Oxpentifylline treatment of venous leg ulcers SIR, -We do not accept the conclusion of Dr Mary Paula Colgan and colleagues that "oxpentifylline used in conjunction with compression bandaging improves the healing of venous ulcers of the leg."' The criteria on which a diagnosis of venous ulceration was based were not stated and it seems that the presence of venous disease was not confirmed by phlebography, plethysmography, or duplex imaging. In a study set in four centres objective criteria for uniform selection of patients should be mandatory. Other diseases that might delay healing of venous ulcers can be overlooked if diagnosis is based only on clinical impression and arterial Doppler pressure measurements. In our clinic over 20 months venous disease was thought to be the only aetiological factor in 137 patients with leg ulceration. This provisional diagnosis was based on clinical examination and arterial Doppler pressure measurement. After photoplethysmography, blood glucose measurement, and serological investigation, however, 24 patients were found to have diabetes or rheumatoid arthritis, or both, in addition to venous disease. In three patients investigation did not confirm the clinical impression of venous disease. The healing rate of the control group (34% at six months) is lower than would be expected from other published work." In our clinic, using graduated compression bandages, the healing rate of ulcerated legs in which venous disease has been confirmed by photoplethysmography and for which other causes of ulceration have been excluded is 73% at six months. The healing rate of ulcerated legs in which venous disease is not the only aetiological factor is lower (67% at six months). The authors do not comment on the low healing rate of their control group. They did not exclude aetiological bias and we suspect that the causes of ulceration in the two groups may not be the same. Although it is difficult to compare different studies directly, we do not accept that the healing rate of leg ulcers treated with oxpentifylline and compression bandaging (64% at six months) is an improvement on current methods of treatment where a venous aetiology has been established firmly. H GAJRAJ ANDREA JOPP-McKAY M C STACEY
University Department of Surgery Fremantle Hospital, Fremantle 6160, Western Australia I Colgan MP, Dormandy JA, Jones PW', Schraibman IG, Shanik DG, Young RAL Oxpentif lline treatment of venotus ulcers of the leg. Br Mlgedj 1990;300 972-5. (21 April.) 2 Backhouse CAlN, Blair SD), Savage Al', Wlalton J, MicCollum CN. Controlled trial of occluIsie dIressings in healing chronic vctnous ulcers Br]Surg 1987;74:626-7. 3 Northeast ADR, Laser GT, Staces INIC, Wilson NM? Browse NL, Burnand KG Elfcct of fibrinolvtic enhancement of venous ulcer healing Brj Surg 1989;76: 1332.
AIJTHOR'S REPLY,-The presence of venous disease was confirmed in all cases by non-invasive testing. TFhough diabetic patients were not excluded if their ulcers were considered to be venous in origin, only three patients in the entire group were diabetic and all three received oxpentifylline. Additionally, at the first visit plasma glucose concentrations were measured to ensure that occult diabetes had not been overlooked. We agree that the overall healing rate may appear low, but this was not a normal population of patients with venous ulcers but a group of patients resistant to treatment. All patients had been attending a specialist clinic for at least two months with no improvement in their ulcers. These patients will, therefore, have a lower healing rate than those reported. We feel that our conclusion that oxpentifylline BMJ
VOLUME 300
30 JUNE 1990
used in conjunction with compression bandaging improves the healing of venous ulcers is indeed justified. MARY PAULA COLGAN
St James's Hospital, PO Box 580, Dublin
Health checks in general practice SIR, -Dr Deborah Waller and others invited men and women in the age group 35-64 to a health check for cardiovascular disease.' The only indication for this invitation was the age of the patient, and the group that attended were described as "the worried well." This inverse care effect could have been reduced if case finding had been used in the target group before invitations were given. This can be done with a questionnaire about signs and symptoms of cardiovascular disease, history of cardiovascular disease or hypercholesterolaemia in relatives under 60 years, diabetes mellitus, hypertension, severe obesity, and smoking 20 or more cigarettes a day. Only those with one or more positive answers to these questions should be invited to a check on cardiovascular disease. People at high risk can then be invited, not just because they are the correct age. A L C J VAN DEN HOGEN
4WXYeerder, 1141 JE, Monnickendam, rhe Netherlands I Waller D, Agass Mi, Mant D, Coulter A, Fuller A, Jones L. Health checks in general practice: another example of inverse
care? Br MedJ7 1990;300:1115-8. (28 April.)
