), obesity (two), weight loss (one), heart failure (one), heart defect (one), wax in ears leading to hearing loss (three). Seven patients required surgerv: three for hernia, two for undescended testicles, one for subluxation of the hip, and one for a septic bunion. Four patients were advised on
contraception. No patient complained or had svmptoms. Rarely the family and sometimes staff reported a change in behaviour. The need for routine medical examinations seems self evident. It is effective to use clinical medical officers, who are accustomed to assessing potential in those without speech, flexible in their approach to different venues, and accustomed to liaison within a health care team. For the patients, being examined in a familiar setting with the reassuring presence of their community nurse is non-threatening. For carers, the long appointment encourages useful dialogue with the clinical medical officer and communitv nurse, which, it is hoped, reduces the burden on them-and on the state. MAUREEN BEAUCHAMP
leiceseirshire Area Health Authorit%, I.cester LEI 6TY I Ashley-Miller M. Community care. Br Medj 1990;300:487-8.
V24 February.i
2 Bicknell J. (Consent and people with mcntal handicap. Br Medj 1989;299:1176-7. I I November.;
Research ethics committees SIR,-I sympathise with Dr Ann Cartwright's view that researchers using samples taken from public records should not be required to submit their protocols to district health authority ethics committees.' I disagree, however, with her assertion that when considering sociological studies ethics committees should confine themselves to questions of confidentiality of data and not question the design of the study. Most of the protocols that ethics committees see concern research on NHS patients. As a patient representative I want the committees to examine the quality of all the research they are asked to approve, but as a former lay member of a research ethics committee I know that they are woefully inadequate to do this. They are not required to have a social scientist as a member-an important omission in the Department of Health's draft guidelines. Their ignorance of another discipline does not prevent medical members, who are in the majority, from thinking that they know how such research should, or should not, be done. I have seen studies either go through on the nod because it was wrongly assumed that there were no ethical issues involved or be criticised in a way that could damage carefully structured questionnaires. There is an increasing volume of research in nursing and midwifery exploring issues that are of great value to patients, yet researchers are deprived of informed and educational advice by the ethics committees that look at their projects. I have no doubt that there are issues of major ethical importance in approval or amendment of
BMJ VOLUME 300
7 APRIL 1990
these studies, and it mav be that they should go to separate committees that are not primarily medical. One problem at present is that proposals from a medical source are more likely to be approved - yet some of such "sociological" research published in medical journals is of such poor quality that no social science based ethics committee would have passed it and no decent sociological journal would have published it. Doctors seem unaware that questions asked before and after patients have left hospital are likely to get different answers and that being interviewed by or receiving a questionnaire from a doctor, nurse, or medical student may affect the patient's reply. Patients who know or suspect that their doctors will see their replies are unlikely to be critical -but the advantage to the doctors is that the "survey" shows how much patients like their care. Many doctors and research ethics committees seem unaware that the abilitv to design and analyse questionnaires does not come automatically with a medical degree. Future management of patients mav be based on the results of such "studies"; poor quality protocols are dangerous to patients and we have a right to be protected from them. The only question is whether the present ethics committees can be trained to do the job or whether we need different committees for social science research.
seem unequivocal about the need for separate submissions to local committees.' On the other hand, the recently published guidelines from the Royal College of Physicians on the practice of ethics committees acknowledges that multiple applications may be impracticable "so that research that is valuable to the community can be impeded or may even not be undertaken."4 The confusing nature of the present situation is thus all too evident. The articles, commentaries, and letters published over the past few months have revealed a range of serious problems. So far, these have not been debated sufficiently thoroughly around the same table by those concerned. T W MEADE
MRC Epidemiology and Medical Care Unit, Northwick Plark Hospital, Harrow, Middlesex HAl 3UJ I Drury Mi. Research ethics committees. Br Med7 1990;300:608. (3 March.) 2 Meade IW. Research ethics committees. Br Alcd J 1990;300: 396. (10 February., 3 Marshall T, Moodie P. Research ethics committees revisited. BrMed] 1989;299:1419-20. (9 December.) 4 Royal College of Physicians. Guidelintes on the practice o ethift/s commtttees tn medical research involving human subjects. 2nd cd. London: Royal College of Physicians, 1 99.
JEAN ROBINSON
()xford (X2 7ti' I Cartwright A. Research ethics committces. Br lcdj 1990;300: 607. (3 March.)