AUTHORS' REPLY, -Dr van den Hogen suggests that initial screening for cardiovascular risk is done by questionnaire so that only those patients at relatively high risk need be invited for appropriate counselling and advice. This would certainly allow more time to be spent dealing with risk factors rather than simply identifying them. Of course, if questionnaires were sent by post this would not necessarily avoid the problem of inverse care because of the possibility of poor response from those at greatest risk. But the new contract requires that risk factors are sought when patients first register with the practice, and this can be done by the doctor, practice nurse, or reception staff when the patient first visits the surgery. In Berinsfield the intention is to use such records to target intervention. As the proportion of general practitioners that uses computerised records grows, data on which selective screening could be based will be increasingly accessible to general practitioners without special questionnaires. But there are difficulties with this approach. Firstly, the proportion of patients with at least one cardiovascular risk factor is high. We analysed recently data collected from 1047 health checks and found that 451 patients (43 1%) had either a personal history of coronary heart disease, diabetes, hypertension, or long term smoking (more than 10 years) or a family history of premature coronary heart disease (under 50 years). If obesity is added to this list of selective criteria the proportion of patients included is even higher. Secondly, 85 of the 1047 patients had a total plasma cholesterol over 7-8 mmolUl but only 50 (58 8%) had one or more of the selective criteria cited. Thirdly, self reports of smoking elicited by questionnaire may be unreliable. None the less, the separation of screening and intervention, with more emphasis on intensive intervention in selected patients, may be a way forward. In the near future we will be able to address Dr van den Hogen's suggestion directly because we will be analysing interim data from the
"Ox-check" trial. This is a randomised controlled trial of the effectiveness of health checks in reducing cardiovascular risk, and completion of a questionnaire, which includes the specific risk factors mentioned by Dr van den Hogen, is a precondition of recruitment to the study and the offer of a health check. L ROE
M AGASS D WALLER
J MUIR
D MANT
A COULTER
Radcliffe Infirmary, Oxford 0X2 6HE
Meningococcal meningitis SIR,-The editorial by Mr Philip D Welsby and Dr Clayton L Golledge on meningococcal meningitis' merits further comment on indications for hospital admission. In Israel recently experience has been accumulating regarding patients with a genetic predisposition to meningococcal infection due to a deficiency of the terminal complement components (C5-C9). A recent review of 111 survivors of sporadic meningococcal disease in 10 Israeli hospitals found an incidence of C7 and C8 deficiency of 18% in patients of Sephardic origin (and an incidence of 40% in patients of Moroccan origin), and cases among Arabs have also been identified.2 The clinical characteristics of the illness were different in this group, including a higher age at first presentation and a high incidence of recurrence (40%), and a high proportion of the patients had a family history of meningococcal infection.2 Some patients may present with fever and no relevant clinical signs (Malnick and others, unpublished observations), as may also occur in children with meningoccocal infection.' Inherited deficiencies of the complement system that are associated with increased susceptibility to meningococcal infections have also been reported in white and black people4 and may be more prevalent than currently realised. One American study of 20 consecutive patients presenting with a first known episode of meningococcal meningitis disclosed three cases of "idiopathic" terminal pathway protein deficiency.5 Thus a history of meningococcal infection or a positive family history should be added to the criteria for hospital admission suggested by Mr Welsby and Dr Golledge. S D H MALNICK A SCHATTNER Division of Internal Medicine,
Kaplan Hospital,
Rehosot and Hadassah Miedical School, Jerusalem, Israel
laD, Golledge CL. Mleningococcal meningitis. Br.Medj 1990;300:1150-1. (5 May. ) Zimran A, Rudenskv B, Kramer MR, et al. Hereditary complement deficiency in survivors of meningo occal disease: high prevalence of C7/C8 deficiency in Sephardic (Moroccan) Jews. QJfMed 1987;240:349-58. Dashefskv B, Teek DW, Klein JO. Unsuspected meningococcemia. 7Pediatr 1983;103:69-73. Ross SC, Densen P. Complement deficiency states and infection: epidemiology, pathogenosis and consequences of neisserial and other infections in an immune deficiency. Medicine 1984;863:243-73. Ellison RT, Kohler PF, Curd JG, Judson FN, Reller lIB. Prevalence of congenital or acquired complement deficiency in patients with sporadic meningococcal disease. N Engl j Med 1983;308:913-6.
1 Welsby 2
3 4
5
Coronary artery bypass surgery SIR,-In their editorial Mr E W J Cameron and Mr W S Walker draw attention to the important issue of selection of the conduit in coronary artery bypass surgery.' Graft failure is more common with saphenous vein conduits than with internal mammary artery grafts. Consequently we were 1725
surprised by the relatively low proportioi of patients (two fifths) receiving internal mammary artery grafts according to the United Kingdom cardiac surgical register. We analysed the last 100 consecutive patients who underwent coronary bypass surgery in our hospital from December 1989 to May 1990. Ninety nine patients received the following arterial grafts: left internal mammary artery (75); both internal mammary arteries (22); right internal mammary artery (1); both internal mammary arteries and gastroepiploic artery (1). A sequential left internal mammary graft was used in three patients, and a free right internal mammary graft was used in three patients. The mean number of distal venous graft anastomoses was 2 5. The operative mortality was zero. Two patients had postoperative mediastinitis; both recovered well. We hope that our favourable experience with arterial conduits will encourage their increased use. TIMO PELLINEN ANTERO SAHLMAN
TUULA KURKI ARI HARJULA
Heart Centre, Deaconess Medical Centre, Alppikatu 2, SF-00530 Helsinki, Finland I Canmeroni EWJ, Walker WS. Coronary artery bypass surgery. BrMedj 1990;300:1219-20. (12 May.)
Births resulting from assisted conception SIR,-The data on the incidence of congenital malformations in babies resulting from assisted conception as reported by the MRC working party on children conceived by in vitro fertilisation' seem to be lacking in one important respect. Women who conceive after in vitro fertilisation (IVF) or gamete intrafallopian transfer (GIFT) would probably have access to skilled detailed fetal ultrasonography in the second trimester to detect congenital malformations, and pregnancies complicated by serious abnormalities may well have been terminated. While some terminations for malformations (three for anencephaly, one for achondroplasia, and one for translocation) are included it is not clear whether or not the working party made any attempt to ascertain all pregnancies that ended in therapeutic termination because of fetal malformation. Nor are there any obvious data included on any fetal losses that occurred before 28 weeks in pregnancies that may have been complicated by an abnormality. It is only with the addition of these data that the true incidence of congenital malformations in these pregnancies will be known. A precise diagnosis of each malformation is required not only as good clinical practice in order to advise parents but also to stand any chance of assessing whether the handling of gametes or preembryos predisposes to errors in development. For example, it is surprising that achondroplasia was one of the disorders diagnosed early enough to allow a therapeutic abortion; in our experience in many such cases there turns out to be a more severe bone dysplasia on expert assessment. The authors point out that malformations like anencephaly occur more commonly in twin pregnancies, but this mainly applies to monozygotic twins. Presumably most of the twins reported in this series were dizygotic. If, indeed, a considerable number were monozygotic this could have implications with regard to the incidence of malformations and perhaps the adverse effect of gamete handling. Monozygotic twinning can be viewed as a malformation and therefore the incidence, for which there are good control data, should be monitored. While we agree that it would be difficult to obtain ideally matched control data for all pregnancies, some form of prospective control
1726
study with respect to monitoring fetal malformations may be feasible. The forthcoming Statutory Licensing Authority set up to regulate IVF services (and, it is to be hoped, to monitor GIFT services as well) should take the lead in this matter. LYN CHITTY MARCUS PEMBREY Institute of Child Health, University of London, London WC IN I EH 1 MRC Working Party on Children Conceived by In Vitro Fertilisation. Births in Great Britain resulting from assisted conception, 1978-87. Br MedJ7 1990;300:1229-33. (12 May.)