SIR,-The establishment of a clinical trials ethics committee by the Royal College of General Practitioners referred to by Sir Michael Drury in his letter' is, I think, another indication that some kind of central committee is needed, though for several reasons I am not confident that it solves the problems I was referring to.Firstly, many multicentre studies are, of course, carried out through hospitals. Rather than having to rely on yet another "central" committee on these occasions-set up, for example, by the Royal College of Physicians or the Royal College of Surgeons-it would be much better to have a single committee responsible for all multicentre studies, however they were conducted. Secondly, our own experience of local reactions to an arrangement similar to the committee described by Sir Michael was not encouraging. In 1988, at the suggestion of the Department of Health, we planned a feasibility study of anonymous HIV testing. To allow an early start and also because of the special considerations involved we approached the convener of the committee on ethical issues in medicine of the Royal College of Physicians (and I am grateful to Sir Douglas Black for agreeing that I may refer to our exchange). He consulted the then president and another member of the committee and approved our proposal. When the general practitioners in our study consulted their nine local ethical committees, however, these committees nearly all asked for separate submissions even though they knew the work had been approved centrally. If the decisions of any future national committee were not binding on local committees, the undoubted problems of multiple applications to separate committees would therefore remain. Eventually, the local committees also accepted the HIV study-with one exception, which illustrates the potential for idiosyncratic decisions. That particular committee declined the proposal on the grounds that anonymous testing would leave subjects positive for HIV unidentified and uncounselled, even though the national scheme for anonymous testing had been announced and was about to start. Finally, the Department of Health guidelines that you referred to in the BMJ's news section
Bovine spongiform encephalopathy SIR,-In his editorial Professor W B Matthews discusses bovine spongiform encephalopathy.' The publicity given to this disease and its possible public health implications has prompted a reexamination of the relation between it and similar diseases found in humans and other animals-the transmissible dementias (box).'
The transmissible dementias Disease
Animal affected
Scrapic
Sheep, goats Mink M\ule deer, elk Cattle Humanis
Transmissible mink encephalopathy Chronic wasting disease Bovine spongiform encephalopathy Sporadic Creutzfeldt-Jakob disease Familial Creutzfeldt-Jakob disease
Gerstmann-Straussler svndrome Kuru
Humans Humans Humans
All of the transmissible dementias share a common factor-the involvement of an aberrant isoform of prion protein.' This protein is the product of a highly conserved gene found in organisms as diverse as the fruit fly and humans.' The abnormal isoform of prion protein is present in brain extracts from patients with transmissible dementias and is the main constituent of the amyloid plaques in these patients.3-6 Its presence distinguishes transmissible dementias from conditions such as Alzheimer's disease and can be regarded as diagnostic of transmissible dementias.4i Distinct prion gene alleles are associated with long and short disease incubation times in mice, and linkage between the GerstmannStraussler syndrome and both a base substitution' and an insertion in the prion gene"" has been shown. The evidence suggests that in all animals the aberrant isoform of prion protein has a central and possible causal role in the genesis and pathology of transmissible dementias. Over the past two years we have screened patients with a wide range of familial dementias and ataxias and those with apparently sporadic symptoms to determine whether these diseases are in fact part of the prion disease range. Using both immunocytochemical and molecular biological techniques we have shown that several families labelled as having atypical dementias were suffering from variants of prion disease. "'
943
The recent work on the pathology and genetics of prion disease in humans has several implications for the epidemiological proposals at the end of Professor Matthews's editorial. Firstly, because of the clinical and pathological variability the true incidence of transmissible dementias (prion disease) in humans can be established with certainty only by a combination of genetic screening and immunocytochemical detection of the abnormal isoform of prion protein. Secondly, on the basis of information already available the incidence of the range of prion diseases in humans is (and probably always was) significantly higher than the figures quoted in the editorial. Thirdly, because of this improved rate of detection any attempts at ascertaining the possible impact of foodstuffs contaminated with bovine spongiform encephalitis on the incidence of transmissible dementias must take this apparent rise in the number of cases into account. Finally, the number of familial cases being identified raises the possibility that in humans the prion disease range has more of a genetic basis than hitherto realised. G W ROBERTS
St Mary's Hospital Medical School, London W2 lPG J COLLINGE
Clinical Research Centre, Harrow, Middlesex 1 Matthews WB. Bovine spongiform encephalopathv. Br Med7 1990;300:412-3. (17 February.) 2 Beck E, Daniel PM. Neuropathology of transmissible spongiform encephalopathies. In: Prusiner SB, McKinley MP, eds. Prtons: novel infectious pathogens causing scrapie and Creutzfeldt-
Jakob disease. San Diego: Academic Press, 1987:331-3. 3 Westaway D, Carlson GA, Prusiner SB. Unravelling prion diseases through molecular genetics. Trends in Neurosciences 1989;12:22 1-7. 4 Roberts GW, Lofthouse R, Brown R, Crow TJ, Barry RA, Prusiner SB. Prion protein immunoreactisvity in human transmissible dementia. N Eng1i Ied 1986;315:1231-3. 5 Roberts GW, Lofthouse R, Allsop D, et al. CNS amvloid proteins in neurodegenerative diseases. Neurolog 1988;38: 1534-40. 6 Hope J, Reekie LJD, Hunter N, et al. Fibrils from brains of cows with new cattle disease contain scrapie-associated protein. Nature 1988;336:390-2. 7 Westaway D, Goodman PA, Mirenda CA, McKinley MP, Carlson GA, Prusiner SB. Distinct prion proteins in short and long scrapie incubation period mice. Cell 1987;51:651-2. 8 Hsiao K, Baker HF, Crow TJ, et al. Linkage of a prion protein missense variant to Gerstmann-Straussler syndrome. Nature 1989;338:342-5. 9 Owen F, Poulter M, Lofthouse R, et al. Insertion in prion protein gene in familial Creutzfeldt-Jakob disease. Lancet
1989;i:51-2. 10 Collinge J, Harding AE, Owen F, et al. Diagnosis of GerstmannStraussler syndrome in familial dementia with prion protein gene analysis. Lancet 1989;ii: 15-7.
SIR,-Drs Christopher Fear and Manikkarasa Devakumar' take issue with Professor W B Matthews over his statement that all patients with Creutzfeldt-Jakob disease must be seen by a neurologist who is familiar with the disease-: instead they would substitute electroencephalography as a baseline investigation. Those following this recommendation are more likely to miss the diagnosis than a physician or psychiatrist prepared to review the diagnosis at each stage in the light of clinical evidence. In practice the electroencephalogram is rarely helpful early in the disease, and the best a neurophysiologist reading it can do is to suggest that the test be repeated at intervals. Brown et al observed that the characteristic periodic electroencephalogram complexes were found comparatively late in the illness.' Although a pathognomic electroencephalogram is eventually seen in about 75% to 80% of patients, this may not be the case in certain groups of patients with Creutzfeldt-Jakob disease -for example, those receiving human growth hormone therapy.4 E M R CRITCHLEY
Department of Neurology, Royal Preston Hospital, Preston PR2 4HT
944
I Fear C, Devakumar M. Bovine spongiform cincephalopathy. BrAled_ 1990;300:817. (24 March. 2 Matthews >'B. Bovine spx)ngiform encephalopathy. lBr Med 7 1990;300:912-3. 17 Februarv. 3 Brown P, Cathala F, Castaigne 1', (iajiusck D)C. CretitzlclditJakob disease: clinical analysis of a consecutive series ot 230 neuropathically verified cases. Ann Neurol 1986;20:597-602. 4 Brown P. Thc dccline anid fall of CreLutzfeldt-Jakob disease associated with htiman growth hormone thcrapy. Neurology
1988;38:1135-7.
Non-invasive mechanical ventilation for acute respiratory failure
arterial carbon dioxide tension with this technique. In summary, we believe that both nasal intermittent positive pressure ventilation and continuous positive airways pressure may be useful adjunct treatments in patients with hvpercapnic respiratorv failure, particularly those who are not suitable for assisted ventilation through an endotracheal tube. In our hands this treatment still requires close supervision by medical, phvsiotherapy, and nursing staff, and at present we prefer to treat these patients in high dependency settings. D BELL M W Me NICOI.