AUTHORS' REPLY, --The MRC register was set up to establish whether the children resulting from assisted conception differed materially from children conceived normally. Thus we concentrated on viable pregnancies and on health outcomes that are reliably measured in the general population. For congenital malformations the emphasis was therefore on conditions diagnosed in liveborn and stillborn infants and on prenatally diagnosed fetal malformations in pregnancies that were terminated. We agree that our data do not include spontaneous abortions due to malformations and so do not necessarily give the "true" incidence of malformations. Adequate statistics for comparison do not exist, and a different study design would have been necessary to estimate the true incidence of malformations. Every contributing clinic was asked specifically about terminations because of malformations, and it is unlikely that any are missing as the clinics are particularly aware of such events. We agree that the risk of various malformations does differ for monozygotic and dizygotic twins. Virtually all the twins resulting from assisted conception are dizygotic, and so we did not ask specifically about this. Among the 219 sets of twins in the MRC register 54 were both boys, 52 were both girls, and 113 were a boy and a girl, which is consistent with the sex distribution of dizygotic twins. From 1 January 1989 the Interim Licensing Authority requires that data on the outcomes of in vitro fertilisation treatment cycles are collected by the clinics themselves and sent directly to the Office of Population Censuses and Surveys. We are continuing to collect information on babies born before then. VALERIE BERAL
Imperial Cancer Research Fund, Cancer Epidemiology Unit, John Radcliffe Infirmary, Oxford OX2 6HE PAT DOYLE
London School of Hygiene and Tropical Medicine, London WC1E 7HT
Allergy to peanuts SIR,-The editorial by Dr Tony Smith and the short report by Dr E S K Assem and colleagues and Drs Kieron L Donovan and J Peters on peanut allergy highlighted the potential severity of reactions to peanuts and the importance of recognising this sensitivity.`3 To assess the prevalence of IgE antibody to peanut and 10 other food allergens we investigated the serum of 848 patients with suspected food allergy and found that peanut was the most commonly positive allergen (table). When a serum IgE value of radioallergosorbent test grade 3 or above was taken as a strong indicator of clinical importance4 peanut sensitivity occurred in 70 patients, mostly in conjunction with sensitivity to other foods, though eight had positive results for peanut alone. This comparatively high prevalence of seropositivity and the potential serious consequences of inadvertent ingestion should encourage clinicians to include IgE to peanuts in their investigations of food allergy and alert sensitised subjects to check for peanut,
Prevalence of sensitivity to food allergens No of seropositive patients
Allergen Peanuts Wheat Eggs Oats Rye Milk
(No singly seropositive) 70 (8) 67 (5) 65 (18) 60 (5) 48 (1) 46 (9)
Allergen Barley Peas Fish Shrimps Tomatoes
No of seropositive patients (No singly seropositive) 44 34 (6) 24 (5) 16 (2) 16
even in foods where this would be unexpected. For example, peanut, being a legume, can be included in the vegetable oil component of some foods. CHARLES McSHARRY
Clinical Immunology Laboratory, Western Infirmary, Glasgow G 1I 6NT 1 Smith T. Allergy to peanuts. BrMedI1990;300:1354. (26 May.) 2 Assem ESK, Gelder CM, Spiro SG, Boderman H, Armstrong RF. Anaphylaxis induced by peanuts. Br Med 7 1990;300: 1377-8. (26 May.) 3 Donovan KL, Peters J. Vegetable burger allergy: all was nut as it appeared. BrMedj 1990;300:1378. (26 May.) 4 Barnetson RStC, Benton EC, MacFarlane HAF, et al. Food allergy: the scientific approach. Laboratory Practice 1985;34: 84-6.
Management of the upper airway SIR,-In response to Dr D Watson's article' I would like to make the following points, which may benefit both the anaesthetist and the nonanaesthetist. Correct placement of the tracheal tube must always be confirmed as any ensuing cyanosis may wrongly be ascribed to causes other than oesophageal intubation. Auscultation ofthe axillae, as recommended in the article, will certainly detect right bronchial intubation but will fail to detect 15% of accidental oesophageal intubations, whereas auscultation of the axillae and epigastrium will detect all cases.2 The total reliability of the oesophageal detector described by Wee3 in assessing the position of the tracheal tube has been confirmed.2 When connected to a correctly placed tube this device, a 50 ml syringe or rubber bulb, allows free aspiration of air from the trachea; resistance to aspiration indicates oesophageal intubation. These two techniques should prevent undetected oesophageal intubation occurring during resuscitation. The laryngeal mask airway has been used to resuscitate a patient in whom tracheal intubation was not feasible.4 The mask airway provides an alternative to endotracheal intubation when the operator is inexperienced or abnormal anatomy exists. Insertion of the airway is generally easy. The laryngeal mask airway will not fully protect the tracheobronchial tree against aspiration as the cuff does not effect a complete seal. Its use, however, may save lives. I therefore recommend that Wee's oesophageal detector device and the laryngeal mask airway become part of the standard resuscitation equipment available today. MARK LYNCH
Department of Anaesthetics, Royal Free Hospital, London NW3 2QG 1 Watson D. ABC of major trauma: management of the upper airway. BrMedj7 1990;300:1388-91. (26 May.) 2 Andersen KH, Hald A. Assessing the position of the tracheal tube. The reliability of different methods. Anaesthesia 1989;44: 984-5. 3 Wee MYK. The oesophageal detector device. Assessment of a new method to distinguish oesophageal from tracheal intubation. Anaesthesia 1988;43:27-9. 4 Calder I, Ordman AJ, Jackowski A, Crockard HA. The brain laryngeal mask airway. An alternative to emergcncy tracheal
intubation. Anaesthesia 1990;45:137-9.
BMJ
VOLUME 300
-
30 JUNE 1990
University Department of Surgery Fremantle Hospital, Fremantle 6160, Western Australia I Colgan MP, Dormandy JA, Jones PW', Schraibman IG, Shanik DG, Young RAL Oxpentif lline treatment of venotus ulcers of the leg. Br Mlgedj 1990;300 972-5. (21 April.) 2 Backhouse CAlN, Blair SD), Savage Al', Wlalton J, MicCollum CN. Controlled trial of occluIsie dIressings in healing chronic vctnous ulcers Br]Surg 1987;74:626-7. 3 Northeast ADR, Laser GT, Staces INIC, Wilson NM? Browse NL, Burnand KG Elfcct of fibrinolvtic enhancement of venous ulcer healing Brj Surg 1989;76: 1332.
AIJTHOR'S REPLY,-The presence of venous disease was confirmed in all cases by non-invasive testing. TFhough diabetic patients were not excluded if their ulcers were considered to be venous in origin, only three patients in the entire group were diabetic and all three received oxpentifylline. Additionally, at the first visit plasma glucose concentrations were measured to ensure that occult diabetes had not been overlooked. We agree that the overall healing rate may appear low, but this was not a normal population of patients with venous ulcers but a group of patients resistant to treatment. All patients had been attending a specialist clinic for at least two months with no improvement in their ulcers. These patients will, therefore, have a lower healing rate than those reported. We feel that our conclusion that oxpentifylline BMJ
VOLUME 300
30 JUNE 1990
used in conjunction with compression bandaging improves the healing of venous ulcers is indeed justified. MARY PAULA COLGAN
St James's Hospital, PO Box 580, Dublin
Health checks in general practice SIR, -Dr Deborah Waller and others invited men and women in the age group 35-64 to a health check for cardiovascular disease.' The only indication for this invitation was the age of the patient, and the group that attended were described as "the worried well." This inverse care effect could have been reduced if case finding had been used in the target group before invitations were given. This can be done with a questionnaire about signs and symptoms of cardiovascular disease, history of cardiovascular disease or hypercholesterolaemia in relatives under 60 years, diabetes mellitus, hypertension, severe obesity, and smoking 20 or more cigarettes a day. Only those with one or more positive answers to these questions should be invited to a check on cardiovascular disease. People at high risk can then be invited, not just because they are the correct age. A L C J VAN DEN HOGEN
4WXYeerder, 1141 JE, Monnickendam, rhe Netherlands I Waller D, Agass Mi, Mant D, Coulter A, Fuller A, Jones L. Health checks in general practice: another example of inverse
care? Br MedJ7 1990;300:1115-8. (28 April.)