J F RIORDAN C PRATT S LAM
Central Mliddlesex Hospital, London NWO 7NS
SIR, - In their paper Dr M W Elliott and colleagues describe the use of nasal intermittent positive pressure ventilation in patients with acute hypercapnic respiratory failure.' Of their group, three patients had chronic obstructive pulmonary disease. We have recently managed four patients with this condition with associated acute hypercapnic respiratory failure. Two patients were treated using nasal intermittent positive pressure ventilation, and in the other two patients we used nasal continuous positive airways pressure. The table gives the patients' details. All four patients had known severe chronic obstructive pulmonary disease and were admitted with severe hypercapnic respiratory failure; two of them had associated respiratory acidosis. None of the patients were considered suitable for endotracheal intubation and assisted ventilation because of end stage chronic disease. All four patients had supplemental oxygen at a flow of between 0-5 1/min and 2-0 1/min entrained through a port in the nasal mask. Arterial oxygen tensions increased in all patients from a range of 3 9-4-7 kPa to 6-3-8 0 kPa with nasal intermittent positive pressure ventilation or continuous positive airways pressure. In three patients there was a fall in arterial carbon dioxide tension; in the other patient, who was treated by nasal intermittent positive pressure ventilation, the arterial carbon dioxide tension was maintained (table). Nasal continuous positive airways pressure was well tolerated, but both patients in whom ventilation was assisted by nasal intermittent positive pressure ventilation required chin straps to prevent excessive air loss during the inspiratory phase. Two patients survived to be discharged from hospital: one was alive and well several months after discharge, and one died of a bronchogenic carcinoma four months after discharge. Of the two who died in hospital, one died of progressive renal failure despite correction of arterial gas imbalance and the other died of progressive respiratory failure three weeks after admission despite an initial response. In agreement with Dr Elliott and colleagues we have found nasal intermittent positive pressure ventilation to be of benefit in two patients with severe hypercapnic respiratory failure, but in addition two other patients benefited from the use of nasal continuous positive airways pressure, which has been used previously to manage patients with hypoxia but without hvpercapnial and in the treatment of obstructive sleep apnoea.' In the two patients described we achieved improvement in arterial oxygen tension with an associated fall in
I Elliott MW, Steven MH, I'hillips GD, Braithwaitc MA. Non-invasive mechanical ventilation for acute respirators failure. Br.MedJ 1990;300:358-60. (10 February.} 2 Kesten S, Rebuck AS. Nasal continuous positive airway pressturc in Poneumocystis carinii pneumonia. Lancet 1988;ii: 1414-5. 3 Sullivan CE, Berthon-Joncs M, Issa FG, Eves L. Reversal ot obstructive sleep apnoca by continuous positive airwav pressure applied through the nares. Lancet 1981 :862-5.
Vasectomy and testicular cancer SIR,-I agree with Dr A R J Cale and colleagues that the possibility ofvasectomy affecting testicular malignancy is important in view of the current trend towards male sterlisation.' I disagree, however, that a large prospective study is the best approach at this juncture. The finding of a significantly higher standardised incidence ratio of testicular tumour for men who had a vasectomy and the short average interval (1-9 years) between vasectomy and malignancy have led Dr Cale and colleagues to suspect "an association" and that "vasectomy accelerates the tumour."' Studies on the relation between vasectomy and testicular cancer have been few and the findings tenuous. Vasectomy was not considered as a potential risk factor in a review of the epidemiology of testicular cancer published in 1989,2 and Moss et al did not find an association between vasectomy and testicular cancer among American men aged 18-40.' A case-control analysis performed specifically to test this hypothesis did find an association, but it was restricted entirely to Catholic men.4 The investigators suspected that the finding may have been spurious owing to selective underreporting of vasectomy among the Catholic controls.4 The finding by Dr Cale and colleagues of an excess ofobserved number of cases in only the 30-35 age group presents a problem similarly deserving an explanation. An association in time between two variables can be due to many reasons. More convincing evidence is needed before an expensive, logistically difficult, and time consuming prospective approach is to be undertaken. The incidence of testicular cancer is very low (5 6 per 100000 white men aged 20-69 in the United States).4 As Strader et al pointed out, even if data of four previous large multicentre studies involving 114000 men (with an average follow up period of five years) were pooled for analysis the study power would only be 0 2 (for the detection of
Details offour patients with chronic obstructive pulmonary disease admitted with severe hvpercapnic respiratoiy failure Age Case and No sex 1 68 F 2* 73 F 3 70 M 4 72 F
FV,1 (1)
1-1 0-6 0-6 08
Method At prcsentation (with air) of sentilation Ppo (kPa) P,cot (kl'a)
NIPPV NIPPV NCPAP NCPAP
3-9 4-7 4-5 4-1
8X4 84 73
90(
FEV1l = Forced expiratory volume in one second. Nll'PV=Nasal intermittent positive pressure ventilation.
Assisted
During tonvsalescence
P,o. (kPa) P(o, (kl'a) 80 653 6-4 69
8X6 78 5 3 57
Pa0 (kPa
P,tco (kPa)
Outcome
6-8 6.6
5-0 63
Alive at four mottths Diied after four months Died after three weeks D)ied atter fihe davs
NCP'AP'- Nasal continttous positivc atrwsavs pressure. * Ihe patiettt also had kyphasis.