AUTHORS' REPLY, -Dr van den Hogen suggests that initial screening for cardiovascular risk is done by questionnaire so that only those patients at relatively high risk need be invited for appropriate counselling and advice. This would certainly allow more time to be spent dealing with risk factors rather than simply identifying them. Of course, if questionnaires were sent by post this would not necessarily avoid the problem of inverse care because of the possibility of poor response from those at greatest risk. But the new contract requires that risk factors are sought when patients first register with the practice, and this can be done by the doctor, practice nurse, or reception staff when the patient first visits the surgery. In Berinsfield the intention is to use such records to target intervention. As the proportion of general practitioners that uses computerised records grows, data on which selective screening could be based will be increasingly accessible to general practitioners without special questionnaires. But there are difficulties with this approach. Firstly, the proportion of patients with at least one cardiovascular risk factor is high. We analysed recently data collected from 1047 health checks and found that 451 patients (43 1%) had either a personal history of coronary heart disease, diabetes, hypertension, or long term smoking (more than 10 years) or a family history of premature coronary heart disease (under 50 years). If obesity is added to this list of selective criteria the proportion of patients included is even higher. Secondly, 85 of the 1047 patients had a total plasma cholesterol over 7-8 mmolUl but only 50 (58 8%) had one or more of the selective criteria cited. Thirdly, self reports of smoking elicited by questionnaire may be unreliable. None the less, the separation of screening and intervention, with more emphasis on intensive intervention in selected patients, may be a way forward. In the near future we will be able to address Dr van den Hogen's suggestion directly because we will be analysing interim data from the
"Ox-check" trial. This is a randomised controlled trial of the effectiveness of health checks in reducing cardiovascular risk, and completion of a questionnaire, which includes the specific risk factors mentioned by Dr van den Hogen, is a precondition of recruitment to the study and the offer of a health check. L ROE
M AGASS D WALLER
J MUIR
D MANT
A COULTER
Radcliffe Infirmary, Oxford 0X2 6HE
Meningococcal meningitis SIR,-The editorial by Mr Philip D Welsby and Dr Clayton L Golledge on meningococcal meningitis' merits further comment on indications for hospital admission. In Israel recently experience has been accumulating regarding patients with a genetic predisposition to meningococcal infection due to a deficiency of the terminal complement components (C5-C9). A recent review of 111 survivors of sporadic meningococcal disease in 10 Israeli hospitals found an incidence of C7 and C8 deficiency of 18% in patients of Sephardic origin (and an incidence of 40% in patients of Moroccan origin), and cases among Arabs have also been identified.2 The clinical characteristics of the illness were different in this group, including a higher age at first presentation and a high incidence of recurrence (40%), and a high proportion of the patients had a family history of meningococcal infection.2 Some patients may present with fever and no relevant clinical signs (Malnick and others, unpublished observations), as may also occur in children with meningoccocal infection.' Inherited deficiencies of the complement system that are associated with increased susceptibility to meningococcal infections have also been reported in white and black people4 and may be more prevalent than currently realised. One American study of 20 consecutive patients presenting with a first known episode of meningococcal meningitis disclosed three cases of "idiopathic" terminal pathway protein deficiency.5 Thus a history of meningococcal infection or a positive family history should be added to the criteria for hospital admission suggested by Mr Welsby and Dr Golledge. S D H MALNICK A SCHATTNER Division of Internal Medicine,
Kaplan Hospital,
Rehosot and Hadassah Miedical School, Jerusalem, Israel
laD, Golledge CL. Mleningococcal meningitis. Br.Medj 1990;300:1150-1. (5 May. ) Zimran A, Rudenskv B, Kramer MR, et al. Hereditary complement deficiency in survivors of meningo occal disease: high prevalence of C7/C8 deficiency in Sephardic (Moroccan) Jews. QJfMed 1987;240:349-58. Dashefskv B, Teek DW, Klein JO. Unsuspected meningococcemia. 7Pediatr 1983;103:69-73. Ross SC, Densen P. Complement deficiency states and infection: epidemiology, pathogenosis and consequences of neisserial and other infections in an immune deficiency. Medicine 1984;863:243-73. Ellison RT, Kohler PF, Curd JG, Judson FN, Reller lIB. Prevalence of congenital or acquired complement deficiency in patients with sporadic meningococcal disease. N Engl j Med 1983;308:913-6.
1 Welsby 2
3 4
5
Coronary artery bypass surgery SIR,-In their editorial Mr E W J Cameron and Mr W S Walker draw attention to the important issue of selection of the conduit in coronary artery bypass surgery.' Graft failure is more common with saphenous vein conduits than with internal mammary artery grafts. Consequently we were 1725
surprised by the relatively low proportioi of patients (two fifths) receiving internal mammary artery grafts according to the United Kingdom cardiac surgical register. We analysed the last 100 consecutive patients who underwent coronary bypass surgery in our hospital from December 1989 to May 1990. Ninety nine patients received the following arterial grafts: left internal mammary artery (75); both internal mammary arteries (22); right internal mammary artery (1); both internal mammary arteries and gastroepiploic artery (1). A sequential left internal mammary graft was used in three patients, and a free right internal mammary graft was used in three patients. The mean number of distal venous graft anastomoses was 2 5. The operative mortality was zero. Two patients had postoperative mediastinitis; both recovered well. We hope that our favourable experience with arterial conduits will encourage their increased use. TIMO PELLINEN ANTERO SAHLMAN
TUULA KURKI ARI HARJULA
Heart Centre, Deaconess Medical Centre, Alppikatu 2, SF-00530 Helsinki, Finland I Canmeroni EWJ, Walker WS. Coronary artery bypass surgery. BrMedj 1990;300:1219-20. (12 May.)
Births resulting from assisted conception SIR,-The data on the incidence of congenital malformations in babies resulting from assisted conception as reported by the MRC working party on children conceived by in vitro fertilisation' seem to be lacking in one important respect. Women who conceive after in vitro fertilisation (IVF) or gamete intrafallopian transfer (GIFT) would probably have access to skilled detailed fetal ultrasonography in the second trimester to detect congenital malformations, and pregnancies complicated by serious abnormalities may well have been terminated. While some terminations for malformations (three for anencephaly, one for achondroplasia, and one for translocation) are included it is not clear whether or not the working party made any attempt to ascertain all pregnancies that ended in therapeutic termination because of fetal malformation. Nor are there any obvious data included on any fetal losses that occurred before 28 weeks in pregnancies that may have been complicated by an abnormality. It is only with the addition of these data that the true incidence of congenital malformations in these pregnancies will be known. A precise diagnosis of each malformation is required not only as good clinical practice in order to advise parents but also to stand any chance of assessing whether the handling of gametes or preembryos predisposes to errors in development. For example, it is surprising that achondroplasia was one of the disorders diagnosed early enough to allow a therapeutic abortion; in our experience in many such cases there turns out to be a more severe bone dysplasia on expert assessment. The authors point out that malformations like anencephaly occur more commonly in twin pregnancies, but this mainly applies to monozygotic twins. Presumably most of the twins reported in this series were dizygotic. If, indeed, a considerable number were monozygotic this could have implications with regard to the incidence of malformations and perhaps the adverse effect of gamete handling. Monozygotic twinning can be viewed as a malformation and therefore the incidence, for which there are good control data, should be monitored. While we agree that it would be difficult to obtain ideally matched control data for all pregnancies, some form of prospective control
1726
study with respect to monitoring fetal malformations may be feasible. The forthcoming Statutory Licensing Authority set up to regulate IVF services (and, it is to be hoped, to monitor GIFT services as well) should take the lead in this matter. LYN CHITTY MARCUS PEMBREY Institute of Child Health, University of London, London WC IN I EH 1 MRC Working Party on Children Conceived by In Vitro Fertilisation. Births in Great Britain resulting from assisted conception, 1978-87. Br MedJ7 1990;300:1229-33. (12 May.)
AUTHORS' REPLY, --The MRC register was set up to establish whether the children resulting from assisted conception differed materially from children conceived normally. Thus we concentrated on viable pregnancies and on health outcomes that are reliably measured in the general population. For congenital malformations the emphasis was therefore on conditions diagnosed in liveborn and stillborn infants and on prenatally diagnosed fetal malformations in pregnancies that were terminated. We agree that our data do not include spontaneous abortions due to malformations and so do not necessarily give the "true" incidence of malformations. Adequate statistics for comparison do not exist, and a different study design would have been necessary to estimate the true incidence of malformations. Every contributing clinic was asked specifically about terminations because of malformations, and it is unlikely that any are missing as the clinics are particularly aware of such events. We agree that the risk of various malformations does differ for monozygotic and dizygotic twins. Virtually all the twins resulting from assisted conception are dizygotic, and so we did not ask specifically about this. Among the 219 sets of twins in the MRC register 54 were both boys, 52 were both girls, and 113 were a boy and a girl, which is consistent with the sex distribution of dizygotic twins. From 1 January 1989 the Interim Licensing Authority requires that data on the outcomes of in vitro fertilisation treatment cycles are collected by the clinics themselves and sent directly to the Office of Population Censuses and Surveys. We are continuing to collect information on babies born before then. VALERIE BERAL