BMJ VOLUME 300
7 APRIL 1990
contraception. No patient complained or had svmptoms. Rarely the family and sometimes staff reported a change in behaviour. The need for routine medical examinations seems self evident. It is effective to use clinical medical officers, who are accustomed to assessing potential in those without speech, flexible in their approach to different venues, and accustomed to liaison within a health care team. For the patients, being examined in a familiar setting with the reassuring presence of their community nurse is non-threatening. For carers, the long appointment encourages useful dialogue with the clinical medical officer and communitv nurse, which, it is hoped, reduces the burden on them-and on the state. MAUREEN BEAUCHAMP
leiceseirshire Area Health Authorit%, I.cester LEI 6TY I Ashley-Miller M. Community care. Br Medj 1990;300:487-8.
V24 February.i
2 Bicknell J. (Consent and people with mcntal handicap. Br Medj 1989;299:1176-7. I I November.;
Research ethics committees SIR,-I sympathise with Dr Ann Cartwright's view that researchers using samples taken from public records should not be required to submit their protocols to district health authority ethics committees.' I disagree, however, with her assertion that when considering sociological studies ethics committees should confine themselves to questions of confidentiality of data and not question the design of the study. Most of the protocols that ethics committees see concern research on NHS patients. As a patient representative I want the committees to examine the quality of all the research they are asked to approve, but as a former lay member of a research ethics committee I know that they are woefully inadequate to do this. They are not required to have a social scientist as a member-an important omission in the Department of Health's draft guidelines. Their ignorance of another discipline does not prevent medical members, who are in the majority, from thinking that they know how such research should, or should not, be done. I have seen studies either go through on the nod because it was wrongly assumed that there were no ethical issues involved or be criticised in a way that could damage carefully structured questionnaires. There is an increasing volume of research in nursing and midwifery exploring issues that are of great value to patients, yet researchers are deprived of informed and educational advice by the ethics committees that look at their projects. I have no doubt that there are issues of major ethical importance in approval or amendment of
BMJ VOLUME 300
7 APRIL 1990
these studies, and it mav be that they should go to separate committees that are not primarily medical. One problem at present is that proposals from a medical source are more likely to be approved - yet some of such "sociological" research published in medical journals is of such poor quality that no social science based ethics committee would have passed it and no decent sociological journal would have published it. Doctors seem unaware that questions asked before and after patients have left hospital are likely to get different answers and that being interviewed by or receiving a questionnaire from a doctor, nurse, or medical student may affect the patient's reply. Patients who know or suspect that their doctors will see their replies are unlikely to be critical -but the advantage to the doctors is that the "survey" shows how much patients like their care. Many doctors and research ethics committees seem unaware that the abilitv to design and analyse questionnaires does not come automatically with a medical degree. Future management of patients mav be based on the results of such "studies"; poor quality protocols are dangerous to patients and we have a right to be protected from them. The only question is whether the present ethics committees can be trained to do the job or whether we need different committees for social science research.
seem unequivocal about the need for separate submissions to local committees.' On the other hand, the recently published guidelines from the Royal College of Physicians on the practice of ethics committees acknowledges that multiple applications may be impracticable "so that research that is valuable to the community can be impeded or may even not be undertaken."4 The confusing nature of the present situation is thus all too evident. The articles, commentaries, and letters published over the past few months have revealed a range of serious problems. So far, these have not been debated sufficiently thoroughly around the same table by those concerned. T W MEADE
MRC Epidemiology and Medical Care Unit, Northwick Plark Hospital, Harrow, Middlesex HAl 3UJ I Drury Mi. Research ethics committees. Br Med7 1990;300:608. (3 March.) 2 Meade IW. Research ethics committees. Br Alcd J 1990;300: 396. (10 February., 3 Marshall T, Moodie P. Research ethics committees revisited. BrMed] 1989;299:1419-20. (9 December.) 4 Royal College of Physicians. Guidelintes on the practice o ethift/s commtttees tn medical research involving human subjects. 2nd cd. London: Royal College of Physicians, 1 99.
JEAN ROBINSON
()xford (X2 7ti' I Cartwright A. Research ethics committces. Br lcdj 1990;300: 607. (3 March.)
SIR,-The establishment of a clinical trials ethics committee by the Royal College of General Practitioners referred to by Sir Michael Drury in his letter' is, I think, another indication that some kind of central committee is needed, though for several reasons I am not confident that it solves the problems I was referring to.Firstly, many multicentre studies are, of course, carried out through hospitals. Rather than having to rely on yet another "central" committee on these occasions-set up, for example, by the Royal College of Physicians or the Royal College of Surgeons-it would be much better to have a single committee responsible for all multicentre studies, however they were conducted. Secondly, our own experience of local reactions to an arrangement similar to the committee described by Sir Michael was not encouraging. In 1988, at the suggestion of the Department of Health, we planned a feasibility study of anonymous HIV testing. To allow an early start and also because of the special considerations involved we approached the convener of the committee on ethical issues in medicine of the Royal College of Physicians (and I am grateful to Sir Douglas Black for agreeing that I may refer to our exchange). He consulted the then president and another member of the committee and approved our proposal. When the general practitioners in our study consulted their nine local ethical committees, however, these committees nearly all asked for separate submissions even though they knew the work had been approved centrally. If the decisions of any future national committee were not binding on local committees, the undoubted problems of multiple applications to separate committees would therefore remain. Eventually, the local committees also accepted the HIV study-with one exception, which illustrates the potential for idiosyncratic decisions. That particular committee declined the proposal on the grounds that anonymous testing would leave subjects positive for HIV unidentified and uncounselled, even though the national scheme for anonymous testing had been announced and was about to start. Finally, the Department of Health guidelines that you referred to in the BMJ's news section
Bovine spongiform encephalopathy SIR,-In his editorial Professor W B Matthews discusses bovine spongiform encephalopathy.' The publicity given to this disease and its possible public health implications has prompted a reexamination of the relation between it and similar diseases found in humans and other animals-the transmissible dementias (box).'
The transmissible dementias Disease
Animal affected
Scrapic
Sheep, goats Mink M\ule deer, elk Cattle Humanis
Transmissible mink encephalopathy Chronic wasting disease Bovine spongiform encephalopathy Sporadic Creutzfeldt-Jakob disease Familial Creutzfeldt-Jakob disease
Gerstmann-Straussler svndrome Kuru
Humans Humans Humans
All of the transmissible dementias share a common factor-the involvement of an aberrant isoform of prion protein.' This protein is the product of a highly conserved gene found in organisms as diverse as the fruit fly and humans.' The abnormal isoform of prion protein is present in brain extracts from patients with transmissible dementias and is the main constituent of the amyloid plaques in these patients.3-6 Its presence distinguishes transmissible dementias from conditions such as Alzheimer's disease and can be regarded as diagnostic of transmissible dementias.4i Distinct prion gene alleles are associated with long and short disease incubation times in mice, and linkage between the GerstmannStraussler syndrome and both a base substitution' and an insertion in the prion gene"" has been shown. The evidence suggests that in all animals the aberrant isoform of prion protein has a central and possible causal role in the genesis and pathology of transmissible dementias. Over the past two years we have screened patients with a wide range of familial dementias and ataxias and those with apparently sporadic symptoms to determine whether these diseases are in fact part of the prion disease range. Using both immunocytochemical and molecular biological techniques we have shown that several families labelled as having atypical dementias were suffering from variants of prion disease. "'
943