Imperial Cancer Research Fund, Cancer Epidemiology Unit, John Radcliffe Infirmary, Oxford OX2 6HE PAT DOYLE
London School of Hygiene and Tropical Medicine, London WC1E 7HT
Allergy to peanuts SIR,-The editorial by Dr Tony Smith and the short report by Dr E S K Assem and colleagues and Drs Kieron L Donovan and J Peters on peanut allergy highlighted the potential severity of reactions to peanuts and the importance of recognising this sensitivity.`3 To assess the prevalence of IgE antibody to peanut and 10 other food allergens we investigated the serum of 848 patients with suspected food allergy and found that peanut was the most commonly positive allergen (table). When a serum IgE value of radioallergosorbent test grade 3 or above was taken as a strong indicator of clinical importance4 peanut sensitivity occurred in 70 patients, mostly in conjunction with sensitivity to other foods, though eight had positive results for peanut alone. This comparatively high prevalence of seropositivity and the potential serious consequences of inadvertent ingestion should encourage clinicians to include IgE to peanuts in their investigations of food allergy and alert sensitised subjects to check for peanut,
Prevalence of sensitivity to food allergens No of seropositive patients
Allergen Peanuts Wheat Eggs Oats Rye Milk
(No singly seropositive) 70 (8) 67 (5) 65 (18) 60 (5) 48 (1) 46 (9)
Allergen Barley Peas Fish Shrimps Tomatoes
No of seropositive patients (No singly seropositive) 44 34 (6) 24 (5) 16 (2) 16
even in foods where this would be unexpected. For example, peanut, being a legume, can be included in the vegetable oil component of some foods. CHARLES McSHARRY
Clinical Immunology Laboratory, Western Infirmary, Glasgow G 1I 6NT 1 Smith T. Allergy to peanuts. BrMedI1990;300:1354. (26 May.) 2 Assem ESK, Gelder CM, Spiro SG, Boderman H, Armstrong RF. Anaphylaxis induced by peanuts. Br Med 7 1990;300: 1377-8. (26 May.) 3 Donovan KL, Peters J. Vegetable burger allergy: all was nut as it appeared. BrMedj 1990;300:1378. (26 May.) 4 Barnetson RStC, Benton EC, MacFarlane HAF, et al. Food allergy: the scientific approach. Laboratory Practice 1985;34: 84-6.
Management of the upper airway SIR,-In response to Dr D Watson's article' I would like to make the following points, which may benefit both the anaesthetist and the nonanaesthetist. Correct placement of the tracheal tube must always be confirmed as any ensuing cyanosis may wrongly be ascribed to causes other than oesophageal intubation. Auscultation ofthe axillae, as recommended in the article, will certainly detect right bronchial intubation but will fail to detect 15% of accidental oesophageal intubations, whereas auscultation of the axillae and epigastrium will detect all cases.2 The total reliability of the oesophageal detector described by Wee3 in assessing the position of the tracheal tube has been confirmed.2 When connected to a correctly placed tube this device, a 50 ml syringe or rubber bulb, allows free aspiration of air from the trachea; resistance to aspiration indicates oesophageal intubation. These two techniques should prevent undetected oesophageal intubation occurring during resuscitation. The laryngeal mask airway has been used to resuscitate a patient in whom tracheal intubation was not feasible.4 The mask airway provides an alternative to endotracheal intubation when the operator is inexperienced or abnormal anatomy exists. Insertion of the airway is generally easy. The laryngeal mask airway will not fully protect the tracheobronchial tree against aspiration as the cuff does not effect a complete seal. Its use, however, may save lives. I therefore recommend that Wee's oesophageal detector device and the laryngeal mask airway become part of the standard resuscitation equipment available today. MARK LYNCH
Department of Anaesthetics, Royal Free Hospital, London NW3 2QG 1 Watson D. ABC of major trauma: management of the upper airway. BrMedj7 1990;300:1388-91. (26 May.) 2 Andersen KH, Hald A. Assessing the position of the tracheal tube. The reliability of different methods. Anaesthesia 1989;44: 984-5. 3 Wee MYK. The oesophageal detector device. Assessment of a new method to distinguish oesophageal from tracheal intubation. Anaesthesia 1988;43:27-9. 4 Calder I, Ordman AJ, Jackowski A, Crockard HA. The brain laryngeal mask airway. An alternative to emergcncy tracheal
intubation. Anaesthesia 1990;45:137-9.
BMJ
VOLUME 300
-
30 JUNE 1990