The recent work on the pathology and genetics of prion disease in humans has several implications for the epidemiological proposals at the end of Professor Matthews's editorial. Firstly, because of the clinical and pathological variability the true incidence of transmissible dementias (prion disease) in humans can be established with certainty only by a combination of genetic screening and immunocytochemical detection of the abnormal isoform of prion protein. Secondly, on the basis of information already available the incidence of the range of prion diseases in humans is (and probably always was) significantly higher than the figures quoted in the editorial. Thirdly, because of this improved rate of detection any attempts at ascertaining the possible impact of foodstuffs contaminated with bovine spongiform encephalitis on the incidence of transmissible dementias must take this apparent rise in the number of cases into account. Finally, the number of familial cases being identified raises the possibility that in humans the prion disease range has more of a genetic basis than hitherto realised. G W ROBERTS
St Mary's Hospital Medical School, London W2 lPG J COLLINGE
Clinical Research Centre, Harrow, Middlesex 1 Matthews WB. Bovine spongiform encephalopathv. Br Med7 1990;300:412-3. (17 February.) 2 Beck E, Daniel PM. Neuropathology of transmissible spongiform encephalopathies. In: Prusiner SB, McKinley MP, eds. Prtons: novel infectious pathogens causing scrapie and Creutzfeldt-
Jakob disease. San Diego: Academic Press, 1987:331-3. 3 Westaway D, Carlson GA, Prusiner SB. Unravelling prion diseases through molecular genetics. Trends in Neurosciences 1989;12:22 1-7. 4 Roberts GW, Lofthouse R, Brown R, Crow TJ, Barry RA, Prusiner SB. Prion protein immunoreactisvity in human transmissible dementia. N Eng1i Ied 1986;315:1231-3. 5 Roberts GW, Lofthouse R, Allsop D, et al. CNS amvloid proteins in neurodegenerative diseases. Neurolog 1988;38: 1534-40. 6 Hope J, Reekie LJD, Hunter N, et al. Fibrils from brains of cows with new cattle disease contain scrapie-associated protein. Nature 1988;336:390-2. 7 Westaway D, Goodman PA, Mirenda CA, McKinley MP, Carlson GA, Prusiner SB. Distinct prion proteins in short and long scrapie incubation period mice. Cell 1987;51:651-2. 8 Hsiao K, Baker HF, Crow TJ, et al. Linkage of a prion protein missense variant to Gerstmann-Straussler syndrome. Nature 1989;338:342-5. 9 Owen F, Poulter M, Lofthouse R, et al. Insertion in prion protein gene in familial Creutzfeldt-Jakob disease. Lancet
1989;i:51-2. 10 Collinge J, Harding AE, Owen F, et al. Diagnosis of GerstmannStraussler syndrome in familial dementia with prion protein gene analysis. Lancet 1989;ii: 15-7.
SIR,-Drs Christopher Fear and Manikkarasa Devakumar' take issue with Professor W B Matthews over his statement that all patients with Creutzfeldt-Jakob disease must be seen by a neurologist who is familiar with the disease-: instead they would substitute electroencephalography as a baseline investigation. Those following this recommendation are more likely to miss the diagnosis than a physician or psychiatrist prepared to review the diagnosis at each stage in the light of clinical evidence. In practice the electroencephalogram is rarely helpful early in the disease, and the best a neurophysiologist reading it can do is to suggest that the test be repeated at intervals. Brown et al observed that the characteristic periodic electroencephalogram complexes were found comparatively late in the illness.' Although a pathognomic electroencephalogram is eventually seen in about 75% to 80% of patients, this may not be the case in certain groups of patients with Creutzfeldt-Jakob disease -for example, those receiving human growth hormone therapy.4 E M R CRITCHLEY
Department of Neurology, Royal Preston Hospital, Preston PR2 4HT
944
I Fear C, Devakumar M. Bovine spongiform cincephalopathy. BrAled_ 1990;300:817. (24 March. 2 Matthews >'B. Bovine spx)ngiform encephalopathy. lBr Med 7 1990;300:912-3. 17 Februarv. 3 Brown P, Cathala F, Castaigne 1', (iajiusck D)C. CretitzlclditJakob disease: clinical analysis of a consecutive series ot 230 neuropathically verified cases. Ann Neurol 1986;20:597-602. 4 Brown P. Thc dccline anid fall of CreLutzfeldt-Jakob disease associated with htiman growth hormone thcrapy. Neurology
1988;38:1135-7.
Non-invasive mechanical ventilation for acute respiratory failure
arterial carbon dioxide tension with this technique. In summary, we believe that both nasal intermittent positive pressure ventilation and continuous positive airways pressure may be useful adjunct treatments in patients with hvpercapnic respiratorv failure, particularly those who are not suitable for assisted ventilation through an endotracheal tube. In our hands this treatment still requires close supervision by medical, phvsiotherapy, and nursing staff, and at present we prefer to treat these patients in high dependency settings. D BELL M W Me NICOI.
J F RIORDAN C PRATT S LAM
Central Mliddlesex Hospital, London NWO 7NS
SIR, - In their paper Dr M W Elliott and colleagues describe the use of nasal intermittent positive pressure ventilation in patients with acute hypercapnic respiratory failure.' Of their group, three patients had chronic obstructive pulmonary disease. We have recently managed four patients with this condition with associated acute hypercapnic respiratory failure. Two patients were treated using nasal intermittent positive pressure ventilation, and in the other two patients we used nasal continuous positive airways pressure. The table gives the patients' details. All four patients had known severe chronic obstructive pulmonary disease and were admitted with severe hypercapnic respiratory failure; two of them had associated respiratory acidosis. None of the patients were considered suitable for endotracheal intubation and assisted ventilation because of end stage chronic disease. All four patients had supplemental oxygen at a flow of between 0-5 1/min and 2-0 1/min entrained through a port in the nasal mask. Arterial oxygen tensions increased in all patients from a range of 3 9-4-7 kPa to 6-3-8 0 kPa with nasal intermittent positive pressure ventilation or continuous positive airways pressure. In three patients there was a fall in arterial carbon dioxide tension; in the other patient, who was treated by nasal intermittent positive pressure ventilation, the arterial carbon dioxide tension was maintained (table). Nasal continuous positive airways pressure was well tolerated, but both patients in whom ventilation was assisted by nasal intermittent positive pressure ventilation required chin straps to prevent excessive air loss during the inspiratory phase. Two patients survived to be discharged from hospital: one was alive and well several months after discharge, and one died of a bronchogenic carcinoma four months after discharge. Of the two who died in hospital, one died of progressive renal failure despite correction of arterial gas imbalance and the other died of progressive respiratory failure three weeks after admission despite an initial response. In agreement with Dr Elliott and colleagues we have found nasal intermittent positive pressure ventilation to be of benefit in two patients with severe hypercapnic respiratory failure, but in addition two other patients benefited from the use of nasal continuous positive airways pressure, which has been used previously to manage patients with hypoxia but without hvpercapnial and in the treatment of obstructive sleep apnoea.' In the two patients described we achieved improvement in arterial oxygen tension with an associated fall in
I Elliott MW, Steven MH, I'hillips GD, Braithwaitc MA. Non-invasive mechanical ventilation for acute respirators failure. Br.MedJ 1990;300:358-60. (10 February.} 2 Kesten S, Rebuck AS. Nasal continuous positive airway pressturc in Poneumocystis carinii pneumonia. Lancet 1988;ii: 1414-5. 3 Sullivan CE, Berthon-Joncs M, Issa FG, Eves L. Reversal ot obstructive sleep apnoca by continuous positive airwav pressure applied through the nares. Lancet 1981 :862-5.
Vasectomy and testicular cancer SIR,-I agree with Dr A R J Cale and colleagues that the possibility ofvasectomy affecting testicular malignancy is important in view of the current trend towards male sterlisation.' I disagree, however, that a large prospective study is the best approach at this juncture. The finding of a significantly higher standardised incidence ratio of testicular tumour for men who had a vasectomy and the short average interval (1-9 years) between vasectomy and malignancy have led Dr Cale and colleagues to suspect "an association" and that "vasectomy accelerates the tumour."' Studies on the relation between vasectomy and testicular cancer have been few and the findings tenuous. Vasectomy was not considered as a potential risk factor in a review of the epidemiology of testicular cancer published in 1989,2 and Moss et al did not find an association between vasectomy and testicular cancer among American men aged 18-40.' A case-control analysis performed specifically to test this hypothesis did find an association, but it was restricted entirely to Catholic men.4 The investigators suspected that the finding may have been spurious owing to selective underreporting of vasectomy among the Catholic controls.4 The finding by Dr Cale and colleagues of an excess ofobserved number of cases in only the 30-35 age group presents a problem similarly deserving an explanation. An association in time between two variables can be due to many reasons. More convincing evidence is needed before an expensive, logistically difficult, and time consuming prospective approach is to be undertaken. The incidence of testicular cancer is very low (5 6 per 100000 white men aged 20-69 in the United States).4 As Strader et al pointed out, even if data of four previous large multicentre studies involving 114000 men (with an average follow up period of five years) were pooled for analysis the study power would only be 0 2 (for the detection of
Details offour patients with chronic obstructive pulmonary disease admitted with severe hvpercapnic respiratoiy failure Age Case and No sex 1 68 F 2* 73 F 3 70 M 4 72 F
FV,1 (1)
1-1 0-6 0-6 08
Method At prcsentation (with air) of sentilation Ppo (kPa) P,cot (kl'a)
NIPPV NIPPV NCPAP NCPAP
3-9 4-7 4-5 4-1
8X4 84 73
90(
FEV1l = Forced expiratory volume in one second. Nll'PV=Nasal intermittent positive pressure ventilation.
Assisted
During tonvsalescence
P,o. (kPa) P(o, (kl'a) 80 653 6-4 69
8X6 78 5 3 57
Pa0 (kPa
P,tco (kPa)
Outcome
6-8 6.6
5-0 63
Alive at four mottths Diied after four months Died after three weeks D)ied atter fihe davs
NCP'AP'- Nasal continttous positivc atrwsavs pressure. * Ihe patiettt also had kyphasis.
BMJ VOLUME 300
7 APRIL 1990
